appl hort horticulture of applied... · appl hort issn 0972-1045 horticulture applied journalournal...

Appl Hort

ISSN 0972-1045

HorticultureHorticulture

AppliedAppliedJJournal ofournal of

THE SOCIETY FOR ADVANCEMENT OF HORTICULTUREJournal of

Vol. 11, No. 1, January-June, 2009

JOURNAL OF APPLIED HORTICULTUREVol. 11, No. 1, January-June, 2009

CONTENTSApplications of GIS to Citriculture in South Texas 3Reginald S. Fletcher (USA).

Extracting within-experiment precision of horticultural experiments useful for meta-analysis 10Guido Knapp, Bimal K. Sinha and Dihua Xu (USA).

Is CropSyst adequate for management-oriented simulation of growth and yield of processing tomato ? 17Onofri Andrea, Beccafi chi Catia, Benincasa Paolo, Guiducci Marcello and Tei Francesco (Italy).

Starch degradation characteristics in relation to physiological and biochemical properties during growth and maturation of apple fruit 23Manasikan Thammawong and Osamu Arakawa (Japan).

Water usage and water use effi ciency of drip-irrigated tomato under defi cit irrigation 31Berhanu Kebebew and Ketema Tilahun (Ethiopia).

Effects of antibrowning agents on the shelf life of fresh-cut green jackfruit (Artocarpus heterophyllus Lam.) 41Boodia Navindra, Ruggoo Arvind and Boodoo B. Hassina (Mauritius).

Sucrose synthase and acid invertase activities in relation to the fl oral structures abortion in pepper (Capsicum annuum L.) grown under low night temperature 35Néji Tarchoun Salah Rezgui and Abdelaziz Mougou (Tunisia).

Use of plastic shades to regulate growth of korarima (Aframomum corrorima (Braun) P.C.M. Jansen) 46S. Eyob (Ethiopia).

Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant 50Renu Singh, Ram C. Yadav and Neelam R. Yadav (India).

Effects of the addition of clinker ash to the propagation medium on rooting of rabbiteye blueberry cuttings 54T. Ban, H. Kitazawa, S. Matsumoto, N. Kobayashi, K. Tokumasa, M. Kobatake and T. Asao (Japan).

Persian walnut (Juglans regia L.) grafting as infl uenced by different bench grafting methods and scion cultivars 56Babak Dehghan, Kourosh Vahdati, Darab Hassani and Reza Rezaee (Iran).

Extraction and determination of α-solanine in eggplant fruits 59Zhiwen Li, Baoli Zhou, Yuwen Ding and Xiang Liu (China).

Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress and normal conditions 64M. Altaf Wani, G.R. Lawania, R.A. Bhat, Iffi at Fayaz, A. Nanda and Gazenfar Gani (India).

Growth and interference of invasive Russian knapweed on “Valcatorce INTA” onion 68Carlos R. Bezic, Armando A. Dall Armellina, Omar A. Gajardo, Lucrecia M. Avilés and Silvia L. Cañón (Argentina).

Caulifl ower hybrids for spring production in a southern mediterranean area 73M. Sciortino and G. Iapichino (Italy).

The effect of different sowing times on development and effi ciency of some 78Chinese cabbage varieties (Brassica campestris sbsp. pekinensis)Funda Eryilmaz Acikgoz (Turkey).

Forthcoming PapersISSR, anthocyanin content and antioxidant activity analyses to characterize strawberry genotypes-Samir C Debnath and Elodie Ricard (USA).The effect of high daytime temperatures on inhibition of fl owering in ‘Koroneiki’ olives (Olea europaea L.) under chilling and non-chilling nighttime temperatures-Nasir S.A. Malik and Joe M. Bradford (USA).Irrigation management in greenhouse tomato production on peat substrate-G.A. Peyvast and N. Mayer (Germany).Effi cacy and physical properties of ground, composted rice hulls as a component of soilless substrate for selected bedding plants-C. Song, Paul V. Nelson, Carl E. Niedziela Jr., and D. Keith Cassel (USA).Extraction of anthocyanins from strawberries (Fragaria × ananassa) and analysis using HPLC/Q-TOF mass spectrometry-Qiong Zhang, Hong-qing Wang and Le-xin Jia (China).Low cost hydroponic devices and use of harvested water for vegetable and fl ower cultivation-A. Das, and D. Sing Majhi (India).Evaluation of the FAO CROPWAT model for defi cit-irrigation scheduling for onion crop in a semiarid region of Ethiopia-Samson Bekele and Ketema Tilahun (Ethiopia).An evaluation of health of cactus collection available in botanical garden of Cluj-Napoca, Romania-Gyorgy Feszt and Lucica Mihalte (Romania).Performance of asparagus under the desert conditions of Arabian Peninsula: A pilot study-N. Kameswara Rao and Mohammed Shahid (UAE).Effect of ripening concentrate on ripening and sensory properties of locally cultivated mangoes (Mangifera indica L.)-William Ofori Appaw , Ibok Oduro, William Otto Ellis (Ghana).Horizontal and vertical soilless growing systems under Cyprus conditions-Damianos Neocleous, Charalambos Kaittanis, Nicos Seraphides and Polycarpos Polycarpou (Cyprus).Expression of gene-related anthocyanin accumulation in berries of cv. Tannat (Vitis vinifera L.)- O. Borsani, G. Gonzalez-Neves, M. Ferrer and J. Monza (Uruguay).Caffeine, phenol and protein contents of thirty-seven clones of Nigerian robusta coffee (Coffea canephora Pierre ex. Froehner)-S.S. Omolaja (Nigeria).Growth, nutrient uptake and nitrogen use effi ciency of Ficus hawaii grown on nutrient fi lm techniques (NFT) using different N-sources- Mohamed, M. El-Fouly, A.A. El-Sayed, A.A. Fawzy, B.M. Mansour H.A. Bosila and Hassan. A. Hamouda (Egypt).Infl uence of fungicides and Phytophthora capsici-resistant/tolerant cultivars on bell pepper yield and farm-gate revenues- Jamie, R. Stieg, S. Alan Walters, Jason P. Bond, and M. Babadoost (USA).Growth, yield and productivity responses of okra-pawpaw mixture to sequence of components in S.W. Nigeria-O.O. Olubode, I.O.O. Aiyelaagbe and J.G., Bodunde (Nigeria).Constraints as perceived by the vegetable growers regarding the adoption of IPM technologies in Caulifl ower cultivation: an empirical study- Prabuddha Ray and Sarthak Chowdhury (India).The effect of high daytime temperatures on inhibition of fl owering in ‘Koroneiki’ olives (Olea europaea L.) under chilling and non-chilling nighttime temperatures- Nasir S.A. Malik and Joe M. Bradford (USA).Leaf N and P in different growth habits of peach: Effects of root system morphology and transpiration- Thomas Tworkoski A, Ralph Scorza, and D.Michael Glenn (USA).In vitro fl owering and shoot multiplication of gentiana trifl ora in air-lift bioreactor cultures-Yaser Hassan Dewir, Nisha Singh, Siveshni Govender and Pragashnee Pillay (Egypt).Comparison of drying characteristics of bitter gourd using cabinet drier, microwave drier and combination of cabinet and infrared drier-K.A. Athmaselvi (India).Induction of multiple shoots in Amomum hypoleucum Thwaites – A Threatened Wild Relative of Large Cardamom- M. Bejoy, M. Dan, N.P. Anish, Githa Ann George and B.J. Radhika (India).Effects of grafted eggplants on allelopathy of cinnamic acid and vanillin in root exudates-Chen Shaoli, Zhou Baoli, Wang Ruhua, Xi Haijun (China).Ripening in fruit of three papaya cultivars: Solo Sunrise, Tainung #2 and Red Lady at two temperatures-Protain, S., M. Mohammed and L.A. Wilson (West Indies).

Journal of Applied Horticulture, 11(1):3-9, January-June, 2009

Applications of GIS to Citriculture in South Texas

Reginald S. Fletcher

United States Department of Agriculture, Agricultural Research Service, Kika de la Garza Subtropical Agricultural Research Center, 2413 E. Hwy. 83, Weslaco, Texas 78596, E-mail: reginald.fl [email protected]

AbstractThe South Texas citrus industry needs an inventory of soil properties within existing citrus (Citrus spp.) orchards, wants data at the county level showing soils that are suitable for citrus production, and would value any information related to the establishment of citrus orchards. This study discusses integration of citrus, soil survey geographic data (SSURGO), and U.S. Census spatial and tabular data with geographical information system (GIS) technology for citriculture. For this study, Hidalgo County Texas was evaluated because it is the major citrus producing county in South Texas. The spatial and tabular data and commercial GIS software were used to inventory selected soil chemical and physical properties within citrus groves, to identify orchards that may be affected by urban expansion, and to select potential sites for establishing new citrus orchards. Results indicated that citrus, SSURGO, and U.S. census spatial and tabular data integrated with GIS technology can be a powerful tool for citriculture. The information provided in this study should appeal to producers, extension agents, scientists, and government agencies within the U.S. and abroad.

Key words: Citriculture, citrus, Citrus spp., Geographic Information System (GIS), soil survey geographic data (SSURGO), Topologically Integrated Geographic Encoding and Referencing (TIGER) data, grapefruit, Citrus paradisi, oranges, Citrus sinensis

IntroductionIn the Lower Rio Grande Valley of Texas, the citrus industry needs inventories on chemical and physical properties of soils within citrus groves to assist them in making management decisions. The life expectancy of citrus trees ranges from 30 to 50 years. Establishment of new groves on soils having a high potential for citrus production and a low potential for commercial buildings and homes would be ideal because these locations can remain in citrus for extended periods, reducing the loss of established orchards to urban expansion.

Spatial and tabular soil data are available in digital formats, allowing analysts to incorporate it into a geographic information system (GIS), a computer system designed to store, retrieve, and manipulate geographic data in various formats. The soil survey geographic data (SSURGO) created by the Natural Resource Conservation Service (NRCS) provides the most comprehensive soil information for the United States (U.S.). It is designed for multiple purposes (National Soil Survey Center, 1995). The U.S. Census Bureau has also created spatial and tabular data in digital format to support its mapping needs for the decennial census and other bureau programs. This data contains line features (e.g., roads, railroad, transportation, etc.,); boundary features (e.g., statistical-census tracts and blocks, government-places and counties, administrative-congressional districts, school districts, etc.,); and landmark features (e.g., point-schools and churches, area-park and cemeteries, and key geographic locations-apartment building and factories).

Researchers have used GIS technology and spatial and tabular data provided by government agencies to assist them in assessing and in solving problems affecting agricultural and natural resources. Mandal and Sharma (2006) demonstrated the application of GIS to develop maps of salt-affected soils in India at the regional, state, and local level. Maps were useful for planning and for making

decisions in reclamation and in management of salt affected soils in the Indo-Ganetic Plain. Boonyanuphap et al. (2004) employed GIS-based analysis to identify suitable land for establishing banana plantations in Thailand. Wu et al. (2001) showed that the use of SSURGO data, integrated with remotely-sensed and other GIS layers were benefi cial for planning and for managing natural resources in Finney County Kansas. Wu et al. (1997) also found SSURGO data to be an excellent source for determining erodible areas and for developing conservation practices. For subtropical South Texas, Richardson et al. (1996) utilized GIS technology to estimate sugarcane production, to map soil type, to manage farms, to classify vegetation communities, to enforce the cotton stalk destruction program, and to monitor the migration of Africanized honey bees into the southern U.S.

The development of SSURGO, along with other digital GIS layers, has been an important advancement. Little information is available on integrating SSURGO and U.S. Census data with GIS technology for citriculture. The objectives of this study were to use citrus, SSURGO, and U.S. Census data with GIS technology (i) to inventory selected chemical and physical soil properties within active citrus groves; (ii) to assess potential of soils for establishing citrus orchards; and (iii) to address alternatives for citrus grove establishment. For this study, Hidalgo County Texas was evaluated because it is the major citrus producing county in South Texas.

Materials and methodsCounty description: Hidalgo County is located at the southern tip of Texas (Fig. 1). The climate is subtropical. Annual rainfall average is 58.42 cm. The elevation ranges from 12.2 to 114 m with the greatest elevation occurring in the western half of the county. The most valuable resources are soil, water, natural gas, and the mild climate.

Journal

Appl

Agriculture is a major source of income for the county. For the state of Texas, the highest vegetable production occurs in Hidalgo County. Irrigated land is primarily used for agricultural production. Producers intensely farm the irrigated land, and their methods are highly specialized. Irrigated crop production includes cotton (Gossypium hirsutum L.), citrus, vegetables, grain sorghum

(Sorghum bicolor L.), and sugarcane. Grapefruit (Citrus paradisi L.) and oranges [C. sinensis (L.) Osbeck] are the primary citrus grown in the county. Drainage and salinity are the major soil problems affecting agriculture.

Development of GIS database: Illustrated in Fig. 2 is a fl owchart showing the development of the spatial and tabular citrus-soil database. Spatial and tabular soil data of Hidalgo County, Texas, and a spatial database template were downloaded from the Soil Data Mart website (http://soildatamart.nrcs.usda.gov/), the offi cial NRCS website for SSURGO data. SSURGO is the most comprehensive soil data in digital format provided by NRCS. The spatial component contained information needed to display and to analyze the data with GIS software. The tabular component included a set of American Standard for Information Interchange fi elds and text delimited fi les. The text delimited fi les matched a table within the SSURGO database template, which is a Microsoft Access database fi le containing empty tables. For this study, the soildb_US_2002.mb template (Microsoft Access 2002 format) was downloaded for further analysis. The spatial SSURGO data were transferred to the Manifold GIS (version 7.0) software.

The template table and the text delimited data were merged using Micrsoft Access 2002. The SSURGO database template has a macro (a rule specifying how a certain input sequence should be mapped to an output sequence according to a defi ned procedure) that allows the user to import the text delimited data into the SSURGO database. To use the macro and to complete the importing procedure, the analyst has to use Microsoft Access. Detailed instructions for this procedure were provided in the SSURGO manual.

The majority of the feeder roots for mature citrus trees occur within 61 cm of the soil surface (Ray and Walheim, 1980). For the soil inventory, chemical and physical properties based on this depth were determined with Soil Data Viewer (version 5.1,

Fig. 1. (a) Map of the continental United States showing the location of Texas and the study area (Enclosed in black square). (b) Close-up showing the Lower Rio Grande Valley and the study area-Hidalgo County.

Fig. 2. Flowchart showing steps used to create citrus soil database.

4 Applications of GIS to Citriculture in South Texas

http://nrcs.usda.gov), a freeware designed to create soil-based thematic maps when linked with ArcMap (GIS software) and tabular output for systems not equipped with ArcMap. The tabular data was used in this study. Soil Data Viewer only works with Microsoft Access databases, hence, the need to merge the text with the tabular template using Microsoft Access.

To calculate the information needed for the 0 to 61 cm depth, the soils data had to be aggregated. The following rules were used to estimate the values: (1) the dominant component, (2) the high tie break rule, and (3) converting null values to zeroes. The dominant component aggregation method returns the attribute value associated with the component occupying the highest percent composition in the map unit. The tie-break rule indicates which value should be selected from a set of multiple candidate values, or which value should be selected in the event of a percent composition tie. For this study, the component having the greatest value was used in tie break situations. Interpret nulls as zeroes is used to determine if a null value for a component should be converted to zero before aggregation occurs. This conversion happens only if a soil map unit has at least one component where this value is not null. For this study, null values were not interpreted as zeroes. Aggregated data were transferred to the GIS software and incorporated into the soil spatial data fi le table.

Soil parameters estimated with the Soil Data Viewer were clay content, pH, organic matter, and electrical conductivity. These properties were selected because they are important to citrus production. Brief defi nitions of the selected categories are as follows. Clay is mineral material composed of particles that are less than 0.002 mm in diameter. pH is a measure of the degree of acidity or alkalinity of a soil. Soils with pH values below 6.6 are considered acidic; soils with pH values above 7.3 are characterized as alkaline; and soils with pH values equal to or within the extremes are characterized as neutral (USDA, 1981). Electrical conductivity (EC) is a measure of the concentration of water-soluble salts in soils. Organic matter is the plant and animal residue in the soil at various stages of decomposition. Output for soil properties were manually entered into the soil database residing in the GIS.

Other soil parameters calculated with the Soil Data Viewer included the soil map unit potential for building small commercial buildings and for building dwellings without basements. The dominant component and high tie break rule were used to assign the map units to a class: not limited, somewhat limited, and very limited.

Brief defi nitions for small commercial buildings and dwellings without basements categories are as follows. Small commercial buildings and dwellings without basements are structures less than three stories high and do not have basements. Ratings are based on soil properties that affect the capacity of the soil to support a load without movement and on the properties that affect excavation and construction costs. Properties infl uencing load-supporting capacity include depth to a water table, ponding, flooding, subsidence, linear extensibility (shrink-swell potential), and compressibility (which is inferred from the Unifi ed classifi cation of the soil).

Ratings are both verbal and numerical. Rating class terms indicate the extent to which the soils are limited by all of the

soil features that affect the specifi ed use. Verbal ratings include the following classes, not limited, somewhat limited, and very limited. Not limited encompasses soils having very favourable characteristics for the specifi ed use. Good performance and very low maintenance can be expected. Somewhat limited includes soils having features that are moderately favourable for the specifi ed use. The limitations can be overcome or minimized by special planning, design, or installation. Fair performance and moderate maintenance can be expected. Very limited indicates that the soil has one or more features that are unfavourable for the specifi ed use. The limitations generally cannot be overcome without major soil reclamation, special design, or expensive installation procedures. Poor performance and high maintenance can be expected.

No information related to the potential of soils for citrus grove establishment appeared in the SSURGO tables. This information was available in the hard copy version of the soil survey (USDA, 1981). Soils were grouped into the following classes for citrus grove establishment, high, medium, low, questionable, or not suitable. High potential indicated that the performance is at or above the level of locally established standards, the cost of measures for overcoming soil limitations are judged locally to be favourable in relation to the expected performance or yields, and soil limitations that continue after corrective measures are installed do not detract appreciably from environmental quality or economic returns. Medium potential signifi es that production or performance is somewhat below locally established standards, the costs of measures for overcoming soil limitations are high, or soil limitations that continue after corrective measures are installed detract from environmental quality or economic returns. Low potential indicates that production or performance is signifi cantly below local standards, measures that are required to overcome soil limitations are very costly, or soil limitations that continue after corrective measures are installed detract appreciably from environmental quality or economic returns. Questionable indicates that not enough evidence existed to determine if the land could or could not be used for citrus grove establishment. Unsuitable means that these soils are not adequate for citrus grove establishment.

The soil survey did not group fi ve soil map units into a potential class. Several steps were taken to try to group the map units into a class; the steps are as follows. The descriptions of the map units were evaluated thoroughly and compared with map units having the same surname. For example, the soil survey did not rate the Harlingen clay-urban complex into a potential class for citrus grove establishment. However, it did provide information on Harlingen clay, which received a rating of not suitable for citrus grove establishment. The Harlingen clay-urban complex consisted of urban areas developed on Harlingen clay soil; therefore, the Harlingen clay-urban complex was categorized as unsuitable for citrus production.

If the nonrated soil could not be placed into a category using the surname procedure, then it was assigned to one of two additional classes, not rated-yield reported and not rated-yield not reported. To create these classes, the Soil Data Viewer was used to determine if citrus yields had been reported for the soil map unit in question. If yield information was reported, then the soil was categorized as not rated for citrus grove establishment-citrus yield reported

Applications of GIS to Citriculture in South Texas 5

(NRYR). If yield information was not provided, then the soil map unit was rated as not rated for citrus grove establishment-no citrus yield reported (NRNYR). The suitability data were transferred to the soil spatial table residing in the Manifold software package.

The citrus database was created in a joint venture between U.S. Department of Agriculture-Animal Plant Inspection Service and U.S. Department of Agriculture- Agricultural Research Service. The 2006 updated version of the database was employed in this study. It consists of spatial and tabular data. The former has polygons representing areas planted to citrus trees and the latter contains attributes for the citrus polygons. Spatial and tabular data were imported directly into the GIS software (Fig. 2). Attributes used in this study were citrus type and active orchards.

To extract the urban area information of Hidalgo County, the TIGER (Topologically Integrated Geographic Encoding and Referencing)/Line 108th CD Census 2000 data were downloaded from the U.S. Census Bureau website (www.census.gov). The TIGER/Line fi les were created from the Census Bureau’s TIGER database of selected geographic and cartographic information. TIGER was developed by the U.S. Census Bureau to support the mapping and related geographic activities required by the decennial and economic censuses and sample survey programs. The TIGER data were imported into the GIS software for further analysis (Fig. 2).

For this study, information related to Urban Areas was employed to assist in answering management questions. The Census Bureau uses the term Urban Area (UA) to refer collectively to Urbanized Areas (UZA) and Urban Clusters (UC). Urbanized Areas (UZA) is a statistical geographic entity consisting of a central core and adjacent densely settled territory that together contain at least 50,000 people, generally with an overall population density of at least 386 people per square km. Urban Clusters (UC) is a new statistical geographic entity designated by the Census Bureau for the 2000 Census, consisting of a central core and adjacent densely settled territory that together contains between 2,500 and 49,999 people. Typically, the overall population density is at least 386 people per square km. Urban Clusters are based on Census block and block group density and do not coincide with offi cial municipal boundaries.

The soil data was merged with the citrus data by employing the topology overlay function of the GIS software. This function created new polygons containing soil and citrus attributes. Using the newly created data layer and structure query language, the inventory of the selected chemical and physical properties of soils within citrus groves were completed. Structure query language and visual basic script were used throughout the study to select appropriate information from tables, estimate area, or perform mathematical calculations.

At the county level, a thematic map showing the potential of the soil map units for citrus grove establishment was developed to assist in selecting new areas for establishing citrus groves. A thematic map showing the census urban area data, citrus location data, and high potential soil citrus grove establishment data versus the building site development data was evaluated to determine alternative sites for establishing citrus groves.

Note: The spatial data utilized for the study were projected to a coordinate system and datum by the various entities that created the data. For this study, the fi nal coordinate system and datum used for the maps were Universal Transmercator (Zone 14 N) and North American Datum 1983, respectively.

It was diffi cult to display the citrus polygons onto thematic maps. For illustrative purposes, centroids (points) extracted from the polygons of the citrus-soil spatial layer was used to represent the locations of citrus groves.

Results and discussionCitrus inventory by type: Summarized in Table 1 is the overall statistics created with the combined citrus-soil database. The total area encompassed by active orchards was 9214.7 ha. Grapefruit and oranges were the dominant citrus types, covering 66% and 33% of the total area, respectively. Tangelos (C. x tangelo Ingram and Moore), tangerines (C. reticula Blanco), and lemons [C. x limon (L.) Burm.f.] were included in the citrus survey; their total area of coverage was less than 1%.

Figs. 3-5 show how the GIS can be used to display the spatial distribution of the orchards at the county level based on citrus type. Grapefruit and orange groves were scattered throughout the south central region of the county (Figs. 3 and 4). Lemons, tangelos, and tangerines were established in isolated areas within the county (Fig. 5).

Soil physical data of citrus groves: The soil physical data indicated that producers established their groves on soils with high potential for citrus grove establishment (Table 1). Thirty-four percent of citrus production occurred on soils with lower ratings. The dominant soil mapping units for citrus grove establishment was Brennan fi ne sandy loam, 0 to 1% slopes; followed by Hidalgo sandy clay loam, 0 to 1% slopes; Hidalgo fi ne sandy

Fig. 3. Spatial distribution of active grapefruit orchards within Hidalgo County.


loam, 0 to 1% slopes; Brennan fi ne sandy loam, 1 to 3% slopes; and Raymondville clay loam (Table 1). Raymondville clay loam ranked number fi ve in area listed under citrus production. It had a low potential rating for citrus grove establishment; while the top four soils had a high potential ranking.

Most of the groves appeared on soils that were categorized as somewhat limited to building development, followed by not limited and very limited (Table 1). Groves occurring on soils rated not limited and somewhat limited for building site development may succumb to urban expansion in the future.

Clay content ranged from 70 to 455 g kg-1. Increases in clay content represent heavier textured soils. Eighty-two percent of the groves occurred on soils having clay contents of 186, 203, and 256 g kg-1. Shrink swell capacity of the soils played a role in determining if they were appropriate for agricultural production and for building site development. There was not a pattern between clay content and potential rating for establishing citrus groves.

Soil chemical data-citrus groves: Citrus production occurred on strongly alkaline (pH values: 8.5-9.0) to slightly acid soils (pH values: 6.1-6.5) (Table 1). The majority of the production was established on moderately alkaline soils (pH values: 7.9-8.4; Table 1). At pH values above 7.8, macronutrients such as phosphorus

and micronutrients such as copper, iron, zinc, and manganese are not readily available for plant uptake. Therefore, producers have to supply these nutrients to the trees. Macronutrients such as calcium, magnesium, and potassium would normally be high in these soils.

The statistics indicated that producers have chosen to grow citrus trees on non-saline soils (0 to 0.2 S m-1), limiting the effect of soil salinity on citrus yields. Organic matter content ranged from 0.5 to 20 g kg-1 with the most hectares of citrus appearing on soils having 16.3 g kg-1 of organic matter. Organic matter totals seemed to fl uctuate; for example, in comparison of the top soils by area for organic matter content, the values ranged from 6.6 g kg-1 to 16.3 g kg-1. It is well known by producers in this county that they have to increase the organic matter content in soils used for agricultural production.

Spatial distribution of soil potential at the county level for citrus grove establishment: Fig. 6 shows the spatial distribution of the soil potential for active citrus grove establishment and the location of citrus groves within the county. Soils with high potential for citrus grove establishment occupy most of the county. Based on this map, producers can expand citrus production into the north central section of the county. If they have to settle for

Table 1. Overall statistics for active citrus groves in Hidalgo County, Texas, including area of citrus by variety, soil map unit ranked by area*, soil clay content ranked by area*, soil organic matter ranked by area*, soil potential, soil pH level, soil electrical conductivity, and soil characteristics for building site development (small commercial buildings and dwellings without a basement)Electrical Conductivity Value Range (S m-1) Area (ha)Non-saline 0.00 to 0.20 9151.0 Slightly-saline 0.21 to 0.40 45.6Moderately saline 0.41 to 0.80 1.0Strongly-saline 0.81 to 1.60 17.1

pH pH Values Area (ha) Building site development Area (ha)Slightly acid 6.1 to 6.5 41.1 Not limited 3480.0Neutral 6.6 to 7.3 753.5 Somewhat limited 5240.0Mildly alkaline 7.4 to 7.8 2967.5 Very limited 494.2Moderately alkaline 7.9 to 8.4 5452.1 Strongly alkaline 8.5 to 9.0 0.5

Potential Area (ha) Organic matter g kg-1 Area (ha)High potential 8033.2 15.3 266.6Medium potential 597.6 5.7 276.6Low potential 355.2 16.9 401.9Not suitable 163.1 6.6 2893.2Questionable 17.3 16.3 4705.1NRYR 48.4NRNYR 0.0

Clay g kg-1 Area (ha) Variety Area (ha)193 270.7 Grapefruit 6124.5224 274.5 Oranges 3045.1203 2342.1 Tangelo 38.1256 2363.1 Tangerine 6.8180 2889.2 Lemon 0.2

Soil Map Unit Area (ha)Raymondville clay loam 266.5Brenan fi ne sandy loam, 1 to 3% slopes 331.6Hidalgo fi ne sandy loam, 0 to 1% slopes 2271.5Hidalgo sandy clay loam, 0 to 1% slopes 2351.9Brennan fi ne sandy loam, 0 to 1% slopes 2557.6*Only the top fi ve soils by area are included in the table. NRYR = Not rated yield reported and NRNYR = not rated no yield reported.


using soils with medium potential for citrus grove establishment, then areas in the extreme western and the central north east section of the county are adequate. Soils having low potential would be their last alternative. These areas are scattered throughout the county.

Spatial distribution of high soil potential for citrus vs building site development, urban area, and citrus grove location: A thematic map showing high potential soils for citrus grove establishment versus the building development categories are

shown in Fig. 7. High potential soils were chosen for map development because they were the soils of choice for growers to establish citrus groves (Table 1). Combinations mapped were high potential for citrus grove development and some what

Fig. 7. Thematic map showing the high soil potential for citrus grove establishment vs soil potential for building site development, urban area identifi ed by the 2000 census, and the location of citrus groves based on citrus polygons and soil interactions in Hidalgo County. HPC = high potential for citrus grove establishment, NLB = not limited for building site development, and SWLB = somewhat limited for building site development.

Fig. 4. Spatial distribution of active orange orchards within Hidalgo County.

Fig. 6. Thematic map showing the soil potential for citrus grove establishment and the location of active citrus groves based on citrus polygons and soil interactions in Hidalgo County.

Fig. 5. Spatial distribution of active lemon, tangelo, and tangerine orchards in Hidalgo County.


limited potential for building site development (92300 ha) and high potential for citrus grove establishment and the some what limited for building development combination (71300 ha). None of the soils were classifi ed into the high potential for citrus grove establishment and very limited potential for building site development class. Soils ranked into this category would have been an ideal choice for establishing citrus groves. Results specify that prime land for citrus grove establishment can have a rank of not limited or somewhat limited for building site development.

The urban area 2000 census map overlaid onto the citrus and soil layers was helpful in indentifying citrus groves appearing in urban areas (Fig. 7). Thirty-four percent of the citrus production occurred in urban areas. In these areas, buildings are continuously being built because of the increase of population. Therefore, some of these groves may be lost to commercial building over time. In the future, citrus producers may have to get accustomed to the concept of urban farming, the growing, processing, and distribution of food through intensive plant cultivation in and around cities (Bailkey and Nasr, 1999). This practice could limit treatments commonly used to control pests in citrus orchards. To reduce potential losses to urban expansion, it is believed that new groves should be established as far as possible from the urban area boundaries. It appears that the northwest central and the northeast central regions of the county would be ideal choices for establishing new groves. The overall results of this study concurred with others in that GIS technology is valuable for assessing and for managing agricultural resources (Richardson et al., 1996; Boonyanuphap et al., 2004; Mandal and Sharma, 2006).

Results of this study showed that citrus, SSURGO, and U.S. census spatial and tabular data integrated with GIS technology can be a powerful tool for citriculture. This information was employed to inventory soil chemical and physical properties

within citrus groves of Hidalgo County, Texas, to estimate and to identify orchards that may be affected by urban expansion, and to select potential sites for establishing new citrus orchards. The techniques used in this study and fi ndings of the study should appeal to the citrus industry in the U.S. and abroad.

Acknowledgments: I thank Dr. David Bartells and Mr. Russel Sheets for supplying the citrus database.

ReferencesBailkey, M. and J. Nasr, 1999. From brown fi elds to green fi elds:

Producing food in North American cities. Community Fd. Security News, Fall 1999/Winter 2000: 7.

Boonyanuphap, J., D. Wattanachaiyingcharoen and K. Sakurai, 2004. GIS-based land suitability assessment for Musa (ABB group) plantation. J. Appl. Hort., 6(1): 3-10.

Mandal, A.K. and R.C. Sharma, 2006. Computerized database of salt-affected soils for agro-climatic regions in the Indo-Gangetic Plain of India using GIS. Geocarto Intl., 21(2): 47-57.

National Soil Survey Center, 1995. Soil Survey Geographic (SSURGO) Data Base: Data Use Information. USDA NRCS Misc. Publ. no. 1527. Natural Resources Conservation Serv., Fort Worth, TX.

Ray, R. and L. Walheim, 1980. Citrus, How to Select, Grow, and Enjoy. Hort. Publishing Co., Inc. Tucson, AZ.

Richardson, A.J., C.L. Wiegand, G.L. Anderson and A.H. Gerbermann, 1996. Six exemplary applications of GIS technology to subtropical Texas agriculture and natural resources. Gecarto Intl., 11(1): 49-60.

USDA. 1981. Soil Survey of Hidalgo County, Texas. U.S. Govt. Printing Offi ce, Washington, D.C.

Wu, J., M.D. Nellis, M.D. Ransom, K.P. Price and S.L. Egbert, 1997. Evaluating soil properties of CRP land using remote sensing and GIS in Finney County, Kansas. J. Soil Water Conserv., 52: 352-358.

Wu, J., M.D. Ransom, G.J. Kluitenberg, M.D. Nellis and H.L. Seyler, 2001. Land-use management using a soil survey geographic database for Finney County, Kansas. Soil Sci. Soc. Amer. J., 65: 169-177.


Journal of Applied Horticulture, 11(1): 10-16, January-June, 2009

Extracting within-experiment precision of horticultural experiments useful for meta-analysis

Guido Knapp1, Bimal K. Sinha2 and Dihua Xu2

1Department of Statistics, TU Dortmund University, 44221 Dortmund, Germany, 2Department of Mathematics and Statistics, University of Maryland, Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA

AbstractFor combining results from independent experiments, it is essential that information about the precision of the estimates of treatment effects is available. In publications of horticultural experiments, the results of multiple comparisons tests are often reported without suffi cient information about the precision of the experiments. Based on limited information of the precision of an experiment such as treatments with the same letter are not signifi cantly different, we develop a method for extracting a possible range of the precision of the experiment which can then be used for meta-analysis. The procedure is demonstrated using a real data example where alternatives to methyl bromide are studied in pre-plant soil fumigation. We also provide an R program which computes the possible range of the precision.

Key words: Duncan’s multiple range test, Student-Newman-Keuls multiple range test, Fisher’s LSD test, standardized mean differences, ratio of means, random effects meta-analysis.

IntroductionMeta-analysis (MA) has become a common and widely accepted tool in different spheres of sciences including horticulture for combining tests of signifi cance as well as methods for comparing differences between treatments. Olkin and Shaw (1995) presented methods for the latter, the standardized difference of two normal means is the effect size of interest and they also presented the statistical meta-analysis in the fi xed effects model, that is, they assumed homogeneous effect sizes in all the experiments eligible for meta-analysis. A major obstacle to conducting meta-analysis is the non-availability of appropriate experimental results. As Olkin and Shaw (1995) said “The minimum information required for quantitative research synthesis includes means, sample sizes, and either standard errors or standard deviations.”

Shaw and Larson (1999) conducted a meta-analysis of strawberry yield response to preplant soil fumigation with combinations of methyl bromide-cloropicrin and several alternative systems. They followed the lines of Olkin and Shaw (1995) and an inclusion criterion was that means, sample sizes, and either standard errors or standard deviations of the treatments had to be reported in the published articles. Since the published results of four studies, basically eligible for that meta-analysis, lacked these necessary parameters for inclusion, these four studies were omitted.

The quality of reporting results in the published articles determines the quality of the meta-analysis. Unfortunately, Olkin and Shaw’s summation about the minimum information required for meta-analysis is still often ignored by presenting results of horticultural experiments in publications nowadays. The lack of information about the precisions of the estimated means in most published articles was a major obstacle for the meta-analysis of Porter et al. (2006) on validating the yield performance of alternatives to methyl bromide for preplant fumigation. To circumvent the problem of missing within-experiment precision, Porter et al.

(2006) simply assumed that the between-experiment variability is so large compared to within-experiment variability that the latter one can be ignored in the analysis. Obviously, this assumption is a very critical one and gives all the different experiments the same weight in the meta-analysis.

But sometimes the results of multiple testing procedures for comparing the treatment means are reported using the presentation that groups of means with the same letter are not signifi cantly different. For instance, Student-Newman-Keuls’ or Duncan’s multiple range test are applied in the analysis of horticultural experiments. In this paper we present a method to demonstrate how to extract information on the within-experiment precision when only the results of a multiple range test are reported besides the estimated means of the treatments.

The paper is organized as follows: In Section 2 we briefl y summarize the basic ideas of multiple range tests. In Section 3 we present the extraction method using a simulated data set. Section 4 contains the results of the extraction method using the reported results from two published articles. Since the sample sizes (number of observations or replicates) are often not available from the published articles, we present a meta-analytical approach in Section 5 using the ratio of means as effect size of interest. In Section 6, we give some concluding remarks and show how to extract the precision when simultaneous test procedures like Fisher LSD, Scheffe, or Tukey test were used in the statistical analysis of the horticultural experiments. In the Appendix, an R code is given to extract the precision from experiments when the results of Duncan’s multiple range test or Fisher LSD are known.

Multiple Range TestsLet Yij, i=1, …, r, j=1, …, n, be r independent samples of n independently, normally distributed random variables with a common variance σ2 and expectations E(Yij) = μi, i=1,…r, j=1, ..., n.

Journal

Appl

Following Miller (1981), the basic credo of multiple range tests is: the difference between any two means in a set of r means is signifi cant provided that the range of each and every subset which contains the given means is signifi cant according to an αp-level studentized range test where p is the number of means in the subset concerned.

Let S2 be the error mean sum of squares from one-way analysis of variance and assume S2 is a multiple of a χ2-random variate with r(n-1) degrees of freedom. The αp-level studentized range test is then conducted by comparing the range (divided by S n/ ) of the p means involved with the critical value of the studentized range distribution.

In Student-Newman-Keuls multiple range test, the αp-levels are chosen as (1)

whereas Duncan’s multiple range test uses, see Miller (1981),

(2)

Given a subset of p means, if ( )max

pY denotes the largest mean in this subset and ( )

minpY the smallest one, the corresponding null

hypothesis is then rejected at level αp when

(3)

Extracting method using a simulated data setLet us consider a simulated data set with fi ve treatments each with ten replications. The standard deviation is σ = 10 and we choose the following true means: μ1 = μ2 = 510 and μ3 = μ4 = μ5 = 490.

Then, starting with the global null hypothesis, that is, p = 5, the following hypotheses are sequentially tested:

In case, for instance, the null hypothesis H3,10 cannot be rejected at

level α3, the hypotheses H2,10 and H2,2

0 are also not rejected without further testing.

We used the function RAND(’NORMAL’) in SAS 9.1.3 and the ROUND function to obtain the data given in Table 1.

Using SAS PROC GLM, we obtain an error mean square of S2 = 85.5244. For Student-Newman-Keuls (SNK) and Duncan’s multiple range test, we get the same grouping, see Table 2. The grouping in Table 2 means that treatments with the same letter are not signifi cantly different. Thus we can conclude that we cannot reject and Consequently, we also accept and without

further testing. All the other null hypotheses are rejected. Table 2. Grouping for Student-Newman-Keuls (SNK) and Duncan’s multiple range test

Grouping Mean Treatment A 512.2 1 A 511.4 2 B 492.2 5 B 490.7 4 B 483.9 3

Now, let us assume that only the results from Table 2 are available. The question is: can we extract any information about the within-experiment precision, that is, S n/ , from this table? Note that it holds in the above example

Recall that we reject the null hypothesis at level αp if (4)

that is, if

(5)

or if

(6)

Conversely, we do not reject the null hypothesis at level αp if

(7)

that is, if

(8)

or if

(9)

It is obvious from (5) and (8) that for each subset of p means the critical range is identical, namely

Furthermore, if a hypothesis is rejected at level αp, then we know from (6) that the standard error S n/ is at most

that is, each rejected hypothesis gives an information about an upper bound of the standard error. Calculating the upper bounds from all rejected hypotheses and taking the minimum of all possible upper bounds, we obtain a sharp upper bound of the within-experiment standard error. To facilitate the computation, it is suffi cient to consider only the rejected hypothesis with the smallest range of means for subsets of magnitude p, as this range provides the smallest upper bound for the within-experiment standard error.

2p p … r= , = , , ,

11 (1 ) 2pp p … r−= − − , = , , .

( ) ( )max min

( 1) p

p p

p r nY Y q

S n , − ,

−> .

/

Table 1. Simulated data set for fi ve treatmentsTreatment Observations 1 506 509 502 514 504 513 514 538 506 516 2 501 513 518 505 524 498 516 512 512 515 3 479 488 482 491 487 479 479 504 482 468 4 489 485 490 498 500 487 496 479 501 482 5 473 496 495 498 498 482 501 491 482 507

85 5244 2 92445610

Sn

.= = . .

( ) ( )max min ( 1) p

p pp r nq S nY Y , − ,− > /

( ) ( )max min

( 1) p

p p

p r n

Y Y S nq , − ,

−> / .

( ) ( )max min

( 1) p

p p

p r nY Y q

S n , − ,

−≤ ,

/

( ) ( )max min ( 1) p

p pp r nq S nY Y , − ,− ≤ /

( ) ( )max min

( 1) p

p p

p r n

Y Y S nq , − ,

−≤ / .

( 1) pp r nq S n, − , / .

( ) ( )max min

( 1) p

p p

p r n

Y Yq , − ,

−,

( ) ( )max min

( 1) p

p p

p r nY Y q

S n , − ,

−> ,

/

Extracting within-experiment precision of horticultural experiments useful for meta-analysis 11

p = 5p = 4p = 3p = 2

H :50 2 3 4 5μ μ μ μ μ1= = = =

H :4,10 H :4,2

0 2 3 4μ μ μ μ1= = = 2 3 4μ μ μ μ= = = 5

H :3,10 H :3,2

0 H :3,30 2 3μ μ μ1= = 2 3 4μ μ μ= = 3 4μ μ μ= = 5

H :2,10 H :2,2

0 H :2,30 H :2,4

0 2μ μ1= 2 3μ μ= 3 4μ μ= 4 5μ μ=

Conversely, if a hypothesis is not rejected at level αp, then we know from (8) that the standard error is at least

that is, each non-rejected hypothesis gives an information about a lower bound of the standard error. Calculating the lower bounds from all non-rejected hypotheses and taking the maximum of all possible lower bounds, we obtain a sharp lower bound of the within-experiment standard error. To facilitate the computation, it is again suffi cient to consider only the non-rejected hypothesis with the largest range of means for subsets of magnitude p, as this range provides the largest lower bound for the within-experiment standard error. Let us now apply this method to the results from Table 2. The critical values of Student-Newman-Keuls multiple range test are q5,45,.05=4.0184, q4,45,.05=3.7727, q3,45,.05=3.4275 and q2,45,.05=2.8484 and the critical values of Duncan’s multiple range test are q5,45,.1855=3.1617, q4,45,.1426=3.0919, q3,45,.0975=2.9954 and q2,45,.05=2.8484

In our example only the hypotheses and were tested and not rejected. Consequently, we can extract the following possible lower bounds:

For the determination of the upper bound, we consider, for each p, the smallest observed range of the rejected null hypotheses and calculate the bounds using the critical values of both tests. This results in

The resulting range for possible values of the standard error is [2.4216,5.6989] using Student-Newman-Keuls multiple range test and [2.7709,6.6768] using Duncan’s test.

Note that for p > 2, Student-Newman-Keuls multiple range test is

more informative for the upper bound of the standard error than Duncan’s multiple range test.

In the above calculation we have used the information that the number of replications is ten for each treatment. Sometimes, this information is not available from the published articles and, thus, no information about the error degrees of freedom can be deduced.

In Table 3, critical values of Student-Newman-Keuls multiple range test are presented for p=2(1)10 and several error degrees of freedom (ddf). The corresponding critical values of Duncan’s multiple range test are given in Table 4. For given p and increasing ddf, the critical values decrease but converge to a certain value. In case no information on the ddf are available, a possible choice would be using the limiting value, that is, Cp, ∞,α which can be, for instance, determined with the SAS function PROBMC [e.g., x = PROBMC(’RANGE’, ., 1- alpha , ., p) with alpha = 1-(1-α)p-1 or alpha = α].

Applying the limiting values in the above calculations we obtain the following possible lower bounds :

The possible upper bounds are given as

The resulting range for possible values of the standard error is [2.5004,5.9177] using Student-Newman-Keuls multiple range test, and [2.8440,6.8530] using Duncan’s test. With respect to the upper bound, the use of the limiting critical value is a conservative choice.

We have explicitly demonstrated the idea of the extraction method in this section, but our example is restricted to the situation that the means can be divided into disjunct subgroups of non-signifi cant means (or treatments). In practice, however,

( ) ( )max min

( 1) p

p p

p r n

Y Yq , − ,

−,

Table 3. Critical values Cp, ddf, 0.05 of Student-Newman-Keuls multiple range test

ddf Number of means p 2 3 4 5 6 7 8 9 1010 3.1511 3.8768 4.3266 4.6543 4.9120 5.1242 5.3042 5.4605 5.598420 2.9500 3.5779 3.9583 4.2319 4.4452 4.6199 4.7676 4.8954 5.007950 2.8405 3.4159 3.7584 4.0020 4.1904 4.3437 4.4727 4.5839 4.6814100 2.8058 3.3646 3.6950 3.9289 4.1093 4.2557 4.3785 4.4842 4.57681000 2.7752 3.3194 3.6393 3.8647 4.0379 4.1781 4.2954 4.3962 4.484310000 2.7721 3.3150 3.6338 3.8584 4.0309 4.1704 4.2872 4.3875 4.4751 ∞ 2.7718 3.3145 3.6332 3.8577 4.0301 4.1696 4.2863 4.3865 4.4741

Table 4. Critical values Cp, ddf, 0.05 of Duncan’s multiple range test

ddf Number of means p 2 3 4 5 6 7 8 9 1010 3.1511 3.2928 3.3763 3.4297 3.4652 3.4891 3.5052 3.5156 3.521820 2.9500 3.0965 3.1896 3.2546 3.3026 3.3392 3.3678 3.3905 3.408650 2.8405 2.9876 3.0843 3.1544 3.2082 3.2511 3.2862 3.3155 3.3403100 2.8058 2.9527 3.0505 3.1217 3.1771 3.2219 3.2590 3.2904 3.31731000 2.7752 2.9218 3.0200 3.0925 3.1494 3.1957 3.2345 3.2676 3.296410000 2.7721 2.9188 3.0170 3.0896 3.1466 3.1931 3.2320 3.2653 3.2943∞ 2.7718 2.9184 3.0167 3.0893 3.1463 3.1928 3.2317 3.2651 3.2941

03 30 5 32 10 1 2

H Range SNK DuncanH 2 4216 2 7709H 0 2809 0 2809

Y YY Y

,

,

− . .− . .

1 3

1 4

1 5

2 5

Range SNK Duncan5 7 0426 8 95094 5 6989 6 95363 5 8351 6 67682 6 7407 6 7407

p

Y YY YY YY Y

− . .− . .− . .− . .

03 30 5 32 10 1 2

H Range SNK DuncanH 2 5042 2 8440H 0 2886 0 2886

Y YY Y

,

,

− . .− . .

1 3

1 3

1 4

3 5

Range SNK Duncan5 7 3361 9 16064 5 9177 7 12713 6 0341 6 85302 6 9269 6 9269

p

Y YY YY YY Y

− . .− . .− . .− . .

12 Extracting within-experiment precision of horticultural experiments useful for meta-analysis

the subgroups of non-signifi cant means (or treatments) usually overlap. In the Appendix a computer program is given written in R (R Development Core Team, 2008) which can be used to extract the possible range of standard errors in the general scenario of overlapping subgroups.

Application to real data examplesWe consider some results from Bartual et al. (2002) for demonstrating the extraction method in a real data scenario. Bartual et al. (2002) studied alternatives to methyl bromide in preplant soil fumigation. Seven treatments were investigated, where treatment 1 is a non treated control, treatment 2 is the standard methyl bromide treatment, and treatments 3 to 7 are alternative treatments. We refer to Bartual et al. (2002) for details.

The published article contains the information that the experimental design consisted of two years crop with a complete randomized block with three replications in the fi rst year. The treatments were repeated on the same plot for a second year but with only two replicates. Duncan’s multiple range tests were done for statistical comparison among treatments.

Several outcomes were measured like marketable yield, fi rst quality fruit yield, fi rst quality fruit size, and percentage of second quality fruit yield. The observed outcomes for marketable yield and fi rst quality fruit size along with the results of Duncan’s multiple range tests are reproduced in Table 5.

With seven treatments and three replicates in a completely randomized block design, the denominator degrees of freedom are ddf = 12 for the fi rst year. Only two replicates in the second year provide ddf = 6. Table 5. Marketable yield and fi rst quality fruit size in the fi rst and second year of planting

Treatment Marketable yield First quality fruit sizeFirst year Second year First year Second year

1 319 C 392 C 17.6 C 17.3 B 2 544 A 738 A 19.4 A 19.5 A 3 513 A 683 AB 19.6 A 20.1 A 4 562 A 579 AB 18.7 B 18.2 B 5 554 A 542 BC 18.6 B 18.2 B 6 427 B 410 C 18.6 B 17.7 B 7 284 C 193 D 18.4 B 16.1 C

Applying the extraction method from the previous section, we obtain the standard errors of estimated marketable yield in both years as [14.7927,27.910] (fi rst year) and [44.3336,47.1219] (second year)Replacing the denominator degrees of freedom by infi nity (∞) yields [16.2432,31.0267] (fi rst year) and [54.4815,57.9080] (second year)

The interval for the second year is relatively tight, because we know that the critical range of two means for p = 3 is between 159 and 169.

For the fi rst quality fruit size, the limits of the standard errors are [0.0906,0.2272] (fi rst year) and [0.2466,0.3468] (second year)

Replacing the denominator degrees of freedom by infi nity (∞) yields [0.0994,0.2525] (fi rst year) and [0.2983,0.4329] (second year)


Y1- Y2g= S

Since there is one replicate less in the second year, one expects that the result of the fi rst year is more precise. This is refl ected in the deduced ranges of the standard error.

Note that we can easily extract the error mean sum of squares if the number of replicates (blocks) in a completely randomized block design is known.

A small meta-analysis Olkin and Shaw (1995) used the standardized mean difference as the effect size for combining differences of treatments from several independent experiments. In our scenario, this combination method is possible if all the experiments provide the number of replicates besides the treatment means and the results from the multiple range test. In case the number of replicates (blocks) is unknown, we can use the extracted information of the standard errors to combine the results of several experiments using the ratio of means as the effect size for comparing treatments. Combining results using the standardized mean difference or the ratio of means will be demonstrated in the next two sections. We will consider only the combination of two independent experiments. Let and be generally two effect size estimates of a common effect size, say θ, with estimated variances and , respectively. Then, the combined estimator of θ is given as

with and 2 1ˆ ˆ1w w= − .

Since the extraction method provides a range of possible values of the standard errors, we only will consider point estimates of the common effect θ in the present paper. For further statistical inference on θ, we refer to the textbooks of Hedges and Olkin (1985) and Hartung, Knapp, and Sinha (2008), or to the overview of Olkin and Shaw (1995).

Standardized mean difference Recall that the standardized mean difference is defi ned as

μ1- μ2δ= σ ,

where μ1 is the expected value in the fi rst group (treatment), μ2 the expected value in the second group (treatment), and σ the common standard deviation of both groups (treatments). An estimator of δ, called Hedges’s g, is given as (Hartung et al. (2008))

,

where is the sample mean in the fi rst group (treatment), the sample mean in the second group (treatment), and S the pooled standard deviation of both groups (treatments) or the root error mean sum of squares in case of more than two treatments considered in an experiment. Since g is biased for δ, an approximately unbiased estimator of δ is given as (Hartung et al. (2008))

with N = n1 + n2 and ni, i=1,2, is the number of replicates in the ith group (treatment).

g*= (1- )g 4N - 93

For the variance of g, it holds that (Hartung et al. (2008))

which can be easily estimated by

where, .

Let us now consider the estimated marketable yield in the fi rst and second year from the real data example of Section 4. We fi rst want to determine the effect of the (active) standard treatment (No. 2) compared to the control treatment (No. 1). Since the ranges of the standard errors are [14.7927,27.9101] in the fi rst year and [44.3336,47.1219] in the second year with knowing the number of replicates, the ranges for the root error mean sum of squares are [25.6217,48.3417] (fi rst year) and [62.6972,66.6404] (second year). Consequently, the true observed standardized mean differences (Hedges’s g) lies in the range [4.6544,8.7816] (fi rst year) and [5.1920,5.5186] (second year).

Applying the bias correction to Hedges’s g, see Hedges and Olkin (1985) or Hartung, Knapp, and Sinha (2008), we obtain the possible range of observed values for g* as [3.7235,7.0253] (fi rst year) and [2.9669,31535] (second year)

Note that the larger the estimated standardized mean difference the larger the estimated variance given a fi xed sample size. The estimated variances for Hedges’s g are [5.9247,19.3844] (fi rst year) and [225.6405,254.792] (second year)

For the bias corrected estimates g* we obtain the following estimated variances [4.0318,12.6460] (fi rst year) and [74.3514,83.871] (second year)

Let us combine the results for the marketable yield of the two years considering the extreme cases, that is, using the limits of the extracted intervals. Combining the results using Hedges’s g, we obtain the following range of possible values of the common effect size: [4.6682,8.5509]. Using the bias corrected g*’s yields the range: [3.6846,6.5180].

Ratio of means: Let be the ratio of two (normal) means and

be the logarithm of ρ. An estimate of ξ is readily given as

Using the delta method, an estimate of the variance of p can be deduced as

where S2 is the pooled sample variance of both groups (treatments) or the error mean sum of squares in case of more than two treatments considered in an experiment. Note that for identical replications, that is, n1 = n2 = n, the variance estimate reads

and the knowledge of the standard error besides the treatment means is suffi cient to calculate this variance estimate.

Let us combine the results for the fi rst quality fruit size of the two years for the ratio of means of treatment 2 and 1 using the extracted standard errors with infi nite degrees of freedom, that is, assuming that the number of replicates is unknown. Recall that

the extracted standard errors are [0.0994,0.2525] (fi rst year) and [0.2983,0.4389] (second year)

Since the effect size here depends only on means, we obtain, in contrast to the standardized mean difference, exactly one estimate of the effect size from each experiment. But we get a range of possible values of the estimated variances for estimated effect sizes. For the fi rst year of the fi rst quality fruit size, we get

with

and for the second year

with

Clearly, the weighted average lies between 0.0974 and 0.1197. We obtain the minimum value of all possible weighted averages by using the smallest possible value of and the largest possible value of . Conversely, the maximum value of all possible weighted averages is given by using the largest possible value of and the smallest possible value of . This leads to the following range of possible estimates of the common effect size: [0.0985,0.1067].

Backtransforming the results to the original scale we obtain the range of estimates of ρ as

[exp(0.0985), exp(0.1067)]=[1.1035,1.1126]

Concluding RemarksIn this paper we have demonstrated how information about the within-experiment variability can be extracted when several treatments are compared and only the treatment means and the results of a multiple range test, either Student-Newman-Keuls or Duncan, are reported. Based on the results of the multiple range test we can deduce a lower and an upper bound of possible values of the root error mean sum of squares.

The extracted within-experiment variability can be used in meta-analysis, when the results of several independent experiments should be combined. Possible effect sizes are the standardized mean difference and the ratio of means. For using the standardized mean difference, the number of replicates has to be additionally known to determine the common variance estimate or the error mean sum of squares. Moreover, we can only calculate a range of possible effect size estimates leading also to a possible range of variance estimates. Fortunately, there is a one-to-one relationship between effect size estimate and variance estimate. Using the ratio of means, we always get one estimate of the effect size per experiment but a range of possible variance estimates.

The extraction method described in this paper is a little bit elaborate as the critical ranges of the multiple range tests vary with the number of means. Since Tukey’s and Scheffe’s multiple comparisons or Fisher’s LSD tests are simultaneous comparisons for all treatments and consequently have only one critical range, the interval of the standard error can be easily determined using


the largest range of two means which do not lead to rejection of the null hypothesis and the smallest range of two means which lead to a rejection of the null hypothesis.

To demonstrate the less elaborate extraction method for the simultaneous multiple comparison methods, let us consider Fisher’s LSD test. Let and , i ≠ j, be two sample means. Then Fisher’s LSD test rejects the null hypothesis of equal means if

and tv, α/2 is the upper critical value of tv . If LSD is explicitly given, then it holds

Since the error degrees of freedom ν are unknown, we approximate the standard error by

If LSD is not explicitly given but only the information non-signifi cant then a lower bound of the standard error is

which, when ν is unknown, can again be approximated by

If LSD is not explicitly given but it is known which differences of means are signifi cant and which ones are not, then we proceed as follows. Consider the largest range of two means which do not lead to rejection of the null hypothesis, say , i ≠ j, and the smallest range of two means which lead to a rejection of the null hypothesis, say , l ≠ k. Then we have

Consequently,

and we can approximate the sought-after interval by

ReferencesBartual, R., V. Cebolla, J. Bustos, A.Giner and J.M. Lopez-Aranda,

2002. The Spanish project in alternatives to methyl bromide(2): The case of strawberry in the area of Valencia. Acta Horticulturae, 567: 431-434.

Hartung, J., G. Knapp and B.K. Sinha, 2008. Statistical Meta-Analysis with Applications. Wiley, New York.

Hedges, L. and I. Olkin, 1985. Statistical Methods for Meta-Analysis. Academic Press, Orlando, Fl.

Miller, R.G.J. 1981. Simultaneous Statistical Inference. Springer, New York.

Olkin, I. and D.V. Shaw, 1995. Meta-analysis and its applications in horticultural science. HortScience, 30: 1343-1348.

Porter, Ian J., L. Trinder, D. Partington, J. Banks, S. Smith, M. Hannah and N. Karavarsamis, 2006. Validating the Yield Performance of Alternatives to Methyl Bromide for Pre-Plant Fumigation. UNEP, Nairobi, Kenya.

R Development Core Team 2008. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.

Shaw, D.V. and K.D. Larson, 1999. A meta-analysis of strawberry yield response to preplant soil fumigation with combinations of methyl bromide-chloropicrin and four alternative systems. HortScience, 34: 830-845.

Appendix: Computer code in R

Extracting the Standard Error from Multiple Range Test

Description: multi is used to extract the standard error from Duncan’s Multiple Range Test or Fisher’s LSD Test. Idea: For one study, we compare every possible pair of treatments to fi nd out if they have at least one letter in common. If this happens, we do not reject the null hypothesis for that pair of treatments, thus enabling us to get a lower bound for the standard error. On the other hand, if they have no letter in common, we get an upper bound. Usage: multi(y.mean, y.mrt, method, method=c(“Duncan”, “Fisher”), alpha=.05) Arguments: y.mean: a vector of mean values reported in the study. y.mrt: a vector of statistical comparison results reported in the study. method: method used to report comparison results in the study, either Duncan’s Multiple Range Test or Fisher’s LSD Test. For Fisher’s Test, we only consider the situation that neither LSD nor “not signifi cant” is reported. alpha: level for the reported statistical comparison results, default is .05 Details: # Function for comparing two strings # if they have at least one letter in common return “L”, else return “U” charcomp <- function(x, y) { nsep <- nchar(x) + nchar(y) # summing up number of letters in x, and number of letters in y xy <- c( strsplit(x,split=character(0))[[1]], strsplit(y,split=character(0))[[1]]) # combine x and y into a new vector, letters by letters # for example, x = “abc”, y=”a”, then xy = c(“a”,”b”,”c”,”a”) nbind <- length(unique(xy)) # report number of unique elements in xy # for previous example, will be “a”, “b”, “c”, then return 3. if(nsep==nbind) {return(“U”)} # the two values are equal means x and y have no letter in common else {return(“L”)} }# Function for extracting the standard error # from Duncan’s Multiple Range Test or Fisher’s LSD Test multi <- function(y.mean, y.mrt, method=c(“Duncan”,”Fisher”), alpha=.05) { n <- length(y.mean) n.pair <- choose(n, 2) # number of possible pairs c.pair <- combn(n, 2) # all possible combinations of n choose 2 bound <- decision <- rep(NA, n.pair) if(method==”Duncan”) { y.rank <- 13- rank(y.mean, ties.method=”min”) for (i in 1:n.pair) { j <- c.pair[1, i] k <- c.pair[2, i]


mean.diff <- abs(y.mean[j]- y.mean[k]) rank.diff <- abs(y.rank[j]- y.rank[k]) alpha.p <- 1- (1- alpha)^rank.diff bound[i] <- mean.diff / qtukey(1-alpha.p, rank.diff+1, 100000) decision[i] <- charcomp(y.mrt[j], y.mrt[k])} } if(method==”Fisher”) {for (i in 1:n.pair) { j <- c.pair[1, i] k <- c.pair[2, i] mean.diff <- abs(y.mean[j]- y.mean[k]) bound[i] <- mean.diff /(qnorm(1-alpha / 2) * sqrt(2))

decision[i] <- charcomp(y.mrt[j], y.mrt[k])} } c(max(bound[decision==”L”], na.rm=T), min(bound[decision==”U”], na.rm=T))}Value: The lower bound and upper bound of the standard error will be returned. Example: yield <- c(392, 738, 683, 579, 542, 410, 193) dmrt <- c(“c”, “a”, “ab”, “ab”, “bc”, “c”, “d”) multi(yield, dmrt, method=“Duncan“, alpha=0.05) [1] 54.48149 57.90800


Journal

ApplJournal of Applied Horticulture, 11(1): 17-22, January-June, 2009

Is CropSyst adequate for management-oriented simulation of growth and yield of processing tomato?

Onofri Andrea*, Beccafi chi Catia, Benincasa Paolo, Guiducci Marcello and Tei Francesco

Department of Agricultural and Environmental Sciences, University of Perugia, Borgo XX giugno, 74, I-06121, Perugia (Italy). *E-mail: [email protected]

AbstractThe model CropSyst has proven useful for management-oriented simulations of growth and yield of cereals and other fi eld crops, but no scientifi c information is available with reference to processing tomato. The aim of this paper was to parameterise and validate the crop module of CropSyst for the simulation of potential fruit production in processing transplanted tomato (Lycopersicon esculentum Mill.). Parameterisation and calibration were performed by using fi eld data from an experiment carried out in 1997 in Central Italy. The same set of parameters was validated against fi ve independent experiments, carried out on the same location in 1998, 1999, 2000, 2001 and 2002. The simulation of aerial biomass was always very good, with RRMSE values ranging from 7.5 to 13.4% and modelling effi ciencies (EI) always above 0.976. The simulation of LAI was very good during the fi rst part of growing season (up to 40-50 days after transplanting), while the decreasing trend in the fi nal part of growing cycle was not always reliably simulated. Indeed, RRMSE for LAI ranged from 13.5 to 26.8% and EI ranged from 0.849 to 0.966. The differences between simulated and observed fi nal fruit yield were below 10%, except in one year (18% in 2001), confi rming the practical value of this model, for management and legislative purposes. For research purposes, it is confi rmed that the simulation of dry matter partitioning is a crucial issue in vegetable crops such as tomato, wherein the growth of sources and sinks coexists for a main part of crop cycle.

Key words: Processing tomato, CropSyst, simulation, modelling, dry matter partitioning

IntroductionTomato (Lycopersicon esculentum Mill.) is one of the most widespread summer crops in Mediterranean environments and it would be helpful to make predictions of yield, as affected by farming practices and environmental conditions. Among available models, CropSyst (Stöckle et al., 2003) may be particularly useful for practical applications oriented to fi eld management and decision making, thanks to the very user-friendly interface, to the possibility of including several management events either on specifi c dates or synchronised with crop phenology, and to the possibility of simulating crop rotations.

With respect to other models, CropSyst introduces several conceptual simplifi cations and thus works with a smaller set of input parameters. For example, the core of the simulation engine for crop growth is based on two simple functions for radiation- and transpiration-dependent growth (Stöckle and Nelson, 2003), which rely on two input parameters, i.e. the light-to-biomass conversion coeffi cient (LtBC, as kg MJ-1), and the water-to-biomass conversion ratio (BTR, as kg m-3 kPa).

The approach to dry matter partitioning is also very simple and based on one empirical equation, with two main input parameters, the ‘leaf area/plant biomass’ ratio at the early growth stages (LAR, as m2 leaves kg-1 plant) and the stem-leaf partition coeffi cient (SLP, as m2 kg-1), that accounts for the sharp decline of LAR as biomass accumulates over time (Stöckle and Nelson, 2003). On the other hand, dry matter partitioning to commercial yield is very simply simulated by multiplying fi nal accumulated biomass by the harvest index (HI), eventually corrected by water stress during fl owering and fruit ripening.

It has been shown that LtBC, BTR, LAR and SLP, together with other phenological parameters, are those that more strongly affect simulation results and thus must be chosen with care (Confalonieri and Bechini, 2004; Donatelli et al., 1997; Pala et al., 1996).

The above-mentioned simplicity may be regarded as an advantage, because CropSyst is easily parameterised and calibrated. This may contribute to a high level of diffusion, outside research institutions and with very practical aims (legislative support, technical advice and so on). However, the use of CropSyst without an appropriate validation may easily lead to unreliable conclusions.

This is particularly true for tomato and other vegetable crops that show several ecophysiological differences with respect to cereals and other fi eld crops, wherein CropSyst has been more extensively validated. In particular, the conceptual shortcuts introduced by CropSyst with reference to dry matter partitioning may hold for cereals, but may represent a problem in tomato, wherein the development of sources and sinks overlaps for a main part of crop cycle. It should not be forgotten that models developed for specifi c use in tomato adopt a more complex approach to biomass partitioning with respect to CropSyst (Van Keulen and Dayan, 1993; Heuvelink, 1996; Scholberg et al., 1997; Heuvelink, 1999; Ramirez et al., 2004; Boote and Scholberg, 2006). Scientifi c information on the reliability of CropSyst simulations with reference to growth and yield of tomato is not available.

Therefore, the aim of this study was to parameterise and validate the crop module of CropSyst, by using a series of fi eld experiments carried out in Central Italy, on conventionally grown processing tomato.

Fig. 1. Relationship between above ground biomass and LAI. Symbols show observed data (1997), solid line shows fi tted curve.

Materials and methodsField experiments: Six fi eld experiments were carried out in processing tomato (cv. PS1296) from 1996 to 1997 and from 1999 to 2002, at Papiano (Perugia, Central Italy, 43°N, 165 m a.s.l.) on a silty-loam soil with 1.3 % organic matter. These experiments compared development and growth of processing tomato at different N-fertilisation and density levels, on plots of 100 m2 size. Results have been already published elsewhere (Tei et al., 1999, 2001, 2002; Benincasa et al., 2006). From these experiments, we selected only those experimental treatments wherein the crop was grown following ordinary practices with reference to seedbed preparation, transplanting (from 25 May to 3 June), plant density (always 3.2 plants m-2 with rows 0.9-1.2 m apart), N-fertilisation (always 200 kg N ha-1) and pest control. No limitations of growth were introduced by nutrient shortage, weeds or pests. All the details can be found in the cited papers.

In all the experiments 4-6 plants per plot were sampled throughout the growing season at approximately weekly intervals. At each sampling date, above ground dry weight and leaf area index (LAI) were determined. The main phenological indexes were also recorded as well as fi nal commercial fruit yield at harvest time.

Daily meteorological data (maximum and minimum temperature, rainfall, global solar radiation and wind speed) were also collected from a station inside the experimental site.

Parameterisation, calibration and validation of CropSyst: Simulations were run by using CropSyst version 3.04.08 (29 March 2005). Potential evapotranspiration was estimated by using the Priestley-Taylor equation, while soil water redistribution was simulated by the cascade model.

CropSyst was parameterised by using default values, literature data and the experimental dataset obtained in 1997. A LtBC value of 2.4 g MJ-1 was chosen from Cavero et al. (1998), that is perfectly in line with values found by Scholberg et al. (2000a, b) and Tei et al. (2002). Optimal mean temperature for plant growth was set at 20 °C (Boote and Scholberg, 2006), while a value of 0.55 was chosen for the average extinction coeffi cient (k), following Ramirez et al. (2004) and Tei and Guiducci (unpublished data; see also Acock et al., 1978; Jones et al., 1991; Cavero et al., 1998).

On the other hand, LAR and SLP were estimated by fi tting into the dataset of 1997 the following equation (Stockle et al., 2003):

LAI = LAR · DW (1)1+SLP · DWwhere DW is the accumulated above-ground biomass (kg m-2), as recorded with increasing LAI values.

Initial and maximum LAI were taken from the 1997 experiment, as well as fi nal LAI values with respect to maximum LAI. The harvest index (HI) was obtained from the same experiment, as the observed ratio between dry commercial fruit yield and total dry biomass at fi nal harvest (HI = 0.67). Such values are consistent with fi ndings of Scholberg et al. (2000a) and Battilani (2006).

Also the phenological parameters required by CropSyst, i.e. Growing Degree Days [GDD = Tmean - Tbase; where Tmean = (Tmax + Tmin)/2] for fl owering, for reaching maximum LAI and

for physiological maturity, were calculated from data observed in 1997 (Stöckle and Nelson, 2003). A Tbase of 10°C (Scholberg et al., 2000a) and a Tcutoff of 35°C (Boote and Scholberg, 2006) were assumed.

All the other parameters were initially set to default values and simulations were performed by using the dataset of 1997 to calibrate the BTR and maximum water daily uptake (mm).

After calibration, the model was validated by applying the calibrated set of parameters to all the other experiments (1996, 1999, 2000, 2001, 2002). The same set of parameters was always used, except for the initial LAI, that was regarded as an input datum for the model and was always set to the observed value. Such a decision was taken considering that in transplanted tomato the size of plants at transplanting may be very different from year to year; thus the same initial LAI does not hold in practice and may lead to considerable differences in modelling results (Tei et al., 1996a, b)

The agreement between observed and predicted values was expressed by using the Relative Root Mean Squared Error (RRMSE: minimum and optimum value = 0) and the modelling effi ciency (EF: optimal value = 1), as indicated by Martorana and Bellocchi (1999).

Results and discussionThe whole set of crop parameters as used in simulations is reported in Table 1. The calibrated value for the water to biomass conversion coeffi cient (BTR) is much higher than those reported in the CropSyst manual (maxima of 6.0 and 8.5 kg m-3 kPa respectively for C3 and C4 species; Stöckle and Nelson, 2003). Unfortunately, it was not possible to lower the BTR value below 10.72 kg m-3 kPa without severely degrading the quality of simulations.

The relationship between LAI and accumulated dry biomass (Eq. 1) in 1997 (Fig. 1) was used to obtain a reliable estimate of LAR and SLP as shown by low standard errors (12.194 ± 0.873 and 1.368 ± 0.249).

18 Simulating growth and yield of processing tomato with CropSyst

l

Days after transplanting

Aer

ial b

iom

ass

(t dm

ha

)-1

Fig. 2. Above-ground total dry biomass of processing tomato during six fi eld experiments. Symbols show observed data, solid lines show CropSyst simulations by using the parameters reported in Table 1. Vertical lines show standard errors.

The parameterisation of CropSyst led to a good simulation of aerial biomass (Fig. 2). As expected, the simulation is particularly good for the 1997 data, that were used for calibration (Table 2), but simulations were also good for 1996, 1999, 2000, 2001 and 2002 (Table 2).

Simulations of LAI are not as good as those for aerial biomass

(Fig. 3). The time course of LAI is always well simulated at the beginning of crop cycle, approximately until 40-50 days after transplanting, while the simulation quality decreases afterwards and, in some cases, the maximum LAI and the fi nal decreasing trend are not fully reproduced by CropSyst. As a consequence, RRMSE and EF values are less favourable than those observed for aerial

Simulating growth and yield of processing tomato with CropSyst 19

biomass. A similar behaviour was observed when modelling the growth of sweet pepper (Tei et al., 1996b). This confi rms that the simulation of dry matter partitioning represents a crucial issue with vegetables characterised by the contemporary growth of sources and sinks (Benincasa et al., 2006). Indeed, the empirical approach used in CropSyst may be too simple with respect to the more

mechanistic approaches adopted in other models (Marcelis et al., 1998; Scholberg et al., 1997; Boote and Scholberg, 2006).

However, it should be emphasized that such deviations from observed data occur during a period of time when the impact on biomass yield may not be very relevant, as it would be at the beginning of crop cycle. Furthermore, it is important to notice

Days after transplanting

Leaf

Are

a In

dex

Fig. 3. LAI of processing tomato during six fi eld experiments. Symbols show observed data, solid lines show CropSyst simulations by using the parameters reported in Table 1. Vertical lines show standard errors.


that the year to year variations on LAI growth were reproduced in a satisfactory manner, which is a very interesting and promising result.

In all the cases, the simulation of fi nal fruit yield is very good (Table 2) and differences between observed and simulated values are rather small, always below 10% apart from 2001. This result follows the relatively constant value of the harvest index throughout years, ranging from 0.63 to 0.67.

In conclusion, in spite of the scientifi c shortcuts introduced by CropSyst, it seems that, following a careful parameterisation, the practical value of this model may hold also in processing tomato,

when used for management and legislative purposes. This usage is encouraged by the relatively small set of parameters required to run the simulations and by the very user-friendly interface. The set of parameters hereby reported seems to be quite robust for the environmental conditions of Central Italy and may serve as a useful starting point to run simulations in processing tomato.

AcknowledgementsThe authors wish to thank Dr. Luca Bechini (University of Milano) for critically reviewing the manuscript.

ReferencesAcock, B., D.A. Charles-Edwards, D.J. Fitter, D.W. Hand, L.J. Ludwig,

J. Warren Wilson and A.C. Whiters, 1978. The contribution of leaves from different levels within a tomato crop to canopy net photosynthesis: An experimental examination of two canopy models. Journal of Experimental Botany, 29(4): 815-827.

Battilani, A. 2006. Growth indexes for fertigated processing tomato in a Mediterranean sub-humid area. Acta Horticulturae, 724: 103-110.

Benincasa, P., C. Beccafi chi, M. Guiducci and F. Tei, 2006. Source-sink relationship in processing tomato as affected by fruit and nitrogen availability. Acta Horticulturae, 700: 63-66.

Boote, K.J. and J.M.S. Scholberg, 2006. Developing, parameterising and testing of dynamic crop growth models for horticultural crops. Acta Horticulturae, 718: 23-34.

Table 2. Goodness of simulations expressed as Relative Root Mean Squared Error (RRMSE) and modelling effi ciency (EF) for aerial biomass and LAI (Fig. 1 and Fig. 2), simulated and measured fruit yield (kg d.m. ha-1) of processing tomato during the six experiments. All the simulations are based on the parameterisation in Table 1

Year Aerial biomass LAI Simulated MeasuredRRMSE EF RRMSE EF yield yield

1996 13.4 0.976 17.3 0.92 6509 60821997 8.3 0.991 13.5 0.97 7968 81651999 9.3 0.998 14.9 0.95 9093 91802000 12.4 0.976 17.6 0.94 8999 82572001 12.4 0.976 15.7 0.96 5706 67162002 8.4 0.990 26.8 0.85 6060 6249

Table 1. Crop model parameters for transplanted tomato (cv. PS1296) and source of informationParameter Ref. 1 Value UnitsPhotosynthetic pathway - C3 -Perennial - no -Land use - row crop -Aboveground biomass-transpiration coeffi cient (BTR) 1997 10.72 kg m-3 kPALight to aboveground biomass conversion (LtBC) 1997 2.38 g MJ-1

Actual/potential transpiration ratio limiting leaf area growth D 0.95 0-1Actual to potential transpiration ratio that limits root growth D 0.50 0-1Optimum mean daily temperature for growth (Topt) L 20 °CMaximum water uptake D 7 mm d-1

Leaf water potential at the onset of stomatal closure D -700 J kg-1

Wilting leaf water potential D -1200 J kg-1

Maximum rooting depth L 1 mInitial green LAI 1997 0.010 m2 m-2

Maximum expected Leaf Area Index (LAI) 1997 2.8 m2 m-2

Fraction of maximum LAI at physiological maturity 1997 0.81 0-1Initial “LAI/DW” ratio (LAR) 1997 11.516 m2 kg-1

Stem/leaf partition (SLP) 1997 2.504 1-10Leaf Area Duration 1997 575 °C d-1

Extinction coeffi cient for solar radiation (k) L 0.55 0-1Leaf duration sensitivity to water stress D 0 0-3ET crop coeffi cient at full canopy L 1.05 mm mm-1

Degree days to overcome transplanting crisis 1997 15 °C d-1

Degree days from emergence to fl owering 1997 300 °C d-1

Degree days from emergence to maximum LAI 1997 800 °C d-1

Degree days from emergence to begin fruit fi lling 1997 400 °C d-1

Degree days from emergence to maturity 1997 1300 °C d-1

Base temperature L 10 °CCutoff temperature L 35 °CPhenological sensitivity to water stress D 0 0-3Unstressed harvest index 1997 0.63 0-1

Simulating growth and yield of processing tomato with CropSyst 21

Cavero, J., R.E. Plant, G. Shennan, J.R. Williams, J.R. Kiniry and V.W. Benson, 1998. Application of EPIC model to nitrogen cycling in irrigated processing tomatoes under different management systems. Agricultural Systems, 56(4): 391-414.

Confalonieri, R. and L. Bechini, 2004. A preliminary evaluation of the simulation model CropSyst for alfalfa. European Journal of Agronomy, 21(2): 223-237.

Donatelli, M., C. Stockle, E. Ceotto and M. Rinaldi, 1997. Evaluation of CropSyst for cropping systems at two locations of northern and southern Italy. European Journal of Agronomy, 6(1-2): 35-45.

Heuvelink, E. 1996. Dry matter partitioning in tomato: Validation of a dynamic simulation model. Annals of Botany, 77(1): 71-80.

Heuvelink, E. 1999. Evaluation of a dynamic simulation model for tomato crop growth and development. Annals of Botany, 83(4): 413-422.

Jones, J.W., E. Dayan, L.H. Allen, H. Van Keulen and H. Challa, 1991. A dynamic tomato growth and yield model (TOMGRO). Transactions of the American Society of Agricultural Engineers, 34: 663-672.

Marcelis, L.F.M., E. Heuvelink and J. Goudriaan, 1998. Modelling biomass production and yield of horticultural crops: A review. Scientia Horticulturae, 74: 83-111.

Martorana, F. and G. Bellocchi, 1999. A review of methodologies to evaluate agroecosystem simulation models. Italian Journal of Agronomy, 3(1): 19-39.

Pala, M., C.O. Stockle and H.C. Harris, 1996. Simulation of Durum Wheat (Triticum turgidum ssp. durum) growth under different water and nitrogen regimes in a Mediterranean environment using CropSyst. Agricultural Systems, 51(2): 147-163.

Ramirez, A., F. Rodriguez, M. Berenguel, and E. Heuvelink, 2004. Calibration and validation of complex and simplified tomato growth models for control purposes in the Southeast of Spain. Acta Horticulturae, 654: 147-154.

Scholberg, J.M.S, K.J. Boote, J.W. Jones and B.L. McNeal, 1997. Adaptation of the CROPGRO model to simulate the growth of fi eld-grown tomato. In: Systems approach for sustainable agricultural development: Application of systems approach at the fi eld level (Kropff et al., eds.), Kluwer Academic Publishers, Doordrecht, The Netherlands, 133-151.

Scholberg, J.M.S., B.L. McNeal, J.W. Jones, K.J. Boote, C.D. Stanley and T.A. Obreza, 2000a. Growth and canopy characteristics of fi eld grown tomato. Agronomy Journal, 92: 152-159.

Scholberg, J.M.S, B.L. McNeal, K.J. Boote, J.W. Jones, S.J. Lo Cascio and S.M. Olson, 2000b. Nitrogen stress effect on growth and nitrogen accumulation by fi eld grown tomato. Agronomy Journal, 92: 159-167.

Stöckle, C.O., M. Donatelli and R. Nelson, 2003. CropSyst, a cropping system simulation model. European Journal of Agronomy, 18: 289-307.

Stöckle, C.O. and R. Nelson, 2003. Cropping System Simulation Model: User’s Manual. Washington State University, Biological Systems Engineering Department, Washington.

Tei, F., D.P. Aikman and A. Scaife, 1996a. Growth of Lettuce, Onion and Red Beet. 2. Growth Modelling. Annals of Botany, 78(5): 645-652.

Tei, F., A. Onofri and M. Guiducci, 1996b. Simulation of sweet pepper growth based on INTERCOM: validation of the model in central Italy. In: “System analysis and simulation in Agricultural Sciences: a contribution of the RAISA Project” (Eds F. M. J. Goudriaan and H. H. van Laar), Firenze, 25-33.

Tei, F., P. Benincasa and M. Guiducci, 1999. Nitrogen fertilisation on lettuce, processing tomato and sweet pepper: yield, nitrogen uptake and the risk of nitrate leaching. Acta Horticulturae, 506: 61-67.

Tei, F., P. Benincasa and M. Guiducci, 2001. Determination of a critical nitrogen dilution curve for processing tomato. In: 2nd International Symposium on Modelling Cropping Systems, Book of proceedings (Bindi M., Donatelli M., Porter J., Van Ittersum M.K eds), Florence (Italy), 33-34.

Tei, F., P. Benincasa and M. Guiducci, 2002. Critical nitrogen concentration in processing tomato. European Journal of Agronomy, 18: 45-55.

Van Keulen, H. and E. Dayan, 1993. TOMGRO, a greenhouse tomato simulation model. Wageningen, The Netherlands: Simulation report CABO-TT No. 29, Wageningen University and Research Centre, 1-48.



Starch degradation characteristics in relation to physiological and biochemical properties during growth and maturation of apple fruit

Manasikan Thammawong* and Osamu Arakawa**

Faculty of Agriculture and Life Science, Hirosaki University, Bunkyocho, Hirosaki 036-8561, Japan. E-mail: *[email protected], **[email protected]

AbstractFruit maturity indices, i.e. respiration rate and ethylene production, amylose (AM) and amylopectin (AP) content, total hydrolytic activity, and sugar content were investigated during the growth and maturation of ‘Tsugaru’ (early-maturing) and ‘Fuji’ (late-maturing) apples (Malus domestica Borkh.). Different starch degradation characteristics during the growth and maturation processes were observed between ‘Tsugaru’ and ‘Fuji’. By iodine staining, the loss of starch in ‘Tsugaru’ was observed earlier than in ‘Fuji’. The different degradation patterns of starch were also demonstrated through the observations on AM and AP content. In ‘Tsugaru’, AM and AP degraded rapidly between 95 to 110 days after full bloom (DAFB) and almost all starch were lost rapidly at 125 DAFB with simultaneous increases in rate of respiration and production of ethylene. However, in ‘Fuji’, starch degraded gradually throughout growth and maturation process and was clearly degraded at 170 DAFB with a low level of ethylene production and decreased respiration. In both the cultivars, content of AM and AP were highest in the outer cortex and lowest in the inner cortex. Starch degradation was observed simultaneously in 3 different tissue zones and there was little difference in the total hydrolytic activity among tissue zones in both cultivars. These results suggest that starch hydrolysis in the apple fl esh began simultaneously rather than preferentially in any one tissue zone. For sugar content, although differences among tissue zones were not clear, it increased distinctly with loss of starch content. Moreover, sugars from the degradation of accumulated starch and sugar translocation seem to infl uence mainly the sweetness quality as the fruit ripens.

Key words: Amylopectin, amylose, Malus domestica Borkh., starch degradation, total hydrolytic activity.

IntroductionAs starch is degraded when the fruit ripens (Blankenship and Unrath, 1988; Dinar and Stevens, 1981; Prabha and Bhagyalakshmi, 1998), determination of starch hydrolysis by iodine staining (SI) is widely used as a maturity index of an apple to provide an estimation of the starch content (Fan et al., 1995; Ingle and D’Souza, 1989; Lau, 1988).

The starch index value provides valuable guidance to the level of fruit maturity and to the appropriate time of harvesting for immediate consumption or for long-term storage. Despite the common use of SI to monitor starch loss during fruit maturation, SI results varied widely between cultivars and it has been reported that it does not relate well to starch concentration (Fan et al., 1995; Watkins et al., 1993).

Starch concentration of fruit fl esh differed between tissue zones, with the highest starch content in the outer cortex and the lowest content in the core (Brookfi eld et al., 1997; Ohmiya and Kakiuchi, 1990). Apple starch consists of amylose (AM) and amylopectin (AP) which react differently with I2-KI solution. AM is reacted most efficiently with I2-KI to produce blue-black pigment (Fan et al., 1995; McCready and Hassid, 1943). The different degradation patterns of starch in the fruit cortex may be due to the different ratio of AM and AP. However, the relationship of starch degradation pattern and AM or AP content in the fruit fl esh has not been clearly explained.

Although maturity indices of the apple such as respiration rate, ethylene production and starch rating had been extensively studied (Dinar and Stevens, 1981; Kader, 1985; Prabha and Bhagyalakshmi, 1998; Watkins, 2003), other aspects including AM and AP composition, sugar content and the enzyme actions, necessary to understand starch degradation mechanisms and quality development of fruit fl esh, are still poorly understood.

Knowledge of the relationship between these properties would be valuable for the improvement of fruit quality, not only for plant physiologists but also for pomologists and fruit growers. Therefore, the objective of this study was to investigate the degradation mechanism of starch among different tissue zones of the apple fruit fl esh during growth and maturation in relation to the physiological and biochemical properties using ‘Tsugaru’, which produces large amounts of ethylene, and ‘Fuji’, which produces very low amounts of ethylene.

Materials and methodsTwo cultivars of apples, ‘Tsugaru’ fruits from a 6-year-old tree, and ‘Fuji’ fruits from a 31-year-old tree grafted on Marubakaido (Malus prunifolia Borkh.) rootstock, were obtained from the experimental orchard of the Faculty of Agriculture and Life Science, Hirosaki University, Japan. Fruits were picked on fi ve to six different days from July to November 2005, at 10-15 days intervals. The ‘Tsugaru’ fruits were harvested on 70, 80, 95, 110, 125 days after full bloom (DAFB); and the ‘Fuji’ fruits on 95,

Journal

Appl

110, 125, 140, 155, 170 DAFB. Fifteen fruits of each cultivar per harvesting crop were used. The last harvest of each cultivar was a week before commercial harvest.

Measurements

Respiration rate, ethylene production and starch rating: To determine CO2 production, each fruit was weighed and sealed in 1.4 L plastic boxes for 2 h, and 1 mL of headspace gas was injected into a gas chromatography (model GC-18A, Shimadzu Co., Ltd., Japan) equipped with a molecular sieve column (60/80 mesh, GL Sciences, Inc.), and a thermal conductivity detector. Helium was the carrier gas. The injector, oven, and detector temperatures were set at 60oC. To determine the ethylene content, 1 mL of the headspace gas was removed and injected into a gas chromatography (model GC-8A, Shimadzu Co., Ltd., Japan) equipped with an activated alumina column (30/60 mesh, GL Sciences, Inc.) and a fl ame ionization detector. Nitrogen was the carrier gas. The injector, oven, and detector temperatures were set at 120, 100, and 120oC, respectively.

The starch distribution was measured by dipping an apple slice taken from the equatorial region in I2-KI solution (10 g/25 g in 1 L distilled water), and the starch-iodine rating was done using the generic starch-iodine index chart for comparison (Watkins, 2003). This method uses a 1 to 8 scale, with 1 = all starch and 8 = no starch.

Starch, AM, and AP contents: Samples were taken by a cork borer (1 cm ø), and were peeled, cored, and cut. The fruit fl esh between the core and the outer skin were separated into three equal portions, with each sample containing 3-5 g of fruit fl esh. The tissues were frozen in liquid N2 and stored at -80oC before being freeze-dried. Samples were reweighed after freeze-drying, and then ground in a TI-100 vibrating sample mill (Heiko Sample Mill, Heiko Seisakusho Ltd., Japan).

Measurements of starch, AM, and AP were obtained through methods applied in Fan et al. (1995) and Magein and Leurguin (2000). Briefl y, 1 mL of 18% HCl was added to each tube of ground sample (60 to 80 mg). The ground cortex tissue and 18% HCl were thoroughly mixed and then incubated at 20oC for 30 min. 10 mL distilled H2O was added to each tube; the supernatant was mixed and centrifuged at 3,800×g for 10 min. Supernatant of 0.2 to 1.0 mL was used, and 5.0 mL of 1.8% HCl was added to each tube. After mixing, 200 μL of I2-KI (2 mg I2 and 20 mg KI per mL H2O) solution was added. Each sample was mixed, and 10 min later, absorbance was measured at 530 nm (for AP) and 606 nm (for AM) using the spectrophotometer (U-2000 Spectronic, Hitachi, Ltd., Japan).

AM (Sigma Chemical Co., St. Louis) and AP (MP Biomedicals, Inc., Ohio) standards were dissolved in 18% HCl for 30 min and then diluted 10-fold. The two standards were mixed in various ratios, and absorbance was measured at 530 and 606 nm to generate a standard curve. Absorbance coeffi cients (A), for AM and AP standards, were AAM606 = 6.88, AAM530 = 4.54, AAP606 = 5.00, AAP530 = 6.99, and a typical standard curve with an r2 value of 0.92 was obtained. The amount of AM and starch concentration (SC) in the samples were calculated using equations derived by Magel (1991); AP concentration can then be obtained by subtracting the AM concentration from the SC.

Sugar content: For sugar determination, 100 mg of the dried sample was extracted three times, each over a duration of 20 min, with 2 mL of 80% (v/v) ethanol at 80oC. The homogenates were centrifuged at 15,000×g for 10 min to give ethanol-soluble and ethanol-insoluble fractions. The ethanol soluble fractions were pooled and evaporated to dryness with a concentrator, and resolublized in 2 mL of de-ionized water. The soluble fraction was then fi ltered. Glucose, fructose, sucrose, and sorbitol were separated with a high-performance liquid chromatograph (HPLC) (Shimadzu Co., Ltd., Japan). De-gassed, distilled, de-ionized water at 1 mL min-1 and 80oC was used as the mobile phase. A refractive index detector (RI-98, Laboratory System Co., Ltd., Tokyo, Japan) was used to quantify sugar’content following the separation. Sugar content was determined by comparison with standard samples of known concentration of glucose, fructose, sucrose, and sorbitol.

Activity of total hydrolytic enzyme: The total hydrolytic activities of ‘Tsugaru’ and ‘Fuji’ (inner, middle, and outer parts) were investigated by the method described by Steup (1990). Briefly, apple flesh (approximately 5 g fresh weight) was homogenized in 30 mL ice-cold grinding medium which consisted 50 mM 2-(N-morholino)ethane-sulphonic acid (Mes), brought to pH 6.0 with NaOH, 5mM CaCl2, and 5% (v/v) glycerol. The homogenate was fi ltered through several layers of Miracloth and the fi ltrate was centrifuged for 15 min at 40,000×g. The supernatant was passed through a Sephadex G-25 gel which had been previously equilibrated with grinding medium.

The incubation mixture contained 1 mL 2% (w/v) soluble starch, 0.9 mL grinding medium, and 0.1 mL of the fi ltered supernatant. A blank was prepared by mixing 2% (w/v) soluble starch with an equal volume of grinding medium. Mixtures were incubated at 30oC. At 5 min, the incubation mixture was added to an equal volume of alkaline colour reagent, mixed thoroughly, and heated for 5 min in a boiling water bath. Samples were then cooled to room temperature and stored for at least 30 min. Absorbance at 546 nm was measured against a reference (2 mL blank plus 2 mL alkaline colour reagent, treated as above). The alkaline colour reagent was prepared by dissolving 1 g 3,5-dinitrosalicylic acid in a mixture of 40 mL 1 N NaOH and approximately 30 mL H2O at an elevated temperature. Solid potassium sodium tartrate (3g) was added and dissolved. After cooling to room temperature, the mixture was brought to a fi nal volume of 100 mL.

For calibration purposes, varying amounts of maltose were reacted with the alkaline colour reagent. Samples (2 mL each) which contained 0-2.0 mM maltose, 0.2 mL 2% (w/v) soluble starch and grinding medium were prepared. Each sample was mixed with 2 mL alkaline colour reagent and processed as described above. A typical standard curve of maltose has an r2 value of 0.99 (Fig. 7A).

Data analysis: Analysis of Variance (ANOVA) with Completely Randomized Design (CRD) using tissue zones as a factor was performed using SPSS (SPSS, IL, USA), and Tukey’s multiple-range test was used to test signifi cant difference at the 95% confi dence level of each variable.

ResultsFruit maturity: Starch index value, respiration rate, and ethylene production were investigated during the growth and maturation

24 Starch degradation during growth and maturation of apple fruit

processes of the fruit samples. The starch degradation of fruit fl esh and different degradation patterns between ‘Tsugaru’ and ‘Fuji’ were observed by monitoring the starch index values. Starch in ‘Tsugaru’ started degrading at 95 DAFB, and there was great loss at 110 to 125 DAFB, but the starch content of ‘Fuji’ degraded gradually throughout growth and maturation and obvious loss was observed at 170 DAFB (Fig. 1A).

Starch rating, by iodine staining, was quantifi ed at fi ve and six stages of growth and maturation for ‘Tsugaru’ and ‘Fuji’, respectively (Fig. 1B, 1C). The starch-staining pattern of ‘Tsugaru’ showed slight loss of starch at the core area between 80 and 95 DAFB (Fig. 1B-b, 1B-c). Starch degradation was observed all over the fruit fl esh, with starch content of the core area completely lost at 110 DAFB (Fig. 1B-d) and almost all starch completely lost at 125 DAFB (Fig. 1B-e). Contrarily, starch in ‘Fuji’ was lost

gradually during growth and maturation. Starch of the core area was completely degraded while some loss of starch in the middle cortex was observed at 125 DAFB (Fig. 1C-c). Almost all starch was degraded at the last harvest at 170 DAFB (Fig. 1C-f).

The respiration rate of ‘Tsugaru’ decreased initially, followed by a steep increase between 110 and 125 DAFB. Ethylene production of ‘Tsugaru’ also increased signifi cantly between 110 to 125 DAFB, demonstrating the climacteric rise of the fruit. However, in ‘Fuji’, respiration rate decreased gradually and low level of ethylene production was observed during investigation period (Fig. 1D, 1E).

AM and AP contents: AM content of the two cultivars was highest in the outer cortex and lowest in the inner cortex (P≤0.05). The changes in AM content were observed simultaneously in 3 different fruit tissues; inner, middle, and outer parts. AM content of ‘Tsugaru’ changed slightly at the fi rst two harvest dates, and dropped signifi cantly between 95 to110 DAFB. Almost all starch was degraded at 125 DAFB (Fig. 2A).

However, AM content of ‘Fuji’ was highest at the fi rst harvest (95 DAFB) and then decreased gradually during the investigation

Fig. 1. Starch rating scores (A), iodine staining patterns (B; ‘Tsugaru’ and C; ‘Fuji’), respiration rate (D), and ethylene production (E) of ‘Tsugaru’ and ‘Fuji’ apples. Each value is the mean of fi ve replicates.

Fig. 2 . Amylose content of ‘Tsugaru’ (A) and ‘Fuji’ (B) apples during growth and maturation. Each value is the mean of fi ve replicates. Mean values with the same letter are not signifi cantly different at P=0.05 by Tukey’s multiple-range test.

Starch degradation during growth and maturation of apple fruit 25

period (Fig. 2B). AP content also decreased during growth and maturation; however, there was no signifi cant difference in AP content among tissue zones in ‘Tsugaru’ (Fig. 3; P≤0.05).

Sugar content: Total sugar content (sum of sucrose, glucose, and fructose) of ‘Tsugaru’ and ‘Fuji’ gradually increased during growth and maturation (Fig. 4). The individual sugar component of these two cultivars was shown in Fig. 5 and Fig. 6. Although the difference among tissue zones was not clear, total sugar content of ‘Tsugaru’ increased during growth and maturation, and it increased greatly between 110 and 125 DAFB (Fig. 4A). For the individual sugar content of ‘Tsugaru’, sucrose content was high between 110 and 125 DAFB, and was highest in the outer cortex (Fig. 5A). Glucose content increased slightly, and was highest in the inner tissue zone (Fig. 5B). Fructose and sorbitol contents changed slightly and the difference among tissue zone was not clear (Fig. 5C, 5D).

In ‘Fuji’, total sugar and sucrose contents tended to increase during maturation (Fig. 4B, 6A). The glucose and fructose contents changed slightly with the highest content in the inner cortex (Fig. 6B, 6C). The sorbitol content of ‘Fuji’ was higher than ‘Tsugaru’, and the highest sorbitol content of the inner cortex was observed between 155 and 170 DAFB (Fig. 6D).

Total hydrolytic activity: The total hydrolytic activity of ‘Tsugaru’ changed slightly between 70 to 95 DAFB, but dropped signifi cantly at 110 DAFB. In ‘Fuji’, however, the total hydrolytic activity changed slightly and decreased gradually throughout the growth and maturation processes. There was a small difference observed in the tissue zones among both cultivars (Fig. 7B, 7C; P≤0.05).

DiscussionFrom a previous study, the role of ethylene in starch degradation of a detached apple fruit was shown to differ between cultivars and their harvest stages, and is related to ripening and physiological characteristics of the fruit (Thammawong and Arakawa, 2007). Although starch degradation of the detached fruit has been studied in various apple cultivars, the characteristics of starch degradation in the fruit fl esh on the tree during growth and maturation is not well studied. Moreover, the relationship between cultivar variation and starch degradation mechanism, which includes the role of physiological aspects and cellular components such as starch and sugar content, and enzyme activity in each tissue zone of the fruit fl esh, is still unclear.

Starch index value can be used as a tool to indicate fruit maturity based on the level of starch degradation and hence indicate the appropriate time for harvesting (Kingston, 1992; Knee et al., 1989; Lau, 1988; Watkins, 2003). However, the characteristics

Fig. 3. Amylopectin content of ‘Tsugaru’ (A) and ‘Fuji’ (B) apples during growth and maturation. Each value is the mean of fi ve replicates. Mean values with the same letter are not signifi cantly different at P=0.05 by Tukey’s multiple-range test.

Fig. 4. Total sugar content of ‘Tsugaru’ (A) and ‘Fuji’ (B). Each value is the mean of fi ve replicates. Mean values with the same letter are not signifi cantly different at P=0.05 by Tukey’s multiple-range test.


of starch patterns shown by iodine staining vary according to growing conditions, canopy environment, seasonal changes, and cultivar variation (Reid et al., 1982; Smith et al., 1979; Watkins et al., 1982; Watkins et al., 1993). From our results, ‘Tsugaru’ and ‘Fuji’ revealed different patterns of starch loss. The loss of starch in ‘Tsugaru’ (early-maturing cultivar) was rapid and complete between 110 to 125 DAFB (Fig. 1B). However, iodine staining of ‘Fuji’ showed slight changes in starch degradation until commercial harvesting began. This difference between cultivars suggests that iodine staining is recommended more for determining the maturation of late-maturing cultivars such as ‘Fuji’ rather than the early-maturing ‘Tsugaru’.

During starch degradation, physiological aspects of the ripening characteristics were observed to differ between ‘Tsugaru’ and ‘Fuji’. In ‘Tsugaru’, starch content changed slightly between 70-80 DAFB, but loss of starch content occurred rapidly between 95 to 125 DAFB with simultaneous increases in the rate of respiration and ethylene production (Fig. 1A, 1D, 1E, 2A). On the other hand, in ‘Fuji’, degradation of starch occurred gradually throughout growth and maturation processes with a low level of ethylene production and decreased respiration. As ethylene has been suggested to play a role in stimulating physiological changes and in the conversion of starch to sugar (Kader, 1985;

Watkins, 2003), the rapid change of starch content in ‘Tsugaru’ might be due to the induction of internal ethylene production and respiration rate. Furthermore, the response of the apple fruit to endogenous and exogenous ethylene for climacteric induction and starch degradation varied according to cultivar variations (Thammawong and Arakawa, 2007).

The changing patterns of AM and AP content of the ‘Tsugaru’ and ‘Fuji’ showed the highest accumulated starch in the outer cortex and lowest in the inner cortex. The action of starch degrading enzymes has been suggested to play a role in the loss of starch in the apple fruit (Beck and Ziegler, 1989; Frenkel et al., 1968; Garcia and Lajolo, 1988; Jackson, 2003; Zhang and Wang, 2002). Since it has been suggested that starch hydrolysis generally proceeds from the core (carpel) towards the skin (Poapst et al., 1959), the degradation of AM and AP content in both cultivars was observed simultaneously in 3 different regions of the cortex tissue. However, there was only small difference in the total hydrolytic activity among different tissue zones in ‘Tsugaru’ and ‘Fuji’ cultivars (Fig. 7B, 7C). This suggests that starch degradation began simultaneously rather than preferentially in any one tissue zone. The important factors affecting starch degradation patterns are probably the amount of starch in the tissue and the rate of degradation. If starch concentration in the core region

Fig. 5. Sucrose (A), glucose (B), fructose (C), and sorbitol (D) content of ‘Tsugaru’ during growth and development. Each value is the mean of fi ve replicates. Mean values with the same letter are not signifi cantly different at P=0.05 by Tukey’s multiple-range test.


is very low, the lack of iodine staining will occur sooner rather than later during fruit maturation and ripening. Additionally, this supports the hypothesis that different degradation patterns between ‘Tsugaru’ and ‘Fuji’ during maturation might be due to the physiological aspects of the ripening effect such as rate of respiration and ethylene production.

As the sweetness of a ripe apple fruit is associated to its cellular sugar components, sugar from the degradation of accumulated starch and sugar translocated from the leaves were both taken into account. Jackson (2003) suggested that starch hydrolysis is generally accompanied by the appearance of sucrose which is then slowly hydrolyzed to produce more glucose and fructose. However, Whiting (1970) suggested that the amount of sucrose far exceeded the amount produced by starch degradation alone. Our results revealed that the total sugar content (sucrose, glucose, and fructose) increased during maturation. The individual fructose content was the highest in the fruit fl esh of both cultivars; it might have accumulated from the conversion of translocated sorbitol from leaves by sorbitol dehydrogenase (SDH, EC 1.1.1.14) (Bieleski and Redgwell, 1985; Teo et al., 2006). There was a predominant increase of sucrose content that occurred simultaneously with the loss of starch. Sugar content increased signifi cantly in ‘Tsugaru’ between 110 and 125 DAFB

while a gradual increase during growth and maturation was observed in ‘Fuji’. As such, it can be assumed that the high amount of sucrose in ‘Tsugaru’ may be due to the hydrolysis of accumulated starch in the fruit tissue. However, an attached fruit may also gain sucrose from the sugar translocation process as it is transported to the fruit sink together with sorbitol. The simultaneous accumulation of sugar from translocation during starch degradation in the maturation process is clearly supported by the occurrence of sorbitol translocation in ‘Fuji’ (Fig. 6D). To clarify this phenomenon, however, further study of sugar translocation and individual enzyme expression of each cultivar are required. Moreover, in order to improve fruit eating quality, the accumulation of sugar from both starch hydrolysis and translocation should be taken into consideration.

Overall, physiological and biochemical characteristics of the fruit development and maturation process, including respiration rate, ethylene production, AM and AP contents, and sugar content, seem to account conjointly for the degradation pattern of starch in the fruit fl esh. ‘Tsugaru’, an early-maturing cultivar with a short growth and maturation period, produced a high amount of ethylene and had increased rate of respiration in order to induce degradation of starch and production of sugar to develop fruit sweetness as the fruit ripened. However, in the late-maturing ‘Fuji’, gradual starch

Fig. 6. Sucrose (A), glucose (B), fructose (C), and sorbitol (D) content of ‘Fuji’ during growth and development. Each value is the mean of fi ve replicates. Mean values with the same letter are not signifi cantly different at P=0.05 by Tukey’s multiple-range test.


degradation occured simultaneously with low levels of respiration and ethylene production, and sugar translocation seemed to be the main factor in sweetness development.

RefrencesBeck, E. and P. Ziegler, 1989. Biosynthesis and degradation of starch in

higher plants. Annu. Rev. Plant Physiol. Plant Mol. Biol., 40: 95-117.Bieleski, R.L. and R.J. Redgwell, 1985. Sorbitol versus sucrose as

photosynthesis and translocation products in developing apricot leaves. Austra. J. Plant Physiol., 12: 657-668.

Blankenship, S.M. and C.R. Unrath, 1988. Internal ethylene levels and maturity of ‘Delicious’ apples destined for prompt consumption. J. Amer. Soc. Hort. Sci., 113: 88-91.

Brookfi eld, P., P. Murphy, R. Harker and E. Macrae, 1997. Starch degradation pattern indices; interpretation and relationship to maturity. Postharvest Biol. Technol., 11: 23-30.

Dinar, M. and M.A. Stevens, 1981. The relationship between starch accumulation and soluble solids content of tomato fruits. J. Amer. Soc. Hort. Sci., 106: 415-418.

Fan, X., J.P. Mattheis, M.E. Patterson and J.K. Fellman, 1995. Changes in amylose and total starch content in ‘Fuji’ apples during maturation. HortScience, 30: 104-105.

Frenkel, C., I. Klein and D.R. Dilley, 1968. Protein synthesis in relation to ripening of pome fruits. Plant Physiol., 43: 1146-1153.

Garcia, E. and F.M. Lajolo, 1988. Starch transformation during banana ripening: the amylase and glucosidase behaviour. J. Food Sci., 53: 1181-1186.

Ingle, M. and M.C. D’Souza, 1989. Fruit characteristics of ‘Red Deliciuos’ apple strains during maturation and storage. J. Amer. Soc. Hort. Sci., 114: 776-780.

Jackson, J.E. 2003. Eating quality and its retention. In: Biology of apples and pears. J.E. Jackson (ed.), Cambridge University Press, New York, pp. 341-383.

Kader, A.A. 1985. Ethylene-induced senescence and physiological disorders in harvested horticultural crops. HortScience, 20: 54-57.

Kingston, C.M. 1992. Maturity indices for apple and pear. Hort. Rev., 13: 407-432.

Knee, M., S.G.S. Hatfi eld and S.M. Smith, 1989. Evaluation of various indicators of maturity for harvest of apple fruit intended for long term storage. J. Hort. Sci., 64: 403-411.

Lau, O.L. 1988. Harvest indices, dessert quality, and storability of ‘Jonagold’ apples in air and controlled atmosphere storage. J. Amer. Soc. Hort. Sci., 113: 564-569.

Magein, H. and D. Leurguin. 2000. Changes in amylase, amylopectin and total starch content in ‘Jonagold’ apple fruit during growth and maturation. Acta Hort., 517: 487-491.

Magel, E. 1991. Qualitative and quantitative determination of starch by a colorimetric method. Starch, 43: 384-387.

McCready, R.M. and W.Z. Hassid, 1943. The separation and quantitative estimation of amylose and amylopectin in potato starch. J. Amer. Chem. Soc., 65: 1154-1157.

Ohmiya, A. and N. Kakiuchi, 1990. Quantitative and morphological studies on starch of apple fruit during development. J. Jap. Soc. Hort. Sci., 59: 417-423.

Poapst, P.A., G.M. Ward and W.R. Phillips, 1959. Maturation of McIntosh apples in relation to starch loss and abscission. Can. J. Plant. Sci., 39: 257-263.

Prabha, T.N. and N. Bhagyalakshmi, 1998. Carbohydrate metabolism in ripening banana fruit. Phytochem., 48: 915-919.

Reid, M.S., C.A.S. Padfi eld, C.B. Watkins and J.E. Harman, 1982. Starch iodine pattern as a maturity index for ‘Granny Smith’ apples. 1. Comparison with fl esh fi rmness and soluble solids content. N. Z. J. Agric. Res., 25: 229-237.

Smith, R.B., E.C. Lougheed, E.W. Franklin and I. McMillan, 1979. The starch iodine test for determining stage of maturation in apples. Can. J. Plant Sci., 59: 725-735.

Steup, M. 1990. Starch degrading enzymes. In: Methods in Plant Biochemistry. Vol. 3; Enzymes of primary metabolism. P. J. Lea (ed.), Academic Press Limited, London, pp.103-128.

Fig. 7. Standard curve of maltose at different concentrations (A), and total hydrolytic activity of ‘Tsugaru’ (B) and ‘Fuji’ (C) during growth and development. Each value is the mean of fi ve replicates. Mean values with the same letter are not signifi cantly different at P=0.05 by Tukey’s multiple-range test.


Teo, G., Y. Suzuki, S.L. Uratsu, B. Lampinen, N. Ormonde, W.K.Hu, T.M. Dejong and A.M. Dandekar, 2006. Silencing leaf sorbitol synthesis alters long-distance partitioning and apple fruit quality. Proc. Natl. Acad. Sci. USA, 103: 18842-18847.

Thammawong, M. and O. Arakawa, 2007. Starch degradation of detached apple fruit in relation to ripening and ethylene. J. Jap. Soc. Hort. Sci., 76: 345-350.

Watkins, C.B., M.S. Reid, J.E. Harman and C.A.S. Padfi eld, 1982. Starch iodine pattern as a maturity index for ‘Granny Smith’ apples. 2. Differences between districts and relationship to storage disorders and yield. N. Z. J. Agric. Res., 25: 587-592.

Watkins, C.B., P.L. Brookfi eld and F.R. Harker, 1993. Development of maturity indices for the ‘Fuji’ apple cultivar in relation to watercore incidence. Acta Hort., 326: 267-275.

Watkins, C.B. 2003. Principles and practices of postharvest handling and stress. In: Apples: Botany, production and uses. D.C. Ferree and I. J. Warrington (eds.), CABI Publishing, Massachusetts, pp.585-614.

Whiting, G.C. 1970. Sugars. In: The Biochemistry of Fruits and Their Products, Vol. 1. A.C. Hulme (ed.), Acad. Press, London, pp.1-32.

Zhang, D. and Y. Wang, 2002. β-amylase in developing apple fruits: activities, amounts and subcellular localization. Sci. In China (series C), 45: 429-440.



Water usage and water use effi ciency of drip-irrigated tomato under defi cit irrigation

Berhanu Kebebew1 and Ketema Tilahun2

1Oromia Irrigation Development Authority, Addis Ababa, Ethiopia. 2Haramaya University, Ethiopia. Present address: School of Agricultural and Wine Sciences, Charles Sturt University, Locked Bag 588, Wagga Wagga, NSW 2678, Australia. E-mail: [email protected]

Abstract Effi cient irrigation is essential for sustainable use of available water resources. A fi eld experiment was conducted on two tomato cultivars (Melka Shola and Melkassa Marglobe) and four irrigation defi cit levels (0%ETc, 25%ETc, 50%ETc, and 75%ETc). The objective was to determine crop factor (Kcf) and water use effi ciency (WUE). The Kcf values of 0.62, 0.65, 0.70, and 0.71 during the respective four growth stages of the crop were determined. The highest (91.23 kg ha-1 mm-1) and lowest (81.62 kg ha-1 mm-1) water use effi ciencies were recorded in 25 and 0% defi cit levels, respectively. The yield and WUE of Melka Shola cultivar was higher than that of Melkassa Marglobe. Generally, it was found that irrigating the tomato crop with 75% of ETc (i.e. 25%ETc defi cit) is the best irrigation practice in the area. In terms of both yield and WUE, Melka Shola tomato cultivar was found to perform better than Melkassa Marglobe.

Key words: Crop factor, drip irrigation, Ethiopia, tomato, water use effi ciency.

IntroductionIrrigation is one of the most important inputs for agricultural production. However, limited water resources and increasing water demands for other uses are causing a decrease in the quantity of water available for agriculture. The use of water saving technologies such as drip irrigation is therefore essential. The possibility of applying water at a very low rates offers the drip irrigation system the means to deliver water to the soil in small and frequent quantities at a relatively low cost compared to other pressurized systems (Cetin et al., 2002).

Irrigation scheduling can be established by using several approaches such as soil water balance estimates, plant stress indicator and pan evaporation. Irrigation scheduling with pan evaporation is one of the irrigation scheduling methods that has been used widely because of its simplicity and low cost (FAO, 1995). It can also be operated by farmers. Changes in weather conditions that cause variation in pan evaporation will have a similar, but not identical, impact on potential evapotranspiration from a (reference) crop. As such, pan evaporation (Ep) measurements can be used to estimate both reference evapotranspiration (ETo) using a pan factor (Kp) and potential crop evapotranspiration (ETc) using a crop factor (Kcf) (Paul, 2001). The crop factor use depends on the growth stage of the crop and crop type.

The upper limit for yield is set by soil fertility, climatic conditions and management practices. Where all of these are optimal throughout the growing season, yield reaches the maximum value as does evapotranspiration. Any signifi cant decrease in soil water storage from fi eld capacity water content has an impact on water availability to crops, and subsequently, on evapotranspiration and yield (Vaux and Pruitt, 1983). In order to increase the productivity of irrigation water, there is a growing interest in defi cit irrigation, an irrigation practice where water supply is reduced below

maximum level and mild stress is allowed with minimal effects on yield. Under conditions of scarce water supply and drought, defi cit irrigation can lead to greater economic gains by maximizing yield per unit of water for a given crop.

In order to supplement the income and nutritional intake of Ethiopian farmers who live on very fragmented land holdings of less than a hectare, currently there is a great interest in rain water harvesting for family-level vegetable production. In an effort to use the harvested rain water effi ciently, gravity drip irrigation system is also being made available to the farmers on credit basis. However, there is no documented study on this irrigation package which can help prepare guideline to be used by the farmers. The objectives of this study were to determine (i) crop factor using pan evaporation, and (ii) water use effi ciency of drip-irrigated tomato using defi cit irrigation at Awash Melkassa, Ethiopia.

Materials and methodsSite description: The study was conducted at Melkassa Agricultural Research Center in the Central Rift Valley of Ethiopia. It is located at 8o24´ N latitude and 39o21´ E longitude and has an elevation of 1552 m above mean sea level. The mean annual rainfall is 950 mm. The mean maximum and minimum temperature is 28 and 14oC, respectively. The soil of the experimental site is loam (sand 37%, silt 42%, and clay 21%). The fi eld capacity and permanent wilting point of the soil is 38% and 22%, respectively.

Field experiment: Runoff water was harvested in a dome-shaped under-ground water storage structure. It was then pumped using a treadle pump into an elevated tanker 1.5 m above ground. Treadle pump is a simple, low cost, foot-operated water lifting device. Low head or gravity type drip irrigation set was used for the experiment. In addition to the pump and elevated water

Journal

Appl

tanker, the system has a main line, fi lter, submains, manifi olds, and laterals. Each submain has a length of 3 m and supplies water to four drip laterals spaced 0.80 m apart.

The experimental treatments were factorial, consisting of two tomato cultivars (Melka Shola and Melkassa Marglobe) and four irrigation water defi cit levels: 0 (optimal), 25, 50, and 75% (most stressed) expressed as percentage reductions from the potential crop water requirement ETc. The treatments were conducted under three replications. The selected cultivars are known for their higher yield and disease resistance. The plants were transplanted in plots of 3 x 5 m at 0.30 m plant spacing and 0.80 m row spacing. Each treatment plot consisted of four rows of tomato with total number of 68 plants per plot. Irrigation water was applied equally for all experimental plots for the duration of plant establishment after transplanting i.e. for 10 days. There after, the plots were irrigated with drippers according to their respective treatment levels. Irrigation was carried out with drip emitters of an average fl ow rate 350 mL hr-1 at 1.5 m operating head.

Fertilizer was applied as per agronomists’ recommendation as: DAP was side dressed at a rate of 200 kg ha-1 at transplanting and 100 kg ha-1 urea was applied in split at transplanting and 45 days after transplanting. To protect disease infection, Ridomil Gold RZ 63% was applied as 3.5 kg ha-1. For insect protection Cypermethin or Karate was used at a rate of 100 g ha-1.

Dripper characterization: Dripper emission uniformity (EU) was calculated as follows

(1)

where EU = emission uniformity (%), qn = average low quarter emitter fl ow rate (L h-1), qa = average emitter fl ow rate (L h-1).

The application effi ciency of the drippers was calculated as (Vermeiren, 1998)

(2)

where, Ks is a coeffi cient which expresses the storage effi ciency of the soil as (average water stored in the root volume/average water applied). It takes into account the losses of drip irrigation water application (Ks = 1 or 100% for loam soil) (Vermeiren, 1998).

Crop water requirement and irrigation requirement: Pan evaporation Ep was measured using pan evaporimeter installed just by the side of the experimental plots. Reference crop evapotranspiration ETo was determined using FAO Penman Monteith equation (Allen et al., 1998) as:

ETo = (3)

Where, ETo = reference crop evapotranspiration (mm day-1), Rn = net radiation at crop surface (MJ m-2 day-1), G = soil heat fl ux (MJ m-2 day-1), T = average temperature (oC), U2 = wind speed measured at 2m height (m/s), ea-ed = vapor pressure defi cit (kpa), Δ = Slope vapor pressure curve (kpa oC-1), γ = Psychometric constant (kpa oC-1), 900 = conversion factor. Daily weather data used in Eq. (3)

was obtained from Melkassa Agricultural Research Centre weather station.

Pan factor Kp was determined from Ep and ETo using Eq. (4). Tomato crop coeffi cients available in literature (Allen et al., 1998) were used to determine potential crop water requirement ETc from ETo (Eq. 5). Crop factor Kcf was determined from Ep and ETc using Eq. (7).

ETo = Ep x Kp (4)

ETc = ETo x Kc (5)

ETc = Ep x Kp x Kc (6)

ETc = Ep x Kcf (7)

Crop water requirements determined by conventional methods should be corrected by a reduction factor Kr for drip irrigation. Kr = Gc/0.85 or 1.0, whichever is the smallest where Gc is percentage ground cover (Vermeiren, 1998). Net irrigation water requirement (IRn) was calculated as

(8)

where R is rainfall during the growing period. Gross irrigation water requirement (IRg) can be calculated as

(9)

(10)

(11)

where Ea = irrigation system application effi ciency, A = the wetted area allocated to each plant.

Water use effi ciency: Crops’ response to defi cit irrigation is different due to the difference in their ability to tolerate water stress during their growth period (Vaux and Pruitt, 1983). Crop yield response to water defi cit can be described (Doorenbos and Kassam, 1979) as: (12)

where Ya = actual yield (kg ha-1), Ym = maximum yield (kg ha-1), ETa = actual evapotranspiration (mm), ETm = maximum evapotranspiration (mm) or ETc, and Ky = yield response factor.

The crop water use effi ciency, WUE, can be determined as

(13)

Field data collection: To determine plant growth, water use and water use effi ciency, plants were marked at random locations along the entire plots in each replication. Two plants per row were marked, giving a sample size of eight plants per plot for both tomato cultivars. Plant heights were measured every week

100⎟⎟⎠

⎞⎜⎜⎝

⎛=

a

n

qq

EU

[ ])34.01(

)(273

900)(408.0

2

2

U

eeUT

GR dan

++Δ

−+

+−Δ

EUKE sa *=

RKETIR rcn −= *

RE

KETIR

a

rcg −=

*

RE

KKKEIR

a

rcppg −=

***

RE

AKKEIR

a

rcfpg −=

***

( )m

m

m

a

y

a

a

ETY

ETET

KETY

WUE

⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢

⎣

⎡−

==1

32 Water usage and water use effi ciency of drip-irrigated tomato under defi cit irrigation

⎥⎦⎤

⎢⎣⎡ −=−

ETmETaKy

YmYa 11

starting from the beginning of the treatment period to the end of the midseason of growth stage. The two middle rows of each treatment were harvested and weighed at the end of the season.

Results and discussionDripper characterization: The emitter fl ow rate collected from randomly selected drippers was used to calculate the performance of the irrigation system. The average low quarter fl ow rate (9 out of 36 sample fl ow rates collected in this study) was 311 mL h-1 and the average fl ow rate was 329 mL h-1. The emission uniformity EU was calculated to be 95% (Eq. 1). For the loam soil with Ks = 1, the application effi ciency was found to be 95% (Eq. 2).

Crop water use: The crop water requirement and irrigation requirement for the tomato crop during the growing season determined on two-day basis is presented in Table 1a-d for the months of October, November, December and January. Eqs. 3, 4, 5, 7 and 8 were used to calculate ETo, Kp, ETc, Kcf, and IRn.

The crop coeffi cient Kc values were interpolated for the growing stages of the tomato crop from Doorenbos and Pruitt (1977). During the crop establishment period (until 28 October), equal amount of water (67 mm) was applied to all the treatment plots. The total amount of water applied to the non-stressed (0% defi cit) treatment is 528 mm (Table 1a-d) + 67 mm = 595 mm. The amount of water applied to the 25, 50, and 75% defi cit levels is 467 mm, 331 mm, and 199 mm, respectively.

The pan factor Kp and tomato crop factors Kcf determined for the growth stages and the total growing season are presented in Table 2. The Kcf values are of great importance for farmers interested to optimally schedule the limited harvested water using pan evaporation data. The Kcf value varies during the growing season with the average seasonal value indicating that about 70% of pan evaporation can be considered as potential crop evapotranspiration.

The water use effi ciency of the experiment was signifi cantly

Table 1. Two-day average tomato water requirement at Melkassa Agricultural Research Centera. October (mm day-1)**Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31ETo 6.00 5.59 6.56 6.40 6.49 6.73 6.33 6.43Ep 8.95 7.85 9.18 9.68 9.87 10.98 8.60 8.84Kp 0.67 0.71 0.71 0.66 0.67 0.62 0.74 0.73Kc 0.75 0.79 0.81 0.83 0.85 0.87 0.89 0.91ETc 0.50 0.56 0.58 0.55 0.57 0.54 0.65 0.66Kcf 4.47 4.40 5.32 5.42 5.60 5.93 5.59 5.83R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00IRn 4.82 4.75 5.75 5.85 6.05 6.40 6.03* 6.30* The day on which irrigation treatments started and total amount of water applied upto this day was 67.12 mm.b. November (mm day-1)Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31ETo 5.59 6.61 6.40 6.44 6.44 6.36 5.57 4.00 4.53 4.55 5.07 6.57 7.14 5.62 5.94Ep 8.48 10.17 9.77 9.72 9.30 9.51 7.75 6.16 5.22 7.79 7.28 10.40 10.01 9.02 8.65Kp 0.66 0.65 0.65 0.66 0.69 0.67 0.72 0.65 0.87 0.58 0.70 0.63 0.71 0.62 0.69Kc 0.93 0.95 0.97 0.99 1.01 1.03 1.05 1.07 1.09 1.11 1.13 1.15 1.15 1.15 1.15ETc 0.61 0.62 0.64 0.66 0.70 0.70 0.76 0.68 0.95 0.65 0.79 0.73 0.82 0.72 0.90Kcf 5.17 6.30 5.95 6.54 6.51 6.66 5.89 4.19 4.96 5.06 5.75 7.59 8.20 6.49 7.79R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.60 4.80 0.00 0.00 0.00 0.00 0.00IRn 5.58 6.80 6.43 7.07 7.03 7.19 6.36 4.53 5.35 5.47 6.21 8.20 8.86 7.01 8.40c. December (mm day-1)Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31ETo 5.62 5.16 5.41 5.77 5.58 6.17 5.57 5.48 5.20 5.95 5.34 4.86 5.73 5.92 4.94 5.43Ep 8.74 8.97 8.10 8.86 8.13 9.56 8.66 7.97 9.07 8.13 7.81 7.87 8.25 9.24 7.02 8.08Kp 0.65 0.58 0.67 0.65 0.69 0.64 0.64 0.69 0.57 0.73 0.68 0.62 0.70 0.64 0.70 0.67Kc 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15ETc 0.74 0.66 0.77 0.75 0.79 0.73 0.74 0.79 0.66 0.84 0.79 0.71 0.80 0.74 0.81 0.77Kcf 6.47 5.92 6.24 6.65 6.42 6.98 6.37 6.29 5.98 6.84 6.17 5.59 6.60 6.84 5.69 6.22R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00IRn 6.98 6.39 6.74 7.18 6.94 7.54 6.55 6.80 6.46 7.39 6.66 6.03 7.13 7.38 6.14 6.72d. January (mm day-1)Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31ETo 5.79 6.09 6.56 5.12 5.59 5.53 6.30 6.60 6.16Ep 8.78 8.14 7.94 7.64 8.13 8.27 8.55 7.91 8.43Kp 0.66 0.75 0.70 0.67 0.69 0.67 0.74 0.83 0.73Kc 1.15 1.06 1.06 1.02 0.98 0.93 0.89 0.84 0.80ETc 0.76 0.79 0.74 0.68 0.67 0.62 0.65 0.70 0.58Kcf 6.67 6.46 5.88 5.22 5.45 5.12 5.56 5.54 4.89R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00IRn 7.29 6.97 6.35 5.64 5.88 5.53 6.00 5.98 5.28** The values given in this table are average of two consecutive days expressed as mm day-1 for concise presentation of the data.

Water usage and water use effi ciency of drip-irrigated tomato under defi cit irrigation 33

different (P<0.05) for tomato cultivars (Table 4). The WUE of 25% ETc defi cit was signifi cantly different from the others (Table 3). In both cases, the value was lower for irrigation treatment with high amount of water application. The 25% defi cit level has got higher water use effi ciency which was 1.12 times that of fully irrigated (0% defi cit level) and the 75% defi cit follows it with relative water use effi ciency value of 1.06 times that of fully irrigated tomato (Table 4).Table 3. Infl uence of moisture defi cit levels on yield, water use effi ciency of tomatoParameter Defi cit level (%)

0 25 50 75Total yield (t ha-1) 45.113a 41.555a 24.685b 15.757c

WUE (kg ha-1 mm-1) 81.62a 91.23b 82.96a 86.11a

Relative WUE 1.12 1.02 1.06Ky 0.80b 1.07a 1.08a

Table 4. Yield, water use effi ciency and yield response factor of two cultivars of tomato Characters Tomato cultivars

Melka Shola Melkassa MarglobeTotal yield (t ha-1) 32.688a 30.868a

WUE (kg ha-1 mm-1) 91.67a 79.04b

Ky 0.69a 0.78a

* Means within each row followed by the same letter are not statistically signifi cant at the 1% level according to Duncan’s multiple range test.As crop yield response factor (Ky) increases, the crop water use effi ciency decreases, which in turn implies that benefi t from defi cit irrigation is unlikely. Only those crops and growth stages with a lower crop yield response factor (Ky < 1) can generate signifi cant savings in irrigation water through defi cit irrigation. Seasonal Ky value for tomato is 1.05 (Doorenbos and Kassam, 1979) which is close to the Ky value of the 50% depletion level in this study.

From Table 5, it can be seen that the height of tomato was infl uenced by water defi cit level with a general decrease in crop height as water stress increases.

Pan evaporation data can be used to calculate potential crop evapotranspiration once crop factor is determined. In this study, crop factor values of 0.62, 0.65, 0.70 and 0.71, respectively were determined for crop development, midseason, late season, and total growing season of tomato. As the water stress level increases, the yield of tomato decreases. The 25%ETc defi cit level resulted in the highest water use effi ciency. Melka Shola cultivar has higher yield and water use effi ciency compared to Melkassa Marglobe cultivar. Besides being less water stress resistant, Melkassa Marglobe was easily susceptible to leaf diseases and bacterial wilting during the early growth stages. Yield response factor of Melka Shola is lower than that of Melkassa Marglobe. In the scenario of water shortage, the use of 25%ETc defi cit and Melka Shola cultivar can result in higher water use effi ciency.

AcknowldegementThe authors gratefully acknowledge Dr. Dawit Zerihun and Dr. Tirusew Assefa for suggestions and improvement in the manuscript.

ReferenceAllen, R.G., L.S. Pereira, D. Raes and M. Smith, 1998. Crop

evapotranspitation. FAO Irrigation and Drainage Paper, No. 56, Rome.

Cetin, O., O. Yildirim, D. Uygan and H. Boyaci, 2002. Irrigation scheduling of drip-irrigated tomatoes using class A pan evaporation. Turkish J. Agric. For., 26: 171-178.

Doorenbos, J. and A.H. Kassam, 1979. Yield response to water. Irrigation and Drainage Paper, No. 33. FAO. Rome.

Doorenbos, J. and W.O. Pruitt, 1977. Guidelines for predicting crop water requirements. Irrigation and Drainage Paper, No. 24. FAO. Rome.

FAO. 1995. Irrigation in Africa in Figs.. Water Report No. 7. FAO. Rome.

Paul, E. 2001. Regulated defi cit irrigation and partial root zone drying. Land and Water Resources Department, Australia.

Vaux, H.J. and W.O. Pruitt, 1983. Crop-water production. Advances in Irrigation, 2: 61-93.

Vermeiren, I. 1998. Localized irrigation design, installation, operation, evaluation. FAO Irrigation and Drainage Paper, No. 36. Rome.

Table 2. Average values of the crop water use and crop factor parameters during the growth stages of tomatoParameter Crop growth stages

Development Mid season

Late season

Total growing season

ETo (mm day-1) 6.12 5.56 6.30 5.99Ep (mm day-1) 9.05 8.44 8.18 8.57Kp (mm day-1) 0.68 0.66 0.77 0.69ETc (mm day-1) 5.60 5.50 5.73 5.61Kcf 0.62 0.65 0.70 0.71

Table 5. Average plant growth performance (mm) for defi cit irrigated tomato cultivarsCharacter Defi cit level (%)

0 25 50 75Melka Shola 358 380 357 314Melkassa Marglobe 468 453 368 297

34 Water usage and water use effi ciency of drip-irrigated tomato under defi cit irrigation


Sucrose synthase and acid invertase activities in relation to the fl oral structures abortion in pepper (Capsicum annuum L.) grown under low night temperature

Néji Tarchoun1*, Salah Rezgui2 and Abdelaziz Mougou2

1Centre Régional des Recherches en Horticulture et Agriculture Biologique BP47- 4042 Sousse, Tunisia, 2Institut National Agronomique de Tunisie (INAT) 43, av. Charles Nicolle 1082 cité mahrajène Tunis- Tunisia. *E-mail: [email protected]/ [email protected]

AbstractEffects of low night temperature were investigated on two local hot pepper varieties (‘Beldi’ and ‘Baklouti’) grown at day/night temperature of either low night temperature regime (25°C/10°C) or optimum night temperature regime (25°C/20°C). The negative effect of low night temperature on fl oral structure differentiation was registered on both varieties. The deleterious effect was more sensitive on bud stage than on fl ower buds stage. Abortion of these structures was less important in ‘Beldi’ than in ‘Baklouti’. Floral structure abortion induced by low night temperature was negatively and signifi cantly correlated with soluble acid invertase activity on ‘Beldi’ (r=-0.82), while on ‘Baklouti’, both sucrose synthase and insoluble acid invertase activities were correlated with fl oral abortion (r=-0.78). Under low night temperatures, sucrose synthase and soluble acid invertase activities were reduced to 50%, while the insoluble acid invertase activity was reduced by more than 90%. Enzymatic activities and fl owers abortion correlation show a differential response between these two parameters and the developmental stages of fl owers.

Key words: Abortion, bud and fl ower, hot pepper, low night temperature, sucrose synthase, acid invertase.

IntroductionFlowers retention and fruit set are highly sensitive to environmental factors in many species (Van Doorn and Stead, 1997). High temperature, drought, low light condition or failure of pollination are important factors that may induce abortion (Aloni et al., 1997; Heuvelink and Korner, 2001; Marcelis et al., 2004). These environmental stresses may alter photosynthetic activity (Havaux, 1993; Kitroongrung et al., 1992) and carbohydrate partitioning (Aloni et al., 1991a; Schaffer et al., 1987) causing an imbalance between source-sink structures (Geiger et al., 1996; Minchin and Thorpe, 1996).

Relationships between carbohydrate partitioning and floral structures are studied in few species. Photosynthetic competition between reproductive and vegetative sink seems to play a role in the abortion of Vecia faba fl owers (Aufhammer et al., 1987). Turner and Wien (1994) and Aloni et al. (1996) indicated that competition for assimilates between fl owers and adjacent young leaves may partially determine fl owers abortion in pepper. Then, abortion seems to be dependant not only on the source strength but also on sink strength (assimilate demand) of competing organs.

Several studies indicate that reproductive organs abortion depends on differentiation stages of these organs and the stress type. In fact, under shade conditions, Wien et al. (1989) noted that open fl owers were the most susceptible organs to abortion, while Aloni et al. (1991b), applying heat stress on sweet pepper, concluded that fl ower buds were more susceptible to abortion. Applying shade and heat stress at different stages of fl ower differentiation, Marcelis et al. (2004) noted that fl owers/fruits of sweet pepper were susceptible to abortion few days before anthesis.

Sucrose synthase and acid invertase were found to regulate phloem unloading (Geiger and Servaites, 1991). The magnitude of the activities of both enzymes was suggested as a reliable measurement of sink strength (Black, 1993; Jenner and Hawker, 1993) but was highly dependant on environmental stress (Roitsch, 1999; Sturm and Tang, 1999). Complex mechanisms have to be assumed which integrate the expression of the enzymes involved in carbohydrates production in source tissue and the regulation of source-sink relationship (Roitsch, 1999).

Investigation on low night temperature effect on abortion of vegetable fl oral structures is however scarce; the present experiment was conducted to determine the effect of low night temperature on sucrose synthase and acid invertase activities in relation to the fl oral structures abortion in hot pepper.

Material and methodsPlant material and growth conditions: Seeds of two hot pepper varieties (‘Beldi’ and ‘Baklouti’ from INRAT, Tunisia) were sown in elevated trays containing fertilized peat (NPK, 12-14-24) and allowed to germinate in a growth chamber at 25°C ± 2°C. Seedlings, at 6- 8 leaf stage, were transplanted into 12 liters pots, one plant per pot, containing the same substrate with a layer of about 2 cm of clay (Argex 4/10) and transferred to growth chamber with day/night temperatures of either low night temperature regime (25°C/10°C) or optimum temperature regime (25°C/20°C). The photoperiod was 16 h with light intensity of 250 ± 5 μmol m-2 s-1 (PAR). The relative humidity was maintained at about 70 ± 5%. Ten plants per variety were placed, at random, in each chamber and were watered when needed and fertigated with Nutri chem (N:P:K 22:5: 11) at 1 g L-1.

Journal

Appl

Floral structures sensitivity to low temperature (experiment 1): Evaluation of the fl oral structures sensitive to low night temperature was performed by the transfer of 3 plants of each variety from optimal to low night temperature regime (25/20°C to 25/10°C) and inversely. These plants were acclimated at least seven days in either condition prior to transfer to growing condition. Three developmental stages were carried out using a dimensional and morphological criteria:

- Stage A corresponding to the bud stage,

- Stage B corresponding to the fl oral bud with 3-4 mm diameter and 4-5 mm height (green petals welded to sepals, 4-5 days before anthesis),

- Stage C corresponding to fl oral buds with diameter ≥ 4 mm, height ≥ 5 mm (white petals lightly welded to sepals, 2-3 days before anthesis) (Fig. 1).

Under low night temperature, these three fl oral structures present, usually, a highly abortion percentage compared to the fl owers at anthesis stage; so the late developmental stage was not considered in this experiment.

Based on the developmental period of either stage (Tarchoun, 2003), the transfer period of plants was determined to seven nights. Before transferring the plants of ‘Beldi’ and ‘Baklouti’ varieties, 10 to 15 buds and 6-10 fl oral buds for each stage were studied to follow their evolution. The treatment was repeated three times.

Effect of temperature was evaluated using abortion percentage of buds (stage A) while abortion percentage of fl ower buds (stages B and C) was calculated based on fl ower buds development of stage B to stage C and the fl ower buds development of the stage C to pre-anthesis fl owers (stage D). The abortion percentage was calculated three and fi ve days after transfer.

Enzyme activity essays (experiment 2): Determination of sucrose synthase and acid invertase activities was carried out on fl oral structures at stage A, B and C and compared to the ovary at anthesis stage. Tissue samples of 200-300 mg fresh weight, collected in morning from each growth chamber, were ground in the mortar containing 3 mL of medium extraction: 25 mM HEPES buffer (N-2hydroxyethylpepirazine-N2-2ethansulphonic acid) pH 7.2; 2 mM DDT (DL-Dithiothreitol) ; 5 mM MgCl2 and 3 mM DIDCA (diethyldithiocarbamic acid) as antioxidant (Aloni et al., 1991b).

Sucrose synthase activity, in cleavage sense, was determined as follows: After extraction, the mixture was centrifuged at 10000 g for 30 min at 4°C. Aliquots of 200 μL of the supernatant were incubated for 30 min at 37°C in the medium containing 200 mM sucrose, 5 mM UDP, 50 mM HEPES buffer (pH 7). Reaction was stopped after boiling for 3 min. Fructose content was determined

by the addition of 3 mL of dinitrosalicylic acid reagent (1% dinitrosalicylic acid, 0.2% phenol, 0.05% bisulfate sodium and 1% NAOH) according to Schaffer et al. (1987). To reactivate the reaction, tubes were placed on boiling water bath for 5 min; before cooling, 1 mL of Rochelle salt (potassium sodium tartrate) was added to stabilize the reaction (Miller, 1959).

Soluble and insoluble acid invertase activities were determined by comparable extraction procedure as described for sucrose synthase and the supernatant was collected and placed at 0°C. The pellet was washed three times by the same buffer and centrifuged at 10000 g at 4°C, the supernatant was eliminated in each centrifugation. The later pellet was suspended in 3mL of the extraction medium added to NaCl (1M) and placed at 0°C for 1 hour.

Aliquots of 100μL of both the supernatant (soluble acid invertase) and the suspended pellet (insoluble cell wall acid invertase) were incubated in 1 mL 0.1N phosphate citrate buffer (pH 5) and 20 mM sucrose. The incubation was carried out for 1h at 37°C and was terminated by the addition of 1 mL dinitrosalicylic acid reagent (Aloni et al., 1991b). After boiling for 5 min, 1 mL of Rochelle salt was added (Miller, 1959). Sucrose synthase and acid invertase activities were determined colorimetrically at 580 nm of the optic density.

Statistical analysis: The experiment was carried out as split-split-plot design with temperature as main factor and varieties as the second factor (sub-plot) and the fl oral structures represent the smallest experimental unit (sub-sub-plot). Analysis of variance was performed by SAS system (1985), means were separated by LSD, and fl oral structures abortion was analyzed per date. The relationship between enzymatic activity and fl oral structures abortion was estimated by Pearson correlation coeffi cient using proc corr of SAS (1985).

Results Experiment 1Buds sensitivity to low night temperature: Fig. 2 illustrates the buds development to fl owers-buds on ‘Beldi’ and ‘Baklouti’ varieties after the transfer of plants from 25/20°C to 25/10°C and vice versa. Temperature effect was signifi cant from the third day of treatment and was stabilized at the fi fth day for both temperatures regimes and varieties. The transfer from optimal night temperature to low thermal-conditions caused an increase in buds abortion on the two varieties especially on ‘Baklouti’. Fifteen percent of buds aborted after three days and reached 82 and 88% at the fi fth day for ‘Beldi’ and ‘Baklouti’, respectively (Fig. 2.1). The reciprocal transfer favoured signifi cant decrease of buds abortion and the abortion rate depended on varieties (Fig. 2.2).

Flower buds sensitivity to low night temperature: The development evolution of fl ower buds is represented in Fig. 3. Transferring pepper plants from 25/20°C to 25/10°C and 3 days after treatment, fl ower buds expressed increased abortion of 56 and 69% for ‘Beldi’ and ‘Baklouti’ respectively. This abortion was more important at the fi fth day of treatment and reached 86 to 91% for ‘Beldi’ and ‘Baklouti’, respectively (Fig. 3.1). Although at the end of treatment period, fl ower buds at stage C aborted less than those at stage B; differential behaviour was noted on both local

Fig. 1. Floral growth stage based on the dimension and morphological criteria: bud (stage A) [1] , fl ower bud (stage B) [2] and fl ower bud (stage C) [3] corresponding Fig.1.

1 2 3

Bud

36 Relationships between pepper fl ower abortion and enzymes activity under low temperature

varieties. These fl ower buds at stage C seem to be more tolerant to low night temperature than those at stage B (Fig. 3.2).

The reciprocal transfer (from 25/10°C to 25/20°C) appeared to stabilize abortion at the fi fth day of treatment as well as for the two-stage of fl ower buds with a slight difference between both varieties (Fig. 3.3 and 3.4).

Floral structures sensitivity after three and fi ve days of the treatment: The lowest percentage abortion of fl oral structure was noted under constant optimal night temperature 25/20°C after either 3 or 5 days of treatment, while the highest abortion percentage was recorded following the transfer from 25/20°C to 25/10°C (Table 1). Low night temperature enhanced abortion of fl oral structure that varied from 47 to 74% after 3 and 5 days respectively, while the reciprocal transfer (from low to the optimal night temperature regime) reduced this abortion percentage to 34%.

The floral structure sensitivity was strongly dependent on varieties (Table 2). ‘Beldi’ showed lower abortion percentage than ‘Baklouti’ that seems to be more sensitive to low night temperature; it presented 48% of fl oral structure abortion after 5 days of treatment. Temperature x varieties interaction was not signifi cant. These results describe the response of both varieties grown during winter season under unheated greenhouse. The most pronounced effect of low night temperature was noted on

buds and fl ower buds at stage B compared to those at stage C, as well as after 3 or 5 days of the transfer. The highest abortion of fl oral structure was recorded on buds (stage A) that ranged from 43.5 to 50.2% after 3 and 5 days of the treatment, respectively, while the lowest percent abortion was noted on fl ower buds at stage C (Table 3). The differential sensitivity of fl oral structure to low night temperature could be explained by the enzymatic activity variation.

.

(65-75%)

(63-70%)

1

1

0

0

1

1

2

2

2

2

3

3

3

3

4

4

4

4

5

5

5

5

6

6

7

7

(1)

(2)

(86-91%)

(56-69%)*

Num

ber o

f flo

wer

bud

s at

sta

ge

Num

ber o

f flo

wer

bud

s at

sta

ge C

Beldi

Baklouti

1

1

0

0

1

1

2

2

2

2

3

3

3

3

4

4

4

4

5

55

5

6

6

7

7Days

(3)

(4)

(63-72%)

(48%)

(36-39%)*

(53%)

Num

ber o

f flo

wer

bud

s at

sta

ge

Num

ber o

f flo

wer

bud

s at

sta

ge C

(50%) *

(66-70%)

(82-88%)

(90%)

(60%)

(45%)

Beldi

Baklouti

1

1

11

2

2

2

2

3

3

3

4

4

4

4

5

5

5

6

6

6

6

7

7

7

Flow

er b

ud d

evel

opm

ent

Flow

er b

ud d

evel

opm

ent (24%)

(36%)

(49%)

8

10

Days

(1)

(2)

Fig. 3. Development evolution of fl ower buds, on ‘Beldi’ and ‘Baklouti’ varieties, at stage C and at the pre-anthesis stage (D) following plant transfer from 25/20°C to 25/10°C (1and 2) and vice versa (3 and 4). *numbers between brackets represent the fl ower bud abortion percent after 3 and 5 days of treatment.

Fig. 2. Development evolution of buds to fl ower buds on ‘Beldi’ and ‘Baklouti’ varieties following plant transfer from 25/20°C to 25/10°C (1) and vice versa (2). * Numbers between brackets represent the bud abortion percent after 3, 4 and 5 days of treatment.

Table 1. Abortion percentage of fl oral structure after 3 and 5 days of the treatment at different temperature regime, constant 25/20°C, transfer from 25/20°C to 25/10°C and inverselyTreatment 3 days 5 daysconstant 25/20°C 13.9b* 25.9cTransfer from 25/20°C to 25/10°C 47.2a 74.5aTransfer from 25/10°C to 25/20°C 42.2a 33.9bLSD 6.9 4.7* means not followed by the same letter are signifi cantly different at P≤ 0.05

BB

Relationships between pepper fl ower abortion and enzymes activity under low temperature 37

Table 2. The average of fl oral structure abortion on ‘Beldi’ and ‘Baklouti’ after 3 and 5 days of the transfer from optimal night temperature regime (25/20°C) to low night temperature regime (25/10°C) Variety 3 days 5 days‘Beldi’ 31.5b* 41.2b‘Baklouti’ 37.4a 48.3aLSD 5.7 5.4* means not followed by the same letter are signifi cantly different at P≤ 0.05 Table 3. Differential sensitivity of fl oral structure abortion to low night temperature after 3 and 5 days of the transfer from optimal night temperature regime (25/20°C) to low night temperature regime (25/10°C) Stage/organ 3 days 5 daysBuds (stage A) 43.5a* 50.2aFlower buds (Stage B) 37.7b 50.1aFlower buds (Stage C) 2.1c 33.8bLSD 3.8 3.1* means not followed by the same letter are signifi cantly different at P≤ 0.05 Experiment 2Effect of low night temperature on sucrose synthase and acid invertase activities: The activity of the sucrose synthase (the sucrose cleaving enzyme) and soluble and insoluble acid invertase (expressed on a fresh weight basis) was strongly dependent on temperature regime (Table 4). On the other hand, 50 to 90% of the acid invertase activity was found in the soluble fraction either at optimal or at low night temperature regime. The lowest activity was noted on the insoluble part of acid invertase under temperature regime of 25/10°C.

Effect of varieties on enzymatic activity: Enzymatic activity depends on varieties. In fact, enzymatic activity of ‘Beldi’ was found to be superior to that on ‘Baklouti’ (Table 5). Sucrose synthase activity was greatly suppressed in ‘Baklouti’ (3.4 μmol gfwt-1 min-1) compared to ‘Beldi’ (6.6 μmol gfwt-1 min-1 ). Insoluble and soluble acid invertase followed a similar pattern, but less pronounced for the soluble fraction.

Enzymatic activity in different fl oral structures: Table 6 shows that enzymatic activity depends on the fl oral structures. In fact, in fl ower ovaries (at the anthesis stage), this activity was more intense than in buds (stage A) and fl ower buds (stage B) and less pronounced in fl ower buds at stage C. The activity of the soluble

acid invertase appears to be more important than other enzymes for all fl oral structures and it is characterized by an increasing gradient according to the evolution of these structures. The insoluble acid invertase activity shows a different behaviour; it indicates a decrease between buds stage and fl ower buds stage and takes an increasing trend until the ovary.

Inspite of the similar activity for the sucrose synthase, buds and fl ower buds express an opposite activity for the soluble and insoluble acid invertase. On the other hand, sucrose synthase and soluble acid invertase activities were more intense in fl ower buds at stage C as compared to the previous structures. Abortion of these fl oral structures seems to be controlled differentially by one or the other type of enzymes.

Low night temperature reduced sucrose synthase activity in ‘Baklouti’ by 53% and by 38% in ‘Beldi’ (Table 7). Under low night temperature of 10°C, reduction of soluble and insoluble acid invertase activities was more pronounced than sucrose synthase activity. The insoluble fraction of acid invertase was more affected in both ‘Beldi’ and ‘Baklouti’ varieties with 1 to 0.7 μmol gfwt-1 min-1, respectively.Table 6. The enzymatic activity expressed on μmol gfwt-1 min-1 on four different pepper fl ower structures, buds (stage A), fl ower buds (stage B and C) and fl ower ovaries at anthesisStage/organ ctures Sucrose

synthaseSoluble acid

invertase Insoluble acid

invertase Buds (stage A) 2.9c* 10.7c 4.0bFlower buds (stage B) 2.0c 15.9b 2.3cFlower buds (stage C) 4.6b 19.4a 4.8bOvaries 10.5a 20.2a 14.2aLSD 1.0 2.7 1.1* means not followed by the same letter are signifi cantly different at P≤ 0.05

Table 7. Sucrose synthase and acid invertase activities expressed on μmol gfwt-1 min-1 on ‘Beldi’ and ‘Baklouti’ varieties grown under optimal night temperature regime (25/20°C) or low night temperature (25/10°C)Enzyme ‘Beldi’ ‘Baklouti’

25/20°C 25/10°C 25/20°C 25/10°CSucrose synthase 8.2±2.1* 5.1±0.2 4.7±1.0 2.2±0.3Soluble Acid invertase 26.1±7.1 11.8±4.9 17.9±5.2 10.4±3.2Insoluble Acid invertase 15.7±6.0 1.0±0.2 7.9±1.5 0.7±0.08* means ± SE (n= 12 replications)

Relationships between floral structures abortion and enzymatic activity: Correlation coefficient between floral structures abortion in ‘Beldi’ and ‘Baklouti’ grown under optimal night temperature regime (25/20°C) or low night temperature (25/10°C) and enzymatic activity revealed that, under low night temperature, the abortion of ‘Baklouti’ fl oral structures was associated negatively with sucrose synthase and insoluble acid invertase (r=-0.78**), while for ‘Beldi’, this coeffi cient was only signifi cant for soluble acid invertase (r=-0.82**). However, under optimal temperature regime (25/20°C) fl oral structures abortion of ‘Beldi’ and ‘Baklouti’ was associated to the insoluble and soluble acid invertase, respectively (Table 8). Moreover, the abortion of different fl oral structures seems to be dependant on the enzyme type. In fact, buds abortion was associated mainly to the acid invertase activity as well as in ‘Beldi’ and ‘Baklouti’, while sucrose synthase activity controlled the fl owers buds (stage B) abortion. Although abortion of fl ower buds (stage C) depended on the varieties; soluble and insoluble acid invertase controlled

Table 4. The effect of low night temperature on sucrose synthase and acid invertase activities expressed on μmol gfwt-1min-1 evaluated on different pepper fl ower structuresTemperature Sucrose



invertase 25/20°C 6.5a* 22.0a 11.8a25/10°C 3.6b 11.1b 0.8bLSD 1.7 2.5 0.9* means not followed by the same letter are signifi cantly different at P≤ 0.05 Table 5. The average enzymatic activity expressed on μmol gfwt-1 min-1

evaluated on pepper fl ower structure of two local hot pepper varieties ‘Beldi’ and ‘Baklouti’Variety ties Sucrose



invertase ‘Beldi’ 6.6a* 18.9a 8.4a‘Baklouti’ 3.4b 14.2b 4.3bLSD 0.9 3.3 2.1* means not followed by the same letter are signifi cantly different at P≤ 0.05


this abortion in ‘Baklouti’ while only insoluble fraction of acid invertase was associated with fl ower buds abortion in ‘Beldi’.

DiscussionThe physiological processes like photosynthesis, metabolism of sugars and translocation of assimilate have been studied under stress conditions. Important progress has been made in quantifying and modeling synthesis and distribution of assimilates in leaves of fruits tree species (Grossman and Dejong, 1994; Wermelinger et al., 1991) and of the vegetable species (Heuvelink, 1995; Marcelis, 1996; Marcelis et al., 1998). However, the fl oral structures: buds and fl oral buds, more particularly of the vegetable species cultivated under low night temperature, had little attention.

The fl oral structures abortion at various differentiation stages has been more common under low light intensity and/or high temperature in several Solanaceous crops, including pepper (Aloni et al., 1996; 1997) and tomato (Heuvelink, 1996). The sensitivity of the fl oral structure to low night temperature was exhibited when plants were transfered from optimal condition (25/20°C) to low night temperature (25/10°C) and reciprocally (experiment 1). A greater abortion rate had been noted after three days of this treatment. This result suggests an early effect of temperature on these structures. A similar response has been recorded by Aloni et al. (1991b) who observed that the development of buds is affected by the high temperatures (>30°C) during the fi rst 6 hours post treatment and a complete growth stopping took place after 24h of heat stress.

The abortion of buds and fl ower buds at stages B and C has been found to be dependent on temperature regime and variety, and within the same variety, on the fl oral differentiation stages (Fig. 2, 3). The low night temperatures (25/10°C) enhances buds and fl ower buds abortion which is more pronounced for buds than fl ower buds at stage B and less for those at stage C. The abortion of fl oral structures was more pronounced in ‘Baklouti’ than in ‘Beldi’. This differential behaviour of varieties tested has been found in close association with enzymatic activity that was strongly reduced by the night temperature of 10°C (Table 4). Indeed, the activities of both enzymes, sucrose synthase and the soluble and insoluble acid invertase, were signifi cantly higher under temperature regime of 25/20°C and were more expressed in ‘Beldi’ than in ‘Baklouti’. This result suggests that, in addition of the temperature effect, other factors such as, the genetic aspect seems to regulate the metabolic activity. In fact, under light stress conditions (shading of 60%), Shiffriss et al. (1994), studying the abortion of fl owers of inbred lines and hybrids F1 and F2 of pepper in segregation, had noted variable abortion rates between these different types of plant material and concluded that some genetic factors are responsible.

Geiger et al. (1996) showed that distribution of assimilates is controlled by at least two enzymes: sucrose synthase and acid invertase and this distribution is controlled by the strength of sink organs. It seems, however, that this distribution is governed by the intensity of the organ strength. Buds (stage A) and fl ower buds at stage B presented the weakest enzymatic activity in comparison to ovaries at anthesis stage, while fl ower buds at stage C had intermediate position (Table 6). The differential abortion of these structures would be, particularly, attributed to the activity of these two enzymes that may serve as an indicator for the organs strength (Sun et al., 1992).

Heuvelink (1995) and Marcelis et al. (1998) reported that the temperature is an important factor affecting the distribution of assimilates in plants, while light and the CO2 level affect the strength of the source organs. On the other hand, Bertin (1995) suggested that the abortion of the tomato infl orescences, before the anthesis may be due to the competition for assimilate between the young vegetative organs and the last infl orescences. For other researchers (Sato et al., 2001) the fl oral buds abortion (before the anthesis) at the tomato seems to result in a competition of assimilates between fruits and fl owers of the same bunch or the superior bunch. In this investigation, result suggests that the fl oral structure abortion is attributed to a poor translocation capacity of assimilates. Furthermore, a direct effect of the temperature regime is suggested. Varying sink strength by changing the number or position of early-formed fruits affected abortion in sweet pepper fl owers and the abortion rate showed a linear relationship with the growth rate of the earlier formed competing fruits (Marcelis et al., 2004). This abortion induction may be caused by competition for available assimilates by dominance due to the production of plant growth regulators from the developing fruits, or by a combination of them (Heuvelink and Korner, 2001; Marcelis and Baan Hofman-Eijer, 1997)

The competition for the assimilate translocation between the young leaves and fl owers also constitutes a reason of the fl ower abortion (Van Doom and Stead, 1997) since these young leaves constitute a more powerful sinks than the adjacent fl owers (Aloni et al., 1991a, 1991b). However, Turner and Wien (1994) did not report such effect, since leaves suppression did not improve flowers retention. This controversy assumes that the floral structure abortion would rather be assigned to a competition between the fl oral structures at different differentiation stages.

Flowers, at anthesis stage, have been considered as a strong sinks (Black 1993; Marcelis 1996). In the present study (Table 6), except for insoluble acid invertase, the most intense enzymatic activity has been found in the ovaries; this could explain their low abortion under low night temperature. Working under high temperature, Aloni et al. (1997) found more intense activity of sucrose synthase at post-anthesis stage.

Table 8. Pearson correlation coeffi cients between fl oral structures abortion in ‘Beldi’ and ‘Baklouti’ grown under optimal night temperature regime (25/20°C) or low night temperature (25/10°C) and enzymatic activity expressed as μmol g fwt-1 min-1 Tem ‘Beldi’ ‘Baklouti’

Sucrose synthase Soluble acid invertase

Insoluble acid invertase

Sucrose synthase Soluble acid invertase

Insoluble acid invertase

25/20°C -0.13 ns -0.44 ns -0.70* -0.13 ns -0.72* -0.49 ns

25/10°C -0.12 ns -0.82** -0.17 ns -0.78** -0.35 ns -0.78**

* , ** signifi cant differences at P<0.05 and P<0.01 respectively, ns- differences not signifi cant at P>0.05

Relationships between pepper fl ower abortion and enzymes activity under low temperature 39

The analysis of the association between the fl oral structures abortion and the enzymatic activity revealed that, at low night temperature, abortion control suggests a genetic effect; thus, the soluble acid invertase seems to control abortion in ‘Beldi’ variety, whereas on ‘Baklouti’ variety the combination of sucrose synthase and insoluble acid invertase controls this phenomenon (Table 8).

The abortion of buds seems to be associated with acid invertase activity especially its soluble fraction and to a lesser extent with insoluble fraction either in ‘Beldi’ or ‘Baklouti’. Sucrose synthase seems to be responsible of the fl owers buds (stage B) in both varieties. Compared to the fl owers at anthesis stage, studies on fl oral structure abortion at the fi rst stages of differentiation (buds and fl ower buds) in relation to the metabolic activity are scarce.

The amplitude of variation of fl ower structures to abortion is attributed to the simultaneous effects of the endogenous factors (metabolic activity) and of the exogenous factors (environmental factors). The later factor optimize the availability and the distribution of assimilates between the different organs of the plant. The abortion of buds and fl owers buds under low night temperatures of 10°C, associated to a low enzymatic activity, support this hypothesis.

ReferencesAloni, B., T. Pashkar and L. Karni, 1991a. Nitrogen supply infl uences

carbohydrates partitioning in pepper seedlings and transplant development. J. Amer. Soc. Hort. Sci., 116: 995-999.

Aloni, B., T. Pashkar and L. Karni, 1991b. Partitionning of [14]-C sucrose and acid invertase activity in reproductive organs of pepper plants in relation to their abortion under heat stress. Ann. Bot., 67: 371-377.

Aloni, B., L. Karni, Z. Zaidman and A.A. Schaffer, 1996. Changes of carbohydrates in pepper (Capsicum annuum L.) fl owers in relation to their abortion under different shading regimes. Ann. Bot., 78: 163-168.

Aloni, B., L. Karni, Z. Zaidman and A.A. Schaffer, 1997. The relationship between sucrose supply, sucrose cleaving enzymes and fl ower abortion in pepper. Ann. Bot., 79: 601-605.

Aufhammer, W., I. Gotz and M. Peter, 1987. Yield performance in fi eld bean (Vicia faba L.) in relation to interactions between infl orescence at different nodes. J. Agri. Sci., 105: 662-664.

Bertin, N. 1995. Competition for assimilates and fruit position affect fruit set in indeterminate greenhouse tomato. Ann. Bot., 75: 55-65.

Black, C.C., 1993. Sink strength: it is real or measurable ? Plant Cell Envir., 16: 1037-1038.

Deng, X., S.A. Weinbaum, T.M. DeJong and T.T. Muraoka, 1991. Pistillate fl ower abortion in ‘Serr’ walnut associated with reduced carbohydrate and nitrogen concentrations in wood and xylem sap. J. Amer. Soc. Hort. Sci., 116: 291-296.

Geiger, D.R., K.E. Koch and W.J. Shieh, 1996. Effect of environmental factors on whole plant assimilate partioning and associated gene expression. J. Exp. Bot., 47: 1229-1238.

Geiger, D.R. and J.C. Servaites, 1991. Carbon allocation and response to stress. In: Response of Plants to Multiple Stress. Mooney H.A. et al. (Eds) Academic press, San Diego. p. 103-125.

Grossman, Y.L. and T.M. Dejong, 1994. Peach: a simulation model of reproductive and vegetative growth in peach trees. Tree Physiol., 14: 329-345.

Havaux, M. 1993. Rapid photosynthetic adaptation to heat stress triggered in potato leaves by moderately elevated temperatures. Plant Cell Environ., 16: 461-467.

Heuvelink, E. 1995. Effect of temperature on biomass allocation in tomato (Lycopersicon esculentum). Physiol. Plantarum, 94: 447-452.

Heuvelink, E.1996. Dry matter portioning in tomato: validation of a dynamic simulation model. Ann. Bot., 77: 71-80.

Heuvelink, E. and O. Korner, 2001. Parthenocarpic fruit growth reduces yield fl uctuation and blosson-end rot in sweet pepper. Ann. Bot., 88: 69-74.

Jenner, C.F. and J.S. Hawker, 1993. Sink strength: soluble starch synthase as a measure of a sink strength in wheat endosperm. Plant Cell Environ., 16: 1023-1024.

Kitroongrung, N., S. Jodo, J. Hisai and M. Kato, 1992. Photosynthesis characteristics of melon grown under high temperature. J. Jap. Soc. Hort.. Sci., 61: 107-114.

Marcelis, L.F.M. 1996. Sink Strength as a determinant of gray matter partitioning in the whole plant. J. Exp. Bot., 47: 1281-1291.

Marcelis, L.F.M. and L.R. Baan Hofman-Eijer, 1997. Effect of seed number on competition and dominance among fruits in Capsicum annuum L. Ann. Bot., 79: 687-693.

Marcelis, L.F.M. , E. Heuvelink and J. Goudriaan, 1998. Modeling biomass reduction and yield of horticultural crops: a review. Scientia Hortic., 74: 83-111.

Marcelis, L.F.M., E. Heuvelink, L.R. Baan Hofman-Eijer, J. Den Bakker and L.B. Xue, 2004. Flower and fruit abortion in sweet pepper in relation to source and sink strength. J. Exp. Bot., 55: 2261-2268.

Miller, G.L. 1959. Use of dinitrosalycilic acid reagent for determination of reducing sugars. Ann. Chem., 3: 426-428

Minchin, P.E.H. and M.R. Thorpe, 1996. What determines carbon partitioning between competing sites? J. Exp. Bot., 47: 1293-1296.

Roitsch, T. 1999. Source-sink regulation by sugar and stress. Current Opinion Plant Biol., 2: 198-206.

Statistical Analysis System, 1985. SAS User’s guide Statistics. Ed. Cary, NC.

Sato, S., M.M. Peet and R.G. Gardner, 2001. Formation of parthenocarpic fruit, undeveloped fl owers and aborted fl owers in tomato under moderately elevated temperatures. Scientia Hortic., 90: 243-254.

Schaffer, A.A., B. Aloni and E. Fogelman, 1987. Sucrose metabolism and accumulation in developing fruit of sweet and non-sweet genotypes of Cucumis. Phytochemistry, 26: 1883-1887.

Shiffriss, C., M. Pilowsky and B. Aloni, 1994. Variation in fl ower abortion of pepper under stress shading conditions. Euphytica, 78: 133-136.

Sturm, A. and G.Q. Tang, 1999. The sucrose-cleaving enzymes of plants are crucial for development, growth and carbon partitioning. Trends Plant Sci., 4: 401–407.

Sun, J., T. Loboda, S.S. Sung and C.C. Jr Black, 1992. Sucrose synthase in wild tomato Lycopersicon chmielewskii, and tomato fruit strength. Plant Physiol., 98: 1163-1169.

Tarchoun, N. 2003. Etude de la fl oraison et de la nouaison en conditions de basses températures nocturnes et leur déterminisme génétique chez le piment (Capsicum annuum L.). Thèse de Doctorat d’Etat en Sciences Agronomiques de l’INAT- Tunisie, 188pp.

Turner, A.D. and H.C. Wien, 1994. Photosynthesis, dark respiration and bud sugar concentrations in pepper varieties differing in susceptibility to stress-induced bud abortion. Ann. Bot., 73: 623-628.

Van Doom, W.G. and A.D. Stead, 1997. Abortion of fl owers and fl oral parts. J. Exp. Bot., 48: 821-837.

Wermelinger, B., J. Baumgartner and A.P. Gutterrez, 1991. A demographic model of assimilation and allocation of carbon and nitrogen in grapevines. Ecol. Modeling, 53: 1-26.

Wien, H.C., A.D. Turner and S.F. Yang, 1989. Hormonal basis for low light intensity induced fl ower bud abortion of peppers. J. Amer. Soc. Hort. Sci., 114: 981-985.


Journal of Applied Horticulture, 11(1):41-45, January-June, 2009

Effects of antibrowning agents on the shelf life of fresh-cut green jackfruit (Artocarpus heterophyllus Lam.)

Boodia Navindra*, Ruggoo Arvind and Boodoo B. Hassina

Faculty of Agriculture, University of Mauritius, Réduit, Mauritius. *E-mail: [email protected]

AbstractGreen mature jackfruits were minimally processed into cubes, dipped in solution of citric acid (0 and 1%) and ascorbic acid (0, 1 and 2%), vacuum packed at 550 mbar atmospheric pressure in 80 μm laminated low density polyethylene vacuum pouches and stored at 2-4°C for 15 days. A control was prepared, using water. Quality parameters like colour, fi rmness, pH, titratable acidity and total soluble solids were determined during storage. Colour parameters indicated increase in browning during storage. A signifi cant increase (P<0.05) in titratable acidity and signifi cant decrease (P<0.05) in pH were observed in all treatments. Texture signifi cantly decreased (P<0.05) in all treatments during storage. Combinations of the browning inhibitors were more effective than when applied individually. Citric acid and ascorbic acid when applied together resulted in non-signifi cant change (P>0.05) in microbial counts, browning, and colour lightness. Treatment of 1% citric acid and 2% ascorbic acid in combination with moderate vacuum packaging and low temperature storage was found most effective in inhibiting browning and deterioration of fresh-cut green jackfruit for up to 15 days.

Key words: Antibrowning agents, citric acid, ascorbic acid, Artocarpus heterophyllus, minimal processing, green jackfruit, moderate vacuum packaging.

IntroductionFresh-cut products are defi ned as fruits or vegetables that have been freshly cut, washed and packaged that offer consumers a nutritious, convenient and fresh-like product (Gimenez et al., 2003). There is signifi cant growth in the market for these products worldwide. However, although minimally processed products satisfy consumer demand towards fresh, healthy and convenient products, minimal processing such as peeling, cutting, shredding and grating renders the product highly perishable (Laurila and Ahvenainen, 2002). Fresh-cut processing operations result in a wound response that leads to chemical reactions such as increased respiration rates, cut surface browning, increased ethylene production, water loss, and off-fl avour development, which in turn results in texture loss and reduces product quality and shelf-life (Escalona et al., 2005; Gonzalez-Aguilar et al., 2000; Agar et al., 1999). Apart from biochemical causes, spoilage of fresh-cut products is also caused by the growth of micro-organisms, the growth of which is enhanced in the processing process due to the availability of cell nutrients and increased surface area for growth.

Fresh-cut fruits and vegetables are normally packaged in fi lm bags to reduce the respiration rate and slow the growth of anaerobic spoilage micro-organisms on the product. Creation of a modifi ed atmosphere of low oxygen and elevated carbon dioxide levels inside the package may retard browning, spoilage and maintain the fresh appearance (Gonzalez-Aguilar et al., 2000; Farver and Dodds, 1995). Modifi ed atmosphere packaging is used as a supplement to low temperature storage to reduce the rate of enzyme activity. Moderate vacuum packaging of shredded lettuce with polyethylene (80μm) resulted in a shelf life greater than 10 days at 5°C (Heimdall et al., 1995).

Green jackfruit, as a cooked vegetable, is an appreciated dish by the

Asian community. Therefore, presentation of jackfruit as a fresh-cut product would be very convenient to reduce preparation time in the kitchen and would be a product welcomed by consumers. However, enzymatic browning was found to be the major limiting factor in fresh-cut jackfruit commercialisation. Enzymatic browning occurs when phenolic compounds are oxidised by the copper-containing enzyme, polyphenol oxidase (PPO) (Paul and Palmer, 1972). During fresh-cut preparation, cells are ruptured causing the intermixing of PPO with phenolic substrates, which are normally compartmentalised, resulting in an undesirable brown colour (Dong et al., 2000; Sapers et al., 2002).

Browning and decay in fresh-cut fruits and vegetables can be reduced by the use of natural antibrowning agents (Gonzalez-Aguilar et al., 2000). As an alternative to sulphites, which has been found to cause allergic reactions in asthmatics (Sapers, 1993), the most studied alternative has been ascorbic acid (Laurila and Ahvenainen, 2002). Since ascorbic acid eventually oxidises to dehydro-ascorbic acid, it is best used in combination with other antibrowning agents like citric acid. Ascorbic acid and citric acid have been found to be effective enzymatic browning inhibitors. Langdon (1987) found that immersion of peeled and sliced potatoes in solutions of ascorbic and citric acid followed by vacuum packaging, resulted in product shelf life of greater than 14 days.

The objective of this study was to investigate the effects of ascorbic acid and citric acid as antibrowning agents on the quality and microbial load of moderately vacuum packed fresh-cut green jackfruit stored at 2-4°C.

Materials and methodsPlant materials: Disease free jackfruits of the same maturity were hand-harvested from registered growers in Mauritius. Harvested

Journal

Appl

jackfruits were immediately transported from the fi eld to the University of Mauritius.

Sample preparation: Jackfruits were manually cut at the base and the latex allowed to fl ow. The fruits were then cut into rings, peeled, cored and fi nally cut into cubes of 25 mm. The jackfruit cubes were dipped in the test solutions, which included different concentrations of ascorbic acid and citric acid (Table 1). These cubes were drained and blotted dry with paper towels, placed in laminated polyethylene vacuum bags of thickness 80 μm (oxygen transmission rate: 35 cm3/m2/24h at 23°C; Linpac Ltd., Pontivy, France) and vacuum packaged at 550 mbar atmospheric pressure using the Multivac vacuum packaging machine (Multivac, Wolfertschwenden, Germany). Each pack contained about 100-125 g cut fruit. The packaged fresh-cut fruits were stored at 2-4°C for 15 days. All utensils, containers and work surfaces were sanitized with 3% Oxonia® solution. Table 1. Composition of antibrowning dip solutions used on jackfruit slices Treatment ConcentrationControl (T1) WaterCitric acid (T2) 1 %Ascorbic acid (T3) 1 %Ascorbic acid + Citric acid (T4) 1 % + 1 %Ascorbic acid (T5) 2 %Ascorbic acid + Citric acid (T6) 2 % + 1 %

Physico-chemical determinations: Destructive analysis was carried out in duplicate on days 0, 3, 7, 10, 13 and 15. Two packages selected at random were subjected to physical and chemical analyses. Diced jackfruits selected randomly from each package were blended and then squeezed manually through cheesecloth. The homogenate was used for titratable acidity (TA) determination and the juice analyzed for total soluble solids content (TSS) and pH. The TSS was measured using a standard hand refractometer, corrected at 20°C, and expressed as °Brix (Askar and Treptow, 1993). The pH was determined using a pH meter as per AOAC method (1995). For TA determination, 10g of pulp homogenate was boiled with 50 ml hot water for 5 minutes. The mixture was fi ltered and the fi ltrate titrated with 0.1 N NaOH (Askar and Treptow, 1993) in triplicate. The mean titre value obtained was used to calculate TA, which was expressed in g of citric acid/ 100 g of sample.

Colour measurement: Browning of the jackfruit cubes was measured as L*, a*, b* using a Minolta CR-300 colorimeter (Minolta Company Ltd, Osaka, Japan). Measurements were taken at six different points on each of the cut surface of fi ve dices of the green jackfruit. The L* values for colour indicate lightness, whereby an increase in lightness is indicated by an increase in the L* values. Positive a* values indicate redness and negative a* values indicate greenness. Positive b* values indicate yellowness and negative values blueness.

Determination of fi rmness: Firmness of fi ve jackfruit dices chosen randomly from each package was determined using a hand-held penetrometer at six different places on each dice (HP-Fff; Tracer 0.25 cm2; Bareiss Prufgeratebau GmbH, Oberdischingen, Germany) as per Askar and Treptow (1993).

Microbiological analysis: Microbiological analysis was carried out in triplicate for each treatment on days 0, 6 and 12. A sample

of 10g of jackfruit was taken from each replicate of the six treatments. This was added to 270 ml sterile (0.1 %) peptone water and blended in a stomacher for at least 2 minutes. Serial dilutions (10-1-10-4) were carried out using 1 ml of the macerated sample and 9 ml aliquots of peptone water. The pour plate technique was used to inoculate the medium. Plate count agar and potato dextrose agar were used for enumeration of total viable count and yeast and mould count, respectively. The inoculated plates were incubated at 25°C for 3-5 days. The counts were expressed as log colony-forming units per g (Log10 CFU/g) of sample.

Statistical design and analysis: The statistical design used was a factorial Randomized Block Design with two levels of citric acid (0 and 1%) and three levels of ascorbic acid (0, 1 and 2%). Days in storage (0, 3, 7, 10, 13 or 15) were used as the blocks. Two replicates per treatment combination were subjected to destructive analysis. Data was analyzed using MINITAB Version 13.1. All Least Signifi cant Differences (LSD) were computed at the 5% level of signifi cance.

Results and discussionTitratable acidity and pH: Increase in TA (as g citric acid 100 g-1 fresh tissue) was associated with a decrease in pH during storage (Fig. 1). High acidity was due to the presence of both the acids. Ascorbic acid as well as citric acid was found to have a signifi cant effect on TA (P<0.05). A signifi cant interaction effect (P<0.05) was also noted between citric and ascorbic acids, whereby TA was found to be higher in fresh-cut jackfruit treated with both acids (Table 2). This could be due to respiration.

Utilization of organic acids during respiration was found to cause a decrease in TA in fresh-cut apples (Fan et al., 2005). Increase in acidity during storage could be due to ripening of the fruit, where the degradation of pectin in the cell wall results in the formation of galacturonic acids (Eskin, 1990). Another factor contributing to increased acidity could be due to the fi xation of carbon dioxide formed during respiration into organic acids (Wang, 1990). Table 2. Two-way table of treatment means for TA

Citric acid (%) Ascorbic acid concentration (%)

0 1 20 0.120 0.097 0.1351 0.124 0.156 0.160

SED (interaction) = 0.007Both citric acid and ascorbic acid led to a signifi cant decrease (P<0.05) in the pH during storage (Table 3). There was no evidence (P>0.05) of any interaction between the two treatments indicating that they were acting independently on the pH during storage.

Fig 1. Changes in mean TA and pH during storage of the diced jackfruit LSD (pH) = 0.2303; LSD (TA) = 0.0138

Days

44.14.24.34.44.54.64.74.84.9

0 3 7 10 13 1500.020.040.060.080.10.120.140.160.18

mean pHMea

n pH

Mean TA

mean TA

42 Preservation of fresh-cut green jackfruit

Table 3. Overall pH of diced jackfruits for citric acid and ascorbic acid treatmentsCitric acid (CA) pH Ascorbic acid

(AA)pH

0% 4.70 0% 4.651% 4.32 1% 4.55

2% 4.48

SED (CA) = 0.065, SED (AA) = 0.079Total soluble solids: On day 0, a high °Brix was observed, which decreased on day 3. This could be due to increased respiration as a result of wounding during processing. Sugars and organic acids are the main substrates for respiration in plants (Tovar et al., 2000). Citric acid was found to have a signifi cant effect (P<0.05) on the TSS (Table 4). No signifi cant interaction effect (P>0.05) between ascorbic and citric acid was noted. Changes in the soluble solids may be attributed to changes occurring during ripening (Nakasone and Paul, 1999). However in the present study, it was found that the TSS remained almost constant throughout storage. Table 4. Overall TSS of diced jackfruits for citric acid and ascorbic acid treatments

Citric acid (CA) °Brix Ascorbic acid (AA)

°Brix

0% 4.09 0% 4.221% 4.33 1% 4.14

2% 4.27SED (CA) = 0.064, SED (AA) = 0.078Colour: Changes in a* and L* values have been used to monitor enzymatic browning of cut apples (Soliva-Fortuny et al., 2001). Kim et al. (1995) correlated a* and L* values with PPO activity but there was no correlation with b* values. A significant interaction effect (P<0.05) was noted for a* value suggesting better browning control when using both acids (Table 5).

Days-6-4-202468

10

0 3 6 9 12 15

T1 T2 T3 T4 T5 T6

a* v

alue

s

Fig 2. Changes in a* values during storage of the diced jackfruit. LSD (any 2 means) = 2.072

Fig 3. Changes in L* values during storage of the diced jackfruit. LSD (any 2 means) = 5.20

Table 5. Two-way table of treatment means for a* valuesCitric acid

(%)Ascorbic acid (%) Mean

0 1 20 (+) 5.49 (+) 0.70 (-) 0.22 (+) 1.991 (+) 0.97 (-) 1.33 (-) 2.38 (-) 0.91

Mean (+) 3.23 (-) 0.32 (-) 1.30SED (interacion) = 0.581; SED (AA) = 0.411; SED (CA) = 0.335. F-ratios: CA=74.91; AA=67.28; CA*AA=5.85However, the F-ratio for the interaction effect was much smaller compared to the main effects’ sums of squares. This indicates that changes in a* were mostly driven by the main effects. The increase in a* values during storage indicates an increased occurrence of browning (Fig. 2).

L* values were found to decrease during storage (Fig. 3). The control was found to have the lowest L* value. This is due to the fact that no antibrowning agent was used in the control, thereby causing greater extent of browning (Fig. 3). There was no signifi cant interaction between the effects of citric and ascorbic acid (P>0.05). Both acids were found to have signifi cant effect (P<0.05) on the lightness of the fresh-cut green jackfruit (Table 6). Application of citric acid and ascorbic acid individually were found to inhibit browning moderately. The combination of 1% CA and 2% AA proved to be the most effective treatment in reducing browning for 15 days. Combination of 1% CA and 1% AA was also effective in suppressing browning but to a lesser extent as slight browning was noted on day 15. Table 6. Mean L* values of diced jackfruit for citric acid and ascorbic acid treatments

Citric acid (CA) L* value Ascorbic acid (AA)

L* value

0% 75.2 0% 71.31% 79.9 1% 79.6

2% 81.8

SED (CA) = 0.84, SED (AA) = 1.03Firmness: Firmness is considered as an important quality criterion for fresh-cut products. Both citric acid and ascorbic acid had signifi cant effects (P<0.05) on the fi rmness of the fresh-cut jackfruit where both acids were found to retain the fi rmness (Table 7). Retention of fi rmness could be due to cross-linking between endogenous calcium ions and the freely available carboxyl groups in the cell wall pectin, following acidifi cation (Sapers and Miller, 1995). Percentage loss in fi rmness for all the treatments is depicted in Fig. 4.

Days

0

10

20

30

40

50

60

3 5 7 9 11 13 15

%lo

ssin

firm

nes

s

T1 T2 T3 T4 T5 T6

Fig 4. Percent loss in fi rmness during storage of the diced jackfruit. LSD (any 2 means) = 9.708

505560657075808590

0 3 7 10 13 15Days

T1 T2 T3 T4 T5 T6

L* v

alue

s Preservation of fresh-cut green jackfruit 43

Onset of ripening could be a factor responsible for the fi rmness loss. Fruit pulp loses fi rmness during ripening, which occurs as a result of the decomposition of the primary cell wall constituents, following the solubilization of pectin by several hydrolytic enzymes. (Eskin, 1990; Riquelme et al., 1999). In addition to ripening, loss of fi rmness could also be due to decreased turgor due to water loss (Beaulieu and Gorny, 2002). Table 7. Mean fi rmness of diced jackfruits for main effect treatments

Citric acid (CA) Firmness Ascorbic acid (AA)

Firmness

0% 26.2 0% 24.01% 27.1 1% 27.7

2% 28.3

SED (CA) = 0.70; SED (AA) = 0.86Microbial growth: An increase was observed in both total viable count (TVC) and yeast and mould (Y:M) count over the storage period (Fig. 5).

and citric acid was noted (Fig. 9). This showed that combinations of both acids resulted in an increase in Y:M count. Acidic environments were found to be favourable for yeast and mould growth (Beaulieu and Gorny, 2002; Jay, 1992). Although the Y:M count increased during storage, the counts were still low on day 15. According to Debevere (1996), the recommended Y:M count for fresh-cut vegetables should not exceed Log10 104

CFU/g (4.0).Table 9. Two-way table of treatment means for Y:M count in Log10 CFU/gCitric acid (%) Ascorbic acid concentration (%)

0 1 20 2.81 2.00 2.141 2.05 2.45 2.69

SED (interaction) = 0.162.

The low TVC and Y:M counts can also be explained by the possible development of a high CO2 atmosphere inside the packages. High CO2 atmospheres inhibit most aerobic microorganisms, especially gram-negative bacteria that cause off-fl avours and off-odours (Martin-Belloso, 2006; Farver and Dodds, 1995). Elevated CO2 environments are also fungistatic (Gorny, 1997). However lactic acid bacteria are unaffected by high CO2 and they continue to grow (Farver and Dodds, 1995). Growth of lactic acid bacteria may result in a decrease in pH and may produce antimicrobial compounds, which may suppress the growth of pathogens (Francis et al., 1999).

Slow growth of bacteria, yeasts and moulds was also due to the low storage temperature. Results obtained on shredded chicory salads, and shredded lettuce (Nguyen-The and Carlin, 1994) showed that growth of mesophilic microfl ora was signifi cantly reduced at low storage temperatures. The Institute of Food Science and Technology recommends a storage temperature in the range of 0-5°C (Francis et al., 1999) for prepared salads to limit the growth of pathogens. In this study, the storage temperature was maintained in the range of 2-4°C, thus explaining the lower plate counts observed. Low temperature storage is crucial for maintaining the microbiological quality of fresh-cut green jackfruit.

Two percent ascorbic acid was found to prolong the shelf life of fresh-cut green jackfruit, but its effect was limited to only 7 days at 2-4°C. The combination of 1% citric acid and 1% ascorbic acid was successful in inhibiting browning up to 13 days at 2-4°C, while the combination of 1% citric acid and 2% ascorbic acid was the most suitable treatment in maintaining the quality attributes of the fresh-cut jackfruit for up to 15 days at 2-4°C. This treatment also prevented microbial growth although the acidic conditions did favour the growth of yeast and mould, which were still less than the permissible amount. The use of the combined preservative factors has greater effectiveness in preserving quality attributes and delaying microbial growth than when used singly.

ReferencesAgar, I.T., R. Massantini, B. Hess–Pierce and A.A. Kader, 1999.

Postharvest CO2 and ethylene production and quality maintenance of fresh-cut kiwifruit slices. J. Fd. Sci., 64(3): 433-440.

Askar, A. and H. Treptow, 1993. Quality Assurance in Tropical Fruit Processing. Springer Laboratory, Germany.

Fig 5. TVC and Y:M count during storage of the diced jackfruit. LSD (TVC) = 0.711; LSD (Y:M count) = 0.3619

Both citric and ascorbic acid signifi cantly (P<0.05) decreased the microbial load of the fresh-cut jackfruit (Table 8). No signifi cant interaction effect (P>0.05) was noted between the two treatments. Dipping fresh-cut produce in ascorbic acid/citric acid prior to packaging signifi cantly decreased aerobic and anaerobic counts (McLachan and Stark, 1985; O’Beirne and Ballantyne, 1987). These had an impact on the product’s shelf life. Table 8. Mean TVC (in Log10 CFU/g) of the jackfruit dices for main effect treatments

Citric acid (CA) TVC Ascorbic acid (AA)

TVC

0% 4.04 0% 4.291% 3.37 1% 3.64

2% 3.18

SED (CA) = 0.18, SED (AA) = 0.231 % CA was found to be more effective in reducing plate count as compared to 1% AA. However 2% AA reduced the plate count in almost the same way as 1% CA. The antimicrobial effects of the acids can be explained by the fact that they lower the pH resulting in unfavourable conditions for the growth of bacteria (Piagentini et al., 2003). Although the TVC continued to increase during storage, the total count was quite low and below the recommended level of Log10 107 CFU/g (7.0) as per Mauritian Food Regulations (1999).

No signifi cant effects (P>0.05) on Y:M count were observed when using citric acid and ascorbic acid individually. However a signifi cant interaction effect (P<0.05) between ascorbic acid

44 Preservation of fresh-cut green jackfruit

Days

0

1

2

3

4

5

6

0 3 6 9 12 15

Mean TVC MeanY:M count

log

CFU

/g

Association of the Offi cial Analytical Chemists, 1995. Offi cial methods of analysis of AOAC International, Volume II. (P. Cunniff, ed). AOAC International, USA.

Beaulieu, J.C. and J.R. Gorny, 2002, Fresh-cut fruits, <http: //www.ba.ars.usda.gov/hb66/146freshcutfruits.pdf>

Debevere, J. 1996. Criteria en practische methoden voor de bepaling van de houd baarhei dsdatum in the etikettering. In: Etikettering houdbaarheid en bewaring, (G. Temmerman, G. Cremer, M. Thyssen, J. Devebere, eds). Brugge, Belgium. Die Keure, p. 37-64.

Dong, X., R.E. Wrolstad and D. Sugar, 2000. Extending shelf life of fresh-cut pears. J. Fd. Sci., 65(1): 181-186.

Escalona, V.H., A. Encarna and A. Francisco, 2005. Overall quality throughout shelf life of minimally fresh processed fennel. J. Fd. Sci., 70(1): 13-17.

Eskin, M.N.A. 1990. Biochemistry of Foods. Academic Press, Inc. USA.

Fan, X., B.A. Niemera, J.P. Mattheis, H. Zhuang and D.W. Olson, 2005. Quality of fresh-cut apple slices as affected by low-dose ionizing radiation and calcium ascorbate treatment. J. Fd. Sci., 70(2): 143-148.

Farver, J.M. and K.L. Dodds, 1995. Principles of modifi ed atmosphere and sousvide product packaging. Technomic Publishing Company, Inc. USA.

Francis, G.A., C. Thomas and D. O’Beirne, 1999. The microbiological safety of minimally processed vegetables. Intl. J. Fd. Sci. Technol., 34: 1-22.

Gimenez, M., C. Olarte, S. Sanz, Lomas, J.F. Echavarri and F. Ayala, 2003. Infl uence of packaging fi lms on the sensory and microbiological evolution of minimally processed borage (Borrago offi cinalis). J. Fd. Sci., 68(3): 1051-1057.

Gonzalez-Aguilar, G.A., C.Y. Wang and J.G. Buta, 2000. Maintaining quality of fresh-cut mangoes using antibrowning agents and modifi ed atmosphere packaging. J. Agr. Fd. Chem., 48(9): 4204-4208.

Gorny, J.R. 1997. Modifi ed atmosphere packaging and the fresh-cut revolution. Perishables Handling Nwsl. 90.

Heimdall, H., B.F. Kuhn, L. Poll and L.M. Larsen, 1995. Biochemical changes and sensory quality of shredded and MA packaged iceberg lettuce. J. Fd. Sci., 60(6): 1265-1268.

Jay, M.J. 1992. Spoilage of fruits and vegetables. In: Modern Food Microbiology. Chapman and Hall, London. p.187.

Kim, D.M., K.H. Kim, N.L. Smith and C.Y. Lee, 1995. Changes in fl esh colour and PPO activity by apple cultivars. Fd. Biotechnol., 4(4): 222-225.

Langdon, T.T. 1987. Preventing browning in fresh prepared potatoes without the use of sulphiting agents. Fd. Technol., 41(5): 64.

Laurila, E. and R. Ahvenainen, 2002. Minimal processing in practice. In: Minimal Processing Technologies in the Food Industry, (T. Ohlsson and N. Bengtsson, eds). Woodhead Publishing Ltd. p.219-233.

Martin-Belloso, O. and R. Soliva-Fortuny, 2006. Effect of modifi ed atmosphere packaging on the quality of fresh-cut fruits. Stewart Postharvest Rev., 2(1): 1-8.

Mauritian Food Regulations, 1999. Microbiological Standard for Specifi ed Food. Ready to eat food other than those specifi ed above. Government Gazette of Mauritius No.114. Government Notice No.173.

McLachan, A. and R. Stark, 1985. Modifi ed atmosphere packaging of selected prepared vegetables. Campden Food Preservation Research Association Techical Memorandum. Chipping Campden, UK, 1985, 78 pp.

Nakasone, H.Y. and R.E. Paul, 1999. Tropical fruits. In: Crop Production Science in Horticulture. CABI International, Hawaii. p.334-338.

Nguyen-The, C. and E. Carlin, 1994. The microbiology of minimally processed fresh fruits and vegetables. Critical Rev. Fd. Sci., 34: 371-401.

O’Beirne, D. and A. Ballantyne, 1987. Some effects of modified atmosphere packaging and vacuum packaging in combination with antioxidants on quality and storage life of chilled potato strips. Intl. J. Fd. Sci. Tech., 22: 515-523.

Paul, P.C. and H.H. Palmer, 1972. Fruits and vegetables. In: Food Theory and Applications. John Wiley & Sons, Inc. p. 294-297.

Piagentini, A.M., D.R. Guemes and M.E. Pirovani, 2003. Mesophilic aerobic population of fresh-cut spinach as affected by chemical treatment and type of packaging fi lm. J. Fd. Sci., 68(2): 602-607.

Riquelme, F., M.T. Pretel, G. Martinez, M. Serrano, A. Amoros and F. Romajaro, 1999. Packaging of fruits and vegetables: recent results. In: Food Packaging and Preservation, M. MathLouthi (eds.). Aspen Publishers Inc. p.144-153.

Sapers, G.M. 1993. Browning of foods: control by sulphites, antioxidants and other means. Fd. Technol., 47: 75-84.

Sapers, G.M., K.B. Hicks and R.L. Miller, 2002. Antibrowning agents. In: Food Additives, A.L. Brane, P.M. Davidson, S. Salminen, J.H. Thorngate (eds.). Marcel Dekker. p.543-561.

Sapers, G.M. and R.L. Miller, 1995. Heated ascorbic/citric acid as browning inhibitor for pre-peeled potatoes. J. Fd. Sci., 60(4): 762-776.

Soliva-Fortuny, R.C., N. Grigelmo-Miguel, I. Odriozola-Serrano, S. Gorinstein and O. Matin-Belloso, 2001. Browning evaluation of ready-to-eat apples as affected by modifi ed atmosphere packaging. J. Agri. Fd. Chem., 49(8): 3685-3690.

Tovar, B., L.I. Ibarra, H.S. Garcia and M. Mata, 2000. Some compositional changes in Kent mango (Mangifera indica) slices during storage. J. Appl. Hortic., 2(1): 10-14.

Wang, C.Y. 1990. Physiological and biochemical effects of controlled atmosphere on fruits and vegetables. In: Food Preservation by Modifi ed Atmospheres, M. Calderon and R. Barkai-Golan (eds.). CRC Press, Inc. p.199.

Preservation of fresh-cut green jackfruit 45

Journal

ApplJournal of Applied Horticulture, 11(1): 46-49, January-June, 2009

Use of plastic shades to regulate growth of korarima (Aframomum corrorima (Braun) P.C.M. Jansen)

S. Eyob

Horticulture Department, Awassa College of Agriculture, Hawassa University, P.O.Box 5, Awassa, Ethiopia. E-mail: [email protected]

AbstractKorarima (Aframomum corrorima (Braun) P.C.M. Jansen), a slow growing and persistent under tree shade as an under-story perennial plant, is native to Ethiopia. When it is grown in full sunny condition, all plants die off a few weeks after planting, but the effect of different shading materials on its growth is not known. Half a year old korarima plants were planted under differently coloured plastic shades (red, green, blue and clear) and coffee (Coffea arabica L.) tree shade to regulate the growth. The coffee tree shade was used as control. Varying levels of photosynthetic photon fl ux density (PPFD) and red to far red (R/FR) ratio of light were recorded under different shaded and open conditions. The korarima plant responded differently to the different plastic and coffee tree shades. Average plant height, number of leaves per plant, number of sprouts per plant, chlorophyll content, leaf area, total fresh and dry weights were signifi cantly different when recorded at different stages of growth, highest being recorded under the blue plastic cover. The minimum effi ciency was achieved under control.

Key words: Korarima, Aframomum corrorima, Mesketo, photon fl ux density, plastic shade, tree shade, growth regulation.

Introduction Korarima (Aframomum corrorima (Braun) P.C.M. Jansen) belongs to the family Zingiberaceae and native to Ethiopia (Jansen, 1981; Lock, 1997). It has been used as a spice crop in Ethiopia since time of Queen Sheba. The seed, pod, leaf, rhizome and root of korarima are also used as traditional medicine in Ethiopia (Eyob et al., 2008). It is a perennial aromatic shade loving herb; with scaly rhizomes and leafy stems reaching 1-2m height. It grows slowly and persistent under tree canopy as an under-story plant. It is mainly grown in the natural forests dominated by woody species.

Geoffrey and Martin (2000) described that many changes in the environment can occur when the gap is created in the forest. The changes may include alteration in photosynthetically active radiation (PAR), light intensity and quality, temperature, competition for moisture, humidity and nutrient availability. Perhaps the most critical management requirement is needed for a well-adapted shade loving plant species such as korarima, which grow under forest vegetation. Adaptation to shade does not signifi cantly change leaf radiation adsorption effi ciency in tree species (Baltzer and Thomas, 2005). However, since plants convert only a small percentage of absorbed radiation into photosynthate facilitating growth, absorption effi ciency may not be a relevant measure of adaptation benefi ts. Plants have evolved very sophisticated sensory networks for monitoring the status of several important features of their surroundings. Shahak et al. (2004) reported that the spectral manipulation promotes physiological responses, while light scattering improves light penetration into the plant canopy.

Korarima is traditionally grown in home gardens under tree shades, and in the natural forests in association with wild coffee

(Coffea arabica L.). Review of indigenous production practices in Ethiopia showed a decreasing trend in production areas and yield in two decades as a result of shortage of permanent shade trees in the household farms and destruction of natural forests. As reported previously by Kendrick and Kronenberg (1994) the economic success of crop production requires proper management of the solar radiation resource as light quality has a signifi cant effect on plant growth, development and productivity. With respect to korarima, it is well known fact that its cultivation is unsuccessful where the shade trees are not available. Hence, the need for permanent tree shades hinders large scale production of crop. Uses of covers that block parts of the incoming radiation have been used in areas with high solar radiation and excess temperatures. Therefore, the objectives of this study were to determine the comparative advantages and optimal shade management strategies using locally produced plastic fi lms over tree shades and open conditions for growth of korarima.

Materials and methodsPlant material and experimental design: Approximately half a year old planting materials from cultivar Mesketo were planted (one plant per pot) in plastic pots (Ø 40cm) fi lled with forest soil and compost mixture, and grown in the protected garden of Horticulture Department, Hawassa University, Ethiopia. The plants were managed properly as per agronomic practices. Four plastic pots were replicated three times for each treatment consisting of 12 pots per treatment in completely randomized design. Locally produced photo selective red, green, blue, and clear plastic fi lms partially transparent for sun light were used as experimental treatments. Coffee tree shades were used as control treatment. The plastic fi lms used for this purpose are commonly used as shades in the local open markets and road sides in Ethiopia. The plastic fi lms were mounted on sticks ca

2m above the ground. On all sides, 1 m above the ground, the plastic shades were open to maintain free air circulation at the same level as the shades from the coffee trees. The coffee trees grew ca 35m from plastic shades and were used to simulate natural growing condition of koraima. Twelve pots were placed under open conditions for purpose of comparison.

The photosynthetic photon fl ux density (μmolm-2s-1) and red to far red ratio of light were measured under different plastic shades, coffee trees and open condition by using a SKP 215 PAR Quantum Sensor before the plants were planted (Table 1).

Measurements: Sixty days after planting (DAP), number of fully opened leaves per plant and number of newly emerging young shoots at the base of mother plants and chlorophyll content were recorded on a monthly basis from June 2005 to May 2006. A fi eld portable, hand held chlorophyll content meter (Model CL-01, Hansatech Instruments Ltd, England) was used at day time from 2: 00 to 3:00 p.m. to estimate average total chlorophyll content in leaves of korarima. Leaf area was measured using portable leaf area meter, Model LI-3000A after removing all leaves from plants grown for destructive measurements and average leaf area was estimated for a single plant at 300 DAP. In the end of the experiment fresh and dry weights were recorded. Total fresh weight of sampled plants was measured at 360 DAP and the dry weight was recorded after oven drying at 100°C for 2 days. Soil particles and any debris were removed from roots before taking measurements.

Data analysis: The data were analysed as a completely randomised design with repeated measurements three times. Fisher least signifi cant difference (LSD) test was used to compare means at P=0.05. On randomly selected plants regression analysis was carried out on the number of sprouts per plant and leaf number per plant as independent and dependant variables, respectively to see any relationship and trends in growth of korarima at 240 DAP. The software used for this statistical analysis was MINITAB 13th edition.

Results The varying coloured plastic sheets signifi cantly affected PPFD and R/FR level under plant canopy inside the shades. As a consequence of extreme increase of PPFD (1410 μmolm-2s-1) and R/FR (1.12) under open condition (Table 1), plants showed very slow rate of increase in shoot height for four weeks (data not shown) and died off afterwards while those under shade showed a continuous growth. Under shaded conditions, the plants showed signifi cant differences between the treatments with regard to growth performances at P=0.05.. The variation in

plant height between treatments was signifi cant (Table 2). For all the treatments, there was a gradual increase in plant height from 60 DAP to 210 DAP. However, shoots under the control treatment were weak and slow in the growth. Remarkably strong and erect plants were observed under blue plastic sheet compared to other treatments. Signifi cantly higher plant height (109.0cm) was recorded at 210 DAP in blue plastic sheet compared to the control (68.3cm) when recorded on the same date.

In case of leaf number also, the variation between treatments was signifi cant (Table 3). There was a gradual increase in leaf number in different stages of growth. The increase under the control was low. Signifi cantly increased leaf number was recorded in blue plastic (20.0) as compared to the control (11.3), clear plastic (12.3) and red plastic (12.3) at 210 DAP. Changes in leaf number for the clear plastic and the control were low during growth period.

The varying coloured plastic sheets induced a signifi cant effect on the number of sprouts per plant in almost all the treatments used (Fig. 1). The number of sprouts were most enhanced with about 50% of increase in blue sheet when compared with the control and clear sheet starting from 150 DAP to 210 DAP. The other treatments showed about 20-30% increase over the control and clear plastic.

The variation in leaf area at 300 DAP due to shading effect was signifi cant. Plants grown under blue plastic shade had the highest leaf areas (4.22m2) compared to the control (1.62m2) and other treatments (Table 4).

Table 2. Effect of plastic shading on plant height (cm) of korarima at different stages of growthTreatment Days after planting (DAP)

60 90 120 150 180 210Red plastic 29.3 33.7 48.0 72.7 75.3 77.0 Green plastic 32.3 49.7 61.0 86.00 88.3 81.7 Blue plastic 34.7 71.3 83.3 104.8 106.0 109.0 Clear plastic 22.7 31.3 43.0 61.7 65.7 74.7 Control 23.3 27.0 37.0 56.4 58.3 68.3 LSD 6.3 19.9 21.1 16.2 15.3 15.4

Table 3. Effect of plastic shading on leaf number per plantTreatment Days after planting (DAP)

60 90 120 150 180 210Red plastic 8.0 8.7 9.7 11.0 11.3 12.3Green plastic 9.3 11.3 14.00 16.7 17.3 18.3Blue plastic 11.3 13.7 15.7 17.3 19.0 20.0Clear plastic 5.0 8.7 10.7 11.3 11.7 12.3Control 3.0 4.0 6.7 8.7 10.3 11.3LSD 5 5.8 4 3.5 3.4 3.4

Table 4. Effect of different shade treatments on the leaf area, fresh weight and dry weight per plant Treatment Leaf area (m2)

300 DAPFresh wt (g)

360 DAPDry wt

(g) Red plastic 3.31 2982.5 231.4

Green plastic 3.80 3075.1 262.0Blue plastic 4.22 3766.4 394.9Clear plastic 2.13 2029.2 193.8Control 1.62 1633.9 112.7LSD 0.23 351.1 51.2

Table 1. Photosynthetic photon fl ux density (PPFD) and red to far red (R/FR) ratio of light under different shading and open conditionsShading materials PPFD

(μmolm-2s-1)R/FR

Red plastic 680 0.98Green plastic 462 0.68

Blue plastic 243 0.66Clear plastic 790 1.01Coffee tree shade (control) 102 0.53Open condition 1410 1.12

Use of plastic shades to regulate growth of korarima 47

The shade treatments increased fresh and dry weight. The highest fresh (3766.4g) and dry weight (394.9g) were recorded under blue plastic shade (Table 4). Increase in leaf area accounted wholly for the increase in plant fresh and dry weight.

The chlorophyll content in leaf was affected signifi cantly by shade treatments, the highest being recorded in blue plastic shade when recorded at 90 DAP and the lowest was in the control (Fig. 2). In all treatments the chlorophyll content gradually decreased after 90 DAP. However, under blue plastic shade they seemed to increase again after 180 DAP.

The varying level of PPFD and R/FR induced by different shading materials showed considerable difference on korarima growth. The PPFD (102.2 μmol m-2s-1) and R/FR (0.53) under the control treatments resulted in poor growth performances of korarima. On average 82.8% decrease and 58% increase in PPFD by blue plastic shade over open condition and control, respectively, regulated growth of korarima. Most of the growth characteristics were highest at interaction level of PPFD at 243 μmolm-2s-1 and R/FR at 0.66 under blue plastic.

The simple linear regression analysis showed a significant relationship between sprouts per plant as independent variable and leaf number as a response variable when tested at P=0.05. For a unit increase in sprout number per plant, there was linear increase in leaf number by 1.964 as indicated in Fig. 3. It is clear that 80.4% variability (R2=0.804) in leaf number per plant could be explained by sprout number per plant.

Discussion The potential importance and results obtained from growth performance of korarima are presented in this study. Shading materials signifi cantly infl uenced the rate of vegetative growth of korarima. In the blue plastic shade, plant vegetative growth was signifi cantly promoted. The growth performance of the plant was reduced in all stages of growth under tree shades indicating that tree canopy intercepted most part of incoming solar radiation. This explains the existing fact that korarima is indigenous slowly growing species in the forests of Ethiopia as they appear more or less isolated or in clumps under trees. Total death of plants under open condition might be due to heat from a high portion of near infra-red radiation (NIR, 700 to 1100nm).

It was observed that the plant height under plastic shade showed greater increase in length compared to those of under control. The remarkable increase in plant height, leaf number and number of sprout under blue plastic shade during growth period over other treatments indicate differences in response of the rate of growth under the various light regimes. In our experiment, etiolated and very weak plants with dull green leaves were observed under the control as compared to plastic shades explaining the need of enough light sources during photosynthesis for the manufacture of food. In the shortage of light, suffi cient energy is not available for plants as observed in control.

Leaf area increased faster under blue plastic due to a higher leaf appearance rate, which made possible the growth of young sprouts at the base of mother plant, as photosynthetically active leaves were available. After 90 DAP chlorophyll content in leaf was declined gradually under all shade treatments implying the

Fig. 3. Linear relationship between sprouts number per plant and leaves number per plant at 240 DAP.

0

2

4

6

8

10

12

14

16

60 90 120 150 180 210Days after planting (DAP)

Aver

age

num

bero

fspr

outs

/pla

nt Red plastic

Green plastic

Blue plastic

Clear plastic

Control

Fig. 1. Effects of artifi cial shading materials on average number of sprouts per plant at different stages of growth.

Fig. 2. Effect of artifi cial shading materials on average chlorophyll content of korarima at different stages of growth.

12108642

25

20

15

10

5

0

Number of sprouts per plant

Num

ber o

f lea

ves

per p

lant

Number of leaves per plant= 0.71+1.964 Number of sprouts per plant

R =80.42

0

2

4

6

8

10

12

14

16

18

60 90 120 150 180 210

Days after planting (DAP)

Ave

rage

tota

lchl

orop

hyll

cont

ent u

nits

Red plastic

Green plastic

Blue plastic

Clear plastic

Control

48 Use of plastic shades to regulate growth of korarima

assimilation of photosynthetic products for growth by young growing sprouts. The increased total fresh and dry weights per plant at maturity were associated with enhanced effect of leaf area production under blue plastic.

Compared to other plastic colours, blue plastic decreased PPFD and R/FR to optimum level and resulted in better growth performance of korarima as a result of the partial control of climate inside the plastics. The signifi cance of the climatic control systems inside the greenhouses have been suggested by Arellano et al. (2006). This experimental fi nding shows that artifi cial coloured covering materials are effective in regulating vegetative growth. This is in conformity with fi nding of Donald et al. (2005), Murakami et al. (1997), and Rajapaske and Kelly (1992). Generally with moderate reduction of PAR, there were increase in korarima growth performance. The overall effect of heavy shading resulted in tiny and a poorly grown plant under the control, which is in agreement with work of Bartlett and Remphrey (1998).

The results reported in this work, in brief, suggest that in the absence of enough forest gaps for suffi cient light quality to infi ltrate and initiate better growth of korarima, physiological and morphological response would be affected negatively. The difference in growth response achieved due to shade treatments suggests various possibilities of fi eld management strategies using locally available plastic sheets which could vary depending on light source to optimise production. Moreover, korarima response to various level of PAR suggesting a high degree of fl exibility in terms of growth and resource utilization even though the korarima plant exist in the forest under natural conditions. Further we suggest development of new korarima varieties through conventional breeding and molecular techniques for maximum and effi cient utilization of total light sources under open conditions to promote production at large scale in the future, which also avoids cost of shading materials.

AcknowledgmentsAuthor gratefully acknowledges research support received from Norwegian Agency for Development Cooperation (NORAD) research project in collaboration with Hawassa University, Ethiopia. The author also thanks Associate Professor Maigull Appelgren, Norwegian University of Life Sciences, Department

of Plant and Environmental Sciences and Associate Professor Admasu Tsegaye, Hawassa University, Awassa College of Agriculture, Horticulture Department for their valuable comments and suggestions during preparation of this manuscript.

ReferencesArellano, M.A., S. Garcia, A. Sanchez, J. Soria-Ruiz, D.L.Valera and

M. Urrestarazu, 2006. Greenhouse microclimate and its natural variation in two subtypes of an almería greenhouse. Acta Hort., 719: 147-156.

Baltzer, J.L. and S.C. Thomas, 2005. Leaf optical response to light and soil nutrient availability in temperate deciduous trees. Am. J. Bot., 92: 214-223.

Bartlett, G.A. and W.R. Remphrey, 1998. The effect of reduced quantities of photosynethetically active radiation on Fraxinus pennsylvanica growth and architecture. Can. J. Bot., 76: 1359-1365.

Donald, T.K., C.H. David and M.M. Roman, 2005. Spectral properties of selected UV-blocking and UV-transmitting covering materials with application for production of high-value crops in high tunnels. Photochem. Photobiol., 81: 1047-1051.

Eyob, S., M. Appelgren, J. Rohloff, A. Tsegaye and G. Messele, 2008. Traditional medicinal uses and essential oil composition of leaves and rhizomes of korarima Aframomum corrorima (Braun) P.C.M. Jansen) from southern Ethiopia. S. Afr. J. Bot., 74: 181-185.

Geoffrey, G.P. and J.B. Martin, 2000. Forest canopy stratifi cation-Is it useful? Am. Nat., 155: 473-484.

Jansen, P.C.M. 1981. Spices, Condiments and Medicinal Plants in Ethiopia: Their Taxonomy and Agricultural Significance. Agricultural research reports 906. Center for Agricultural Publishing and Documentation, Wageningen, the Netherlands.

Kendrick, R.E. and G.H.M. Kronenberg, 1994. Photomorphogenesis in Plants. 2nd ed. Kluwer Academic Publishers, Dordrecht, Netherlands.

Lock, J.M. 1997. Zingiberaceae. In: Flora of Ethiopia and Eritrea, Vol. 6, S. Edwards, S. Demissew and I. Hedberg (eds.). Addis Ababa, Ethiopia and Uppsala, Sweden. p. 324-326.

Murakami, K., H. Cui and M. Kiyota, 1997. Control of plant growth by covering materials for greenhouses which alter the spectral distribution of transmitted light, Acta Hort., 435: 123-130.

Rajapaske, N.C. and J.W. Kelly, 1992. Regulation of chrysanthemum growth by light quality. J. Amer. Soc. Hort. Sci., 117: 481-485.

Shahak, Y., E.E. Gussakovsky, E. Gal and R. Ganelevin, 2004. ColourNets: Crop protection and light-quality manipulation in one technology. Acta Hort., 659: 143-151.

Use of plastic shades to regulate growth of korarima 49


Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant

Renu Singh, Ram C. Yadav* and Neelam R. Yadav

Department of Biotechnology and Molecular Biology, CCS Haryana Agricultural University, Hisar. 125004, India*E-mail: [email protected]

AbstractA study was carried out on mature female papaya (Carica papaya L.) plant of Selection 1 cultivar by using axillary bud as an explant and media supplementation with the main aim to assess the effect of growth regulators (auxins, cytokinins) and silver nitrate on in vitro regeneration of female papaya plant. Total of 28 media were used for shoot regeneration while for root regeneration total of eight media were tested supplemented with different growth hormones. Based on the results of this study, for shoot proliferation, MS basal medium supplemented with BAP (1.0 mg L-1) and BAP (2.0 mg L-1) + NAA (0.1 mg L-1) was found to give the best results while MS medium supplemented with IBA (2.0 mg L-1) gave best rooting percentage. Besides, auxins and cytokinins, effect of silver nitrate (AgNO3) on plant regeneration from axillary buds taken from mature female papaya plant was also carried out.

Key words: Auxins, axillary bud, benzylaminopurine (BAP), cytokinins, indole acetic acid (IAA), Carica papaya L., α-naphthalene acetic acid (NAA), proliferation, silver nitrate (AgNO3)

IntroductionPapaya (Carica papaya L.) is a native of tropical America and belongs to family Caricacea. The plant starts fruiting in just one year and gives economically high yield per acre next to banana. The total area under papaya cultivation in India was 57,000 ha (FAO, 2000). Papaya is very nutritious and has much therapeutic value. The fruit contains fair amount of vitamin A, B1, B12 and C. The tremendous yielding potential of this crop is left economically unexploited due to several problems associated with its cultivation. The main problem associated is lack of clonal multiplication methods and in tissue culture bacterial contaminants are the limiting factors (Litz and Conover, 1981).

It is well known that vegetative propagation of papaya through conventional methods has not been successful with commercial viability. Thus, there is ample scope to overcome the above limitations through tissue culture. The plant tissue culture technique has been successfully employed in the regeneration of various fruit crops like almonds, strawberry, citrus, etc (Bajaj, 1986). However, papaya is a problematic crop for micropropagation from grown up plants as it contains a good amount of latex interfering in proliferation. Therefore, the present study was undertaken with a view to evolve a tissue culture technique in papaya and to evaluate the effect of different growth regulators on the vegetative regeneration of female papaya plant using axillary buds.

Materials and methodsThe present investigation was carried out in Tissue Culture Laboratory of the Department of Biotechnology and Molecular Biology, CCS Haryana Agricultural University, Hisar, India. Axillary buds as explants were excised from the fi eld grown sexually differentiated female plants. The chemicals of high purity were used throughout the course of investigation. The excised

explants i.e. axillary buds were cut (6 mm) and washed with single glass distilled water containing few drops of teepol. The explants were then treated with 0.5 per cent Bavistin (fungicide) for 45 minutes. The disinfections steps after bavistin treatment were carried out in the horizontal laminar fl ow cabinet. Explants were further rinsed 4-5 times with distilled water followed by 70 per cent alcohol treatment for 30 seconds and by 50 per cent sodium hypochlorite for 3 minutes. After this treatment, they were rinsed 4-5 times with double distilled water fi nally, given 0.1% HgCl2 treatment for 4 minutes and then the explants were again rinsed 5-6 times with autoclaved double distilled water. In the present study, MS basal medium supplemented with different auxins and cytokinins were initially tried for culture establishment (Murashige and Skoog, 1962). The medium supplemented with kinetin, NAA, BAP and AgNO3 giving better response were further used for regeneration and those which don’t gave better regeneration were discarded in the initial of the experiment. All the cultures were maintained at 25 ± 1°C under 16/8 hours cycles of light (light intensity 50 μmol m-2 s-1) and dark. The regenerated shoots were further cultured on regeneration medium with variable concentration of growth hormones. The multiple shoots formed were then transferred to rooting media in order to get root formation. The data was recorded time to time and presented as the mean of three repeats. Data were analyzed statistically using one-way analysis of variance (ANOVA).

Results and discussionIn general, poor response for shoot regeneration on all the media was observed when axillary bud was used as explant (Fig. 1A). Maximum (70.1%) response was observed on medium supplemented with BAP (Table 1) while very poor response was observed on the media supplemented with kinetin (Table 2). No shoot regeneration was observed on most of the Media supplemented with kinetin.

Journal

Appl

Maximum (70.1%) response (Fig. 1B) was observed on medium supplemented with and moderate response was observed on MS supplemented with BAP (2.0) + NAA (0.1), BAP (1.0) + IAA (0.2) and BAP (2.0) + NAA (0.5) (65.0, 51.1 and 45.8%, respectively), while rest of media gave poor response (Table 1). In case of kinetin maximum shoot regeneration (21.4%) response was observed on medium supplemented with Kinetin (2.5) and minimum response was observed on medium supplemented with Kinetin (5.0) + NAA (0.2) (11.7%).

When AgNO3 was added in the media (Table 3) supplemented with other growth hormones, the regeneration response was observed maximum (59.0%) on MS medium supplemented with BAP (1.0) + AgNO3 (2.0) + Zeatin (3.0). Moderate (50.0%) regeneration response was observed on MS + BAP (2.0) + AgNO3 (2.0) + Zeatin (2.5) medium, while MS medium supplemented with AgNO3 (2.0) and AgNO3 (1.0) gave comparatively poor response (24.8 and 13.2%, respectively). No response was observed on MS medium without any growth regulators.

The regenerated plantlets from micropropagation experiments grown in vitro were transferred to different rooting media for carrying out their rooting in vitro (Fig. 1C). Roots formed were thick and strong (Fig. 1D). No response was observed when medium was devoid of any auxin. Maximum rooting frequency (67.2%) was observed on MS medium supplemented with IBA (2.0); however, it was minimum (22.2 and 27.8%) on medium supplemented with IBA (0.1) and IBA (0.5), respectively (Table 4).

In vitro micropropagation has been successfully used for many horticultural fruit trees (Das et al., 1996). Multiple shoot production from axillary buds obtained from mature trees is now recognized as a better alternative of micropropagation in fruit trees where fi delity of the propagules is of prime importance (Quarashi and Mitra, 1998; Tavares et al., 1996).

The overall conclusion from the above regeneration experiment emerged that the media comprised of MS basal salt + BAP (2.0 mg L-1) + NAA (0.1 mg L-1) gave the best regeneration results in C. papaya. As outlined by Skoog and Miller (1957), root and shoot initiation is basically regulated by the interaction between auxins and cytokinins. This combination of both the hormones was equally effective when regeneration was done with the shoot tips therefore, we can conclude that a proper ratio of cytokinins and auxins helps in both the shoot and root formation from different explants. Axillary buds were associated with relative high level of endogenous growth substances in juvenile tissues in comparison to the adult tissues. This implies the poor regeneration of axillary buds from mature plants than the explants taken from the juvenile plant. The un-branched character of the papaya plant also offers a serious limitation in using axillary buds as explants in the experiment.

Effect of cytokinins: The most striking infl uence on bud-break and shoot multiplication has been found with the use of auxins and cytokinins (Normanly, 1995). The most commonly used cytokinins are BAP, kinetin, 2ip and zeatin, the latter two being natural cytokinin. Superiority of BAP over other cytokinins has been reported and discussed in relation to shoot proliferation in culture of trees (Bonga and Jvon Aderkas, 1992). In the present study, lower level (1.0 mg L-1) of BAP and of kinetin (2.5 mg L-1) induced highest frequency of shoot regeneration. Regeneration

response of BAP was superior over kinetin by taking less time for induction of shoot regeneration. In the present study, highest frequency of shoot regeneration and axillary bud proliferation could be achieved on BAP (1.0 mg L-1) without the need of subculture. BAP is reported to have favoured axillary shoot proliferation in several tree species (Eeswara, 1998; Purohit and Dave, 1996; Purohit and Singhvi, 1998). In the present study, it was observed that an increase in the level of cytokinin from 1.0 to 5.0 mg L-1 produced a negative effect on all the parameters except shoot number (Tables 1, 2).Table 1. Shoot induction response from axillary bud explants on media supplemented with BAP in papayaMedium Composition (mg L-1)

Average number of cultured explants

Average number of responding

explants

Mean (%) response(Mean ±S.E.)*

MSP 50 0 00.0 (04.05 ± 0.01)MS + BAP (1.0) 100 70 70.1 (57.14 ± 0.06)MS + BAP (2.5) 70 24 34.4 (36.14 ± 0.31)MS + BAP (5.0) 74 19 25.4 (30.56 ± 1.10)MS + BAP (1.0) + IAA (0.2)

74 38 51.1 (45.90 ±0.63)

MS + BAP (2.5) + IAA (0.2)

80 15 18.7 (26.01 ± 0.70)

MS + BAP (5.0) + IAA (0.2)

80 10 12.4 (21.01 ± 0.25)

MS + BAP (2.0) + NAA (0.5)

100 46 45.8 (42.88 ± 0.74)

MS + BAP (2.0) + NAA (0.1)

100 65 65.0 (53.99 ± 0.22)

MS + BAP (1.0) + NAA (0.2)

58 12 20.6 (27.32 ± 0.75)

MS + BAP (2.5) + NAA (0.2)

56 10 17.8 (25.32 ± 0.15)

MS + BAP (5.0) + NAA (0.2)

78 8 10.2 (19.04 ± 0.50)

LSD (P=0.05)=1.641, *Transformed valueTable 2. Shoot induction response from axillary bud explants on media supplemented with kinetin in papaya Medium Composition (mg L-1)

Average number of cultured

explants

Average number of responding

explants

Mean (%) response (Mean ±S.E.)*

MSP 50 00 00.0 (04.05 ± 0.01)MS + Kin (1.0) 54 00 00.0 (04.05 ± 0.01)MS + Kin (2.5) 74 16 21.4 (27.91 ± 0.53)MS + Kin (5.0) 58 10 16.9 (24. 61 ± 0.68)MS + Kin (1.0) + IAA (0.2)

84 00 00.0 (04.05 ± 0.01)

MS + Kin (2.5) + IAA (0.2)

48 00 00.0 (04.05 ± 0.01)

MS + Kin (5.0) + IAA (0.2)

48 00 0.00 (04.05 ± 0.01)

MS + Kin (1.0) + NAA (0.2)

60 00 00.0 (04.05 ± 0.01)

MS + Kin (2.5) + NAA (0.2)

80 14 17.5 (25.06 ± 0.55)

MS + Kin (5.0) + NAA (0.2)

84 10 11.7 (21.07 ± 1.24)

MS + Kin (0.5) + 2,4-D (2.5)

74 00 00.0 (04.05 ± 0.01)

MS + Kin (1.0) + 2,4-D (2.5)

100 13 12.0 (20.67 ± 0.28)

LSD (P=0.05)=1.379, *Transformed value

Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant 51

Effect of cytokinins (BAP and kinetin) in combination with auxins (NAA): In the present study combined effect of cytokinin and auxin was promotive on shoot regeneration from nodal explant, however, an inhibitory effect was observed in the combination where levels of cytokinins were high (5.0 mg L-1). Auxins (NAA, 2, 4-D and IAA) were also examined for their infl uence on multiple shoot production and growth of papaya axillary bud cultures. Auxins were found to affect only the growth of the cultures rather than proliferation rate. NAA (0.1-0.2 mg L-1) resulted in better growth of the cultures and hence included in the medium. Similar response by shoot and axillary bud cultures with regard to auxin application were reported in papaya (Buriken et al., 1988; Renveni, 1990). As a promotive effect on shoot regeneration in combination of BAP/kinetin with NAA, maximum shoot regeneration frequency in axillary bud explant (65.0%) was observed on basal medium with BAP (2.0 mg L-1) and NAA (0.1 mg L-1) while kinetin was found to have less promotive effect in combination with NAA. The complementary effect of cytokinin and auxin has been observed by Miller and Drew (1990) during shoot proliferation in papaya. Highest shoot production was achieved on medium containing BAP (2 μM) and NAA (0.5 μM). A combination of BAP (2.0 mg L-1) and NAA (1.0 mg L-1) was found to enhance shoot bud proliferation in cultured shoot tip in papaya (Litz and Conover, 1981). Combination of cytokinin and auxin thus has been found to promote shoot bud proliferation effectively.

Effect of auxins: Various auxins were tried with the objective to induce roots in shoots. In this study, medium containing MS basal + IBA (2.0 mg L-1) was found most appropriate for root induction followed by MS basal + IBA (1.0 mg L-1). However, the induction of rooting in the regenerated shoots of papaya was found to be very diffi cult. Plantlets produced with IBA were normal with regard to the shoot and root growth. IBA has been found to be effective for root induction in papaya-regenerated shoots by Bhattacharya et al., (2002), Buriken et al., (1988), Fitch (1993),

Rajeevan and Pandey (1983), Singh et al., (2000). Although IAA also induced root formation but the root development was poor as compared to IBA. The strength of the basic medium employed is reported to infl uence root initiation in several papaya cultivars. However, better root induction has been obtained in papaya with half strength MS medium (Renveni et al., 1990).

Effect of silver nitrate (AgNO3): AgNO3 is an ethylene inhibitor in plants which is reported to help in somatic embryogenesis (Songstad et al., 1991). AgNO3 may also serve as stress agent inducing endogenous ABA accumulation, Ag+ being a metallic ion may also promote somatic embryo production via an increase in the endogenous ABA levels (Kong and Yeung, 1994, 1995). But in our experiment when AgNO3 was added in the media (Table 3) supplemented with other growth hormones, the regeneration response was observed maximum (59.0%) on MS supplemented with BAP (1.0) + AgNO2 (2.0) + Zeatin (3.0). Thus, overall in conclusion media containing AgNO3 has no promotive effect on shoot formation.

ReferencesBajaj, Y.P.S. 1986. Biotechnology of trees improvement for rapid

propagation and biomass energy production. In: Biotechnology in Agriculture and Forestry, Vol. 1. Trees. Bajaj,Y.P.S (ed.), Springer, Berlin. 1-23.

Banerjee, J. 2002. Tissue Culture and Transformation Studies in Indian Cultivars of Papaya (Carica papaya L.), Ph.D. Thesis, Biotechnology Dept, University of Pune, India.

Table 3. Shoot induction response from axillary bud explants on media supplemented with AgNO3 in papaya Medium composition (mg L-1)

Average number of cultured explants

Average number of responding explants

Mean (%) response (Mean ±S.E.)*

MS 60 0 00.0 (04.05 ± 0.01)MS + AgNO3 (2.0) 120 30 24.8 (30.16 ± 0.52)MS + AgNO3 (1.0) 106 14 13.2 (21.71 ± 0.32)MS + BAP (1.0) + AgNO3 (2.0) + Zeatin (3.0) 100 59 59.0 (50. 48 ± 0.29)MS + BAP (2.0) + AgNO3 (2.0) + Zeatin (2.5) 100 50 50.0 (45.27 ± 0.01)LSD (P=0.05)=1.120, *Transformed

A

C

B

DFig. 1. A. Nodal explants cultured on regeneration medium; B. Multiple shoot formation from axillary explants; C. Regenerated shoot on rooting medium; D. Regenerated shoots with roots on rooting medium

Table 4. Root induction response in shoots regenerated from axillary bud explant on various rooting media in papaya

IAA (mg L-1)

NAA (mg L-1)

IBA (mg L-1)

Number of cultured

explants

Number of responding

explants

Mean (%) response

(Mean ±S.E.)*- - - 6 00 00.0 (04.05 ± 0.01)

2.0 - - 7 00 00.0 (04.05 ± 0.01)- - 0.1 12 2 15.0 (20.43 ± 8.25)- - 0.5 9 2 19.4 (23.30 ± 9.74)- - 1.0 9 4 44.4 (42.02 ± 3.25)- - 2.0 11 7 63.9 (52.46 ± 3.65)- 0.5 0.1 10 3 30.5 (33.79 ± 1.74)- 0.5 0.5 9 3 33.3 (35.53 ± 0.01)

LSD (P=0.05)=1.431, *Transformed value

52 Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant

Bhattacharya, J., S.S. Khuspe, N.N. Renukdas and S.K. Rawal, 2002. Somatic embryogenesis and plant regeneration from immature embryo explant of papaya (Carica papaya L. cv. Washington and Homey Dew.). Indian J. Exp. Bioi., 40: 624-627.

Bonga, J.M. and P. Ven Aderkas, 1992. In vitro culture of tree. Kluwer Academic Publishers, The Netherlands. 236p.

Burikan, S., V. Chommalee and S. Attathon, 1988. Effects of plant growth regulators on papaya ( Carica papaya) cultured in vitro. Kasetsart Journal Natural Sciences, 22 (5): 1-6.

Chengalrayan K., V.B. Mhaske and S. Hazra, 1997. High frequency conversion of abnormal peanut somatic embryos. Plant Cell Report,16:783-786.

Das, S., B.J. Timir and S. Jha, 1996. In vitro propagation of cashew nut (Anacardium occidenatle). Plant Cell Reports, 15: 615-619.

Eeswara, J.P. 1998. A standard procedure for the micropropagation of the Neem tree (Azadirachta indica A. Juss.). Plant Cell Rept., 17: 215-219.

FAOstats. 2001. www.fao.org Fitch, M.M.M. 1993. High frequency somatic embrygogenesis and plant

regeneration from papaya hypocotyls callus. Plant Cell Tissue Organ Culture, 32: 205-212.

Kong, L. and E.C. Yeung, 1994. Effects of ethylene and ethylene inhibitors on white spruce somatic embryo maturation. Plant Science, 104: 71-80.

Kong, L. and E.C. Yeung, 1995. Effects of silver nitrate and polyethylene glycol on white spruce (Picea glauca) somatic embryo development: enhancing cotyledonary embryo formation and endogenous ABA content. Physiol Plant., 43: 298-304.

Litz, R.E. and R.A. Conover, 1981. Effect of sex type, season and other factors on in vitro establishment and culture of Carica papaya L. explants. J. Amer. Soc. Hort. Sci., 106: 792-794.

Miller, R.M. and R.A. Drew, 1990. Effect of explant type on proliferation of Carica papaya L. in vitro. Plant Cell Tissue Organ Culture, 21 (1): 39-44.

Murashige, T. and F. Skoog, 1962. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plantarum, 15: 473-497.

Murashige, T. 1974. Plant propagation through tissue culture. Ann. Rev. Plant Physiol., 25: 135-166.

Normanly, J. 1995. Rethinking auxin biosynthesis and metabolism. Plant Physiol., 107: 323-329.

Purohit, S.D. and A. Dave, 1996. Micropropagation of Sterculia urens Roxb.: An endangered tree species. Plant Cell Report, 15: 704-706.

Purohit, S.D. and A. Singhvi, 1998. Micropropagation of Achras sapota through enhanced axillary branching. Sci. Hort., 76: 219-229.

Quarashi, A. and S.K. Mitra, 1998. Micropropagation of nodal explants from adult trees of Cleistanthus collinus. Plant Cell Report, 17: 430-433.

Rajeevan, M.S. and R.M. Pandey, 1983. Propagation of papaya through tissue culture. Acta Horticulturae, 131: 131-139.

Renveni, R.M. 1990. Propagation of papaya through tissue culture. Plant Cell Tissue Organ Culture, 13: 77-83.

Singh, H.P., S. Singh, R.P. Saxena and R.K. Singh, 1993. In vitro bud break in axillary nodal segments of mature trees of Acacia nilotica. Indian J. Plant Physiol., 36: 21-24.

Skoog, F. and C.O. Miller, 1957. Chemical regulation of growth and organ formation in plant tissue cultured in vitro. Sym. Exp. Biol., 11: 118-130.

Songstad, D.D., C.L. Armstrong and W.L. Petersen, 1991. AgNO3 increases Type II callus production from immature zygotic embryos of inbred B73 and its derivatives. Plant Cell Reports, 9: 699-702.

Tavares, A.C., M.C. Pimenta and M.T. Goncalves, 1996. Micropropagation of Melissa offi cinalis L. through proliferation of axillary shoots. Plant Cell Report, 15: 441-444.

Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant 53


Effects of the addition of clinker ash to the propagation medium on rooting of rabbiteye blueberry cuttings

T. BanA*, H. KitazawaB, S. MatsumotoA, N. KobayashiA, K. TokumasaC, M. KobatakeC and T. AsaoA

AFaculty of Life and Environmental Science, Shimane University, 2059 Kamihonjyo, Matsue, Shimane, 690-1102, Japan, BUnited Graduate School of Agricultural Science, Tottori University, Koyama-cho Minami Tottori, Tottori, 680-8553, Japan, CEnergia Economic and Technical Research Institute, The Chugoku Electric Power Co., Inc., Kagamiyama, Higashihiro-shima, Hiroshima, 739-0046, Japan. E-mail: *[email protected]

AbstractThe recommendation for a propagation medium of rabbiteye blueberry (Vaccinium ashei Reade) in Japan includes the incorporation of peat moss and Kanumatsuti (a volcanic ash deposit). This experiment compared the use of coal ash (clinker ash) and Kanumatsuti to peat moss as soil conditioner for rooting rabbiteye blueberry cutting. The numbers of cuttings survived and the root dry weight of plants propagated in clinker ash- peat moss mixes were almost the same as cuttings propagated in kanumatsuti- peat moss mix. While the quadratic model between the root dry weight and the clinker ash content in the medium was signifi cant, the maximum root dry weight was estimated to reach about 0.2 g when the proportion of clinker ash in the medium was about 40%. These fi ndings indicate that clinker ash can be used in the propagation medium of rabbiteye blueberry.

Key words: Clinker ash, cutting, propagation, rabbiteye blueberry, rooting

IntroductionCoal continues to play a vital role in electricity generation in the world. While coal-fi red electrical generation supplied 39% of the world’s electricity in 2002, this value is likely not to change over the next three decades (World Coal Institute, 2005a). After burning of coal for electric purposes, a huge amount of coal ash is generated. Coal ash mainly consists of SiO2, Al2O3 and Fe2O3 and is classifi ed into fl y and clinker ash depending on the production processes (Schlossberg et al., 2004). Clinker ash, which is coarser than fl y ash, is a dark gray, granular and porous substance (Table 1).

Coal ash production will increase continuously in the coming three decades, hence its recycling for a wide range of uses is required. Coal ash has been utilized as building material and for other constructional purpose (World Coal Institute, 2005b). Since coal ash has desirable horticultural characteristics, including low cost, good drainage and its soil amendment effect, it has been started to be utilized as a horticultural material (Schlossberg et al., 2004; Stevens and Dunn, 2004). Black and Zimmerman (2002) showed the growth and yield of highbush blueberry plants in coal ash- compost mixes were similar to or exceeded those of plants in berryland sand. Rabbiteye blueberry can be propagated by hardwood cuttings and have been propagated in many kinds of medium, including peat moss, sand, sawdust, perlite and pine bark (Austin, 1994). In New Jersey, North California and Massachusetts, a mixture of half sand and half peat moss are used to root blueberry plants (Mainland, 1966).

In this study, we investigated the effects of adding clinker ash to the propagation medium on the rooting of rabbiteye blueberry and evaluated whether it could serve as a suitable component of the propagation medium for blueberry cuttings.

Materials and methodsClinker ash (3~7 mm in diameter) was obtained from the Chugoku Electric Power Co., INC. Misumi power plant in Shimane, Japan. The components of the propagation medium were combined in the proportions listed in Table 2. Kanumatsuti, a volcanic ash deposit, was used in our studies as a substitute for the control, since a mixture of peat moss and Kanumatsuti is generally used as the propagation medium of blueberry plants in Japan (Tamada, 1997). Mature bushes of rabbiteye blueberry ‘Tifblue’ growing on the experimental farm of Shimane University were used. On 14 Feb, 2005, well-matured and healthy shoots of the former growing season were collected and stored at 4˚C. Fifty cuttings (10 cm in length) were prepared from the shoots for each medium combination on Apr, 7, 2005. Before planting the cuttings in the mediums, we confi rmed that the mean diameter of cuttings was not signifi cantly different between treatments (Table 3). The cuttings were planted in 9 cm plastic pots containing 300 mL of each medium. After adequate watering, the pots were placed in plastic propagation tunnels. Mist spray was applied from 08: 30 to 09: 00 and 16: 00 to 16: 30 daily.

The numbers of cuttings that formed callus, rooted and survived cuttings were recorded when almost all of the leaves fallen (about seven months after their planting). Root dry weight was determined by weighting roots after oven drying at 80o for 7 days. The measurements of root dry weight were replicated 5 times per treatment.

Results and discussionAlmost all the cuttings formed callus, but the number of cuttings that rooted and survived decreased with an increase in the proportion of clinker ash in the medium (Table 3). It is

Journal

Appl

well known that solid agents in plant propagation medium can regulate moisture reserves and improve aeration, since these substances have a range of porosity. The results obtained in our study indicate that drought damages was induced with an increase in the proportion of clinker ash in the medium.

The No. 3 and 4 medium resulted in the highest root dry weights and there was no signifi cant differences in root dry weight between the No. 1 (control), 3 and 4 mediums (Table 3). In medium No.

Table 1. Physical properties of clinker ashLocation of power plant Soil phase ( vol % ) Water retention

characteristics (pF=1.5, wt %)

Bulk density

Solid phase Liquid phase Air phaseMisumi in Shimane, Japan 45.2 39.3 15.5 30 0.86Shinonoda in Yamaguchi, Japan 50.7 28.2 21.1 19.1 1.06

Fig. 1. Relationship between the root dry weight and clinker ash composition in the medium. The curve was tested with a polynomial regression analysis. * means signifi cant at the 5% level.

y=-6x10- x +0.004x+0.0815 2

R = 0.448*2

Clinker ash component in medium (%)0 25 50 75 100

0.4

0.3

0.2

0.1

0R

oot d

ry w

eigh

t (g)

ReferencesAustin, M.E. 1994. Propagation. In: Rabbiteye Blueberries. AGSCIENCE,

INC., Florida, USA. p.29-33.Black, B.L. and R.H. Zimmerman, 2002. Mixtures of coal ash and

compost as substrates for highbush blueberry. J. Amer. Soc. Hort. Sci., 127: 869-877.

Mainland, C.M., 1966. Propagation and planting. In: Blueberry Culture. Eck, P. and N.F. Childers (eds.). Rutgers Univ. Press, New Brunswick, Canada. p. 111-131.

Preece, J.E. and P.E. Read, 1993. The Biology of Horticulture. John Wiley & Sons, Inc., New York, USA. p. 323-352.

Schlossberg, M.J., C.P. Vanags and W.P. Miller, 2004. Bermudagrass sod growth and metal uptake in coal combustion by-product-amended medium. J. Environ. Qual., 33: 740-748.

Stevens, G. and D. Dunn, 2004. Fly ash as a liming material for cotton. J. Environ. Qual., 33: 343-348.

Tamada, T., 1997. Guide for blueberry production [11]. Agri. Hort., 72: 92-98 (in Japanease).

World Coal Institute, 2005a. The global coal market. In: The Coal Resource. World Coal Institute, London, U.K. p.13-18.

World Coal Institute, 2005b. How is coal used ? In: The Coal Resource. World Coal Institute, London, U.K. p. 19-25.

Table 3. Effects of the addition of clinker ash to the propagation medium on the rooting of rabbiteye blueberry cuttings

Treatment Average diameter of cuttings (mm)

Number of callus formed cuttings

Number of rooted cuttings

Number of survived cuttings

Root dry weight (g)

1(Control) 8.30 NSa 50 48 48 0.128 abb

2 8.17 NS 49 49 49 0.069ab

3 8.38 NS 50 50 50 0.197a4 8.32 NS 50 46 46 0.168a5 8.48 NS 49 43 41 0.124ab6 8.26 NS 50 26 22 0.017baNon signifi cant at 5% level by the Tukey’s test. dDifferent letters within a column indicate signifi cance at 5% level by the Tukey’s test.

Table 2. Composition of medium treatments expressed as a relative percentageTreatment Component ( vol % )

Clinker ash Kanumatsuti Peat moss1(Control) - 50 502 0 - 1003 25 - 754 50 - 505 75 - 256 100 - 0

2 and 6, the root dry weight was lower than that of the other four mediums. Fig. 1 shows a signifi cant relationship between the root dry weight and the clinker ash content in the medium. While this quadratic model was signifi cant, the maximum root dry weight was estimated to reach about 0.2 g when the proportion of clinker ash in the medium was about 40%. In general, propagation medium of cuttings requires a high moisture holding ability and good drainage (Preece and Read, 1993). The low values in root dry weight from No. 2 medium may be attributed to excessive moisture level, since this medium consisted of only peat moss. In contrast, a high proportion of clinker ash in the propagation medium will inhibit the root growth of rabbiteye blueberry cuttings, because its moisture holding ability will be low.

In the present study, we clarifi ed certain effects of clinker ash when added to the propagation medium on the rooting of rabbiteye blueberry cuttings. The growths of cuttings in the clinker ash- peat moss mixes was similar to or exceeded that of cuttings in normal medium. These fi ndings indicate that clinker ash can be successfully used as a propagation medium for rabbiteye blueberry cuttings.

Effects of the addition of clinker ash to the propagation medium on rooting of rabbiteye blueberry cuttings 55


Persian walnut (Juglans regia L.) grafting as infl uenced by different bench grafting methods and scion cultivars

Babak Dehghan1, Kourosh Vahdati1*, Reza Rezaee2 and Darab Hassani3

1Department of Horticulture, College of Abouraihan, University of Tehran, Tehran, Iran, 2Agricultural and Natural Resources Research Center, West Azerbaijan, Uromia, Iran, 3Department of Horticulture, Seed and Plant Improvement Institute (SPII), Karaj, Iran. *E-mail: [email protected]

AbstractThe study was conducted to determine the effects of different grafting methods and scion cultivars on walnut grafting under controlled conditions in March 2006. Four walnut cultivars (‘Z53’, ‘Hartley’, ‘Pedro’ and ‘Serr’) were grafted using three bench grafting methods (side stub, omega and whip tongue) onto dormant two years old Persian walnut seedlings as rootstock. The plants after grafting were covered with moist sawdust with relative humidity of 85- 90% and stored in a humid room at 26-28 ºC for 21 days. Based on the results, the highest grafting success was observed with omega (84.33%) followed by side stub (41.89%) and whip tongue (24.31%) grafting, respectively. Signifi cant variations were also observed in graft take and scion growth. The differences among walnut cultivars (scion) on grafting take and scion growth were not signifi cant. However, the scions x grafting methods interaction was signifi cant and ‘Hartley’ variety grafted by omega method showed the highest graft take (88.44%) among all combinations. A signifi cant positive correlation (R2 = 0.84) was observed between the callus quality and graft takes in all grafting methods.

Key words: Callus formation, grafting techniques, graft survival, greenhouse, sawdust, walnut cultivars.

IntroductionPersian walnut (Juglans regia L.) is an important nut crop and still being propagated by seedlings in several countries including Iran, which resulted to high variability and poor crop quality (Vahdati, 2000). Selection and vegetative propagation of the superior walnut cultivars of different agroclimatic areas is the most effi cient method for increasing the production and nut quality (Solar et al., 2001). Vegetative propagation in walnut is more diffi cult in comparison with other fruit trees (Kuden and Kaska, 1997; Ozkan and Gumus, 2001) and low success has always been considered a drawback in massive propagation of superior walnut individuals (Ozkan and Gumus, 2001; Vahdati, 2003). Walnut grafting success were reported to be affected by several factors including graft technique, temperature, humidity, phenolic compounds, hormonal condition, nutrition of scion cultivars, and time of taking the scions (Mitrovic, 1995; Mehmet et al., 1997).

Environmental conditions during and following grafting have major impact on callus formation in walnut (Millikan, 1971; Rongting and Pinghai, 1993; Avanzato and Atefi , 1997; Ebrahimi et al., 2006). In Persian walnut, for successful grafting, temperature around the grafting point should be maintained at about 27 °C after grafting (Avanzato and Atefi , 1997; Germain et al., 1997).

In walnut vegetative propagation, fl uctuating temperatures and lack of suffi cient humidity cause undesirable environment under fi eld condition responisble for poor callus formation and grafting failure (Ebrahimi et al., 2006), thus bench grafting methods are being usually preferred. The advantages of these techniques include:1) table grafting could be done under controlled conditions and gives a better result, 2) the operation could be done during winter and 3) grafting could be mechanized to increase labour productivity (Lantos, 1990; Tsurkan, 1990; Unal, 1995).

New methods using bench grafting techniques like hot cable or hot callus have been developed and used by different researchers (Avanzato and Atefi , 1997; Hartmann et al., 2001), but requirement of more skillful workers and expensive facilities restrict their application in most nurseries. Therefore, the main goal of the present study was to evaluate the effi ciency of modifi ed bench grafting methods in order to decrease the high cost of production by grafting in walnut as well as to compare different walnut cultivars in terms of grafting success and scion growth.

Materials and methodsThe experiment was conducted in Department of Horticulture, College of Abouraihan, University of Tehran during March 2006, to evaluate effect of factorial combination of three grafting methods (side stub, omega and whip tongue) and four walnut cultivars (‘Z53’, ‘Hartley’, ‘Pedro’ and ‘Serr’) on callus formation, grafting success and scion growth. The scions were stored at cool and moist condition, until they were used in grafting. The seedling rootstocks were taken in late January and selected for size and uniformity. The study was conducted using a factorial experiment on a complete randomized design with 12 treatments in three replications and 15 seedlings per plot.

All of the grafted combinations were carried out by the same person and standard methods were used for grafting as described by Hartmann et al. (2001) briefl y as following. For side-stub grafting, the basal part of the scions were cut as wedge of about 2.5 cm and an oblique cut at angle of 20º to 30º was made into the basal part of the seedling’s stem then scion were inserted, without any fastening material. In whip grafting, a sloping cut (2-3 cm in length) was made at the top of the rootstock, and then a second downward cut was made starting one-third of the distance from the tip to the base of the fi rst cut. Similarly, a sloping cut

Journal

Appl

was made at the base of the scions to the same length as the cut on the rootstock. Also a second cut was made under the fi rst cut similar to the stock. Finally, the graft compounds were slipped together to be interlocked, and were tied by rubber tape tightly without waxing. The omega grafting method was performed by omega grafting machine (OMEGA-STAR Company of Germany) without any fastening similar to the side stub grafting method.

Grafted plants were maintained in a greenhouse at 26-28 ºC inside the sawdust with the moisture content of 85-90% for 21 days. After completing callusing period, data were recorded for amount of callus formation based on a visual scale of one to four in which 1= poor, 2= medium, 3= high and 4= very high callusing. Subsequently, the percentage of callus formation, graft take, and scions growth were measured in each treatment. The grafted plants then were transferred to the black polyethylene pots (20 × 30 cm) containing sand: soil: leaves compost mixture (1:2:1 w/w), and were put in a humid greenhouse for about three months (Fig. 1). They were brought to the shade house in September, 2007, before transplanting to the fi eld condition.

The collected data were analyzed by SAS statistical software (SAS Institute, North Carolina, USA) and means compared using Duncan’s multiple range test (DMRT) method. Statistical signifi cance indicates means difference (P<0.05).

Results and discussionThe results showed that different grafting methods had signifi cantly different effects on callus quality, graft take and grafting survival (Table 1). Among the methods, omega grafting showed the highest callus quality (2.5 from 4), higher graft take (67.77%) and grafting survival (84.33 %) as well as higher scion growth (12.9 cm) followed by side stub and whip grafting. Percentage of graft takes in side stub and whip tongue methods were 58.88 and 19.44%, respectively. However, percentages of grafting survival were about 41.89% in side stub and 24.31% in whip tongue methods (Table 1).

As demonstrated in Fig. 2, percentage of graft take shows a highly positive correlation with callus quality at different grafting methods and cultivars. This is in agreement with the results of Rongting and Pinghai (1993) that callus quality and amount of callus formation plays an important role in the grafting success.

In this experiment, the percent of grafting success in the most

effi cient treatment was similar to Lantos (1990) who obtained 80% of grafting success using bench grafting method. Among three methods of grafting, omega grafting showed the best result considering all of studied characters which is consistent with Solar et al. (2001).

Signifi cant differences was observed among the cultivars in terms of callus quality and graft survival, but the effect of variety on grafting take or scion growth was not statistically signifi cant (Table 2). This is in contrary with the result of Rongting and Pinghai (1993) and Mitrovic et al. (1997) who reported the

Table 1. Effects of grafting methods on callus quality, graft take, survival and scion growthGrafting type Callus

qualityAGraft takes

(%)Graft

survival (%)Scion

growth (cm)Omega 2.53aB 67.77a 84.33a 12.9aSide stub 2.28a 58.88a 41.89b 5.38bWhip tongue 0.97b 19.44b 24.31b 3.06bAValues are means of callus scoring rating from 1 (low callus) to 4 (very good callus). BMeans with different letters in each column are signifi cantly different at P<0.05.

signifi cant effect of walnut cultivars on graft take. Although scion cultivar could affect grafting result, but it seems that it is mainly affected by scion quality which is a management related issue rather than their genetic makeup (Rezaee and Vahdati, 2008).

The interaction of grafting methods and variety on graft take, grafting survival and scion growth were signifi cant (Table 3). The highest graft take (88.44%) was achieved by omega grafting of ‘Hartley’ and the lowest graft take by whip tongue grafting of ‘Z53’ and ‘Hartley’ cultivars with 0.00 and 13.33 %, respectively. The highest scion growth (14.66 cm) was in ‘ Z53’ variety with omega grafting method and lowest scion growth (2.22 cm) was obtained from ‘Serr’ variety with side stub grafting method. The highest graft survival (96.67 and 90.91%) was obtained by grafting of ‘Z53’ and ‘Pedro’ cultivars using omega and side stub grafting methods, having the highest score of callus quality of 2.76 and 2.75, respectively

Table 2. Callus quality, graft take, survival and scion growth in different walnut cultivarsCultivar Callus qualityA Graft takes

(%)Graft survival

(%)Scion growth

(cm)‘Serr’ 2.24aB 45.92a 48.61ab 5.82a‘Pedro’ 2.08a 54.07a 73.27a 8.16a‘Hartley’ 2.01a 52.59a 42.22ab 8.57a‘Z53’ 1.37b 42.22a 36.67b 5.90aAValues are means of callus scoring ratings from 1 (low callus) to 4 (very good callus). BMeans with different letters in each column are signifi cantly different at P <0.05.

Fig. 2. Correlation between callus quality and graft take in different grafting methods.

Fig. 1: A view of grafted plants after two months of grafting and transferring to pots. A) Whip tongue grafting method, B) Omega grafting, and C) Side stub grafting method.

Persian walnut grafting as infl uenced by different bench grafting methods and scion cultivars 57

The results of this study were different than Ozcan and Gumus (2001) who found the highest success by using whip tongue grafting method. It could be probably related to the experimental condition; so that tight binding of the grafting place by rubber tape in our experiment, resulted to limitation in aeration and blockage of the phloem sap transportation under the graft cover. Adequate aeration and auxins have an important roles on callus formation and grafting success (Ronging and Pinghai, 1993; Vahdati, 2000; Hartmann et al., 2001; Rezaee and Vahdati, 2008).

Interlocking and quality of callus formation between rootstock and scion in the omega and side stub grafting methods were approximately similar. Despite a good interlocking of grafting compound, as shown in Fig. 3, formation of callus bridge was restricted in whip tongue grafting method.

Considering the diffi cult-to-graft nature of walnuts and higher cost of specialized facilities used in methods like hot cable and hot callus (Avanzato and Atefi , 1997; Avanzato, 2001) the results of this experiment are comparably promising, and represents a good alternative method to propagate walnut cultivars under partially controlled condition. Further research is needed for improvement of grafting techniques and condition to obtain more uniform grafting success.

AcknowledgmentWe thank Dr. Reza Amiri for his assistance in analyzing the data. University of Tehran and Iran National Science Foundation (INSF) are also acknowledged for their fi nancial supports.

ReferencesAvanzato, D. 2001. Effect of different hygro-thermic environments on

growth of potted walnut grafted seedlings. Acta Hort., 544: 459-464.Avanzato, D. and J. Atefi , 1997. Walnut grafting by heating the graft

point directly in the fi eld. Acta Hort., 442: 291-294.Ebrahimi, A., K. Vahdati and E. Fallahi, 2006. Improved success of

Persian walnut grafting under environmentally controlled conditions. Int. J. Fruit Sci., 6: 3-12.

Germain, E., F. delort and V. Kanivets, 1997. Precocious maturing walnut population originating from central Asia: their behaviour in France. Acta Hort., 442: 83-90.

Hartmann, H.T., D.E. Kester, F.T. Davies and R. Geneve, 2001. Plant Propagation: Principles and Practices. 7th ed. Prentice Hall, New Jersey.

Kuden, A. and N. Kaska, 1997. Studies on the patch budding of walnuts in different budding periods under subtropical conditions. Acta Hort., 442: 299-301.

Lantos, A. 1990. Bench grafting of walnut. Acta Hort., 284: 53-56.

Mehmet, S., T. Karadanize, F. Balta and E.Tekintas, 1997. Changing of fl avan contents at some organs of walnut seedling (Juglans regia L.) exposed to the controlled grafting conditions. Acta Hort., 442: 181-184.

Millikan, D.F. 1971. ‘Propagation of Juglans species by fall grafting.’ Annual Report: Northern Nut Growers. Association, 61: 41-44.

Mitrovic, M. 1995. Effect of the cutting date of walnut scion wood on the take and callusing of grafts. Jugoslovensko Vocarstvo, 29: 59-63.

Ozkan, Y. and A. Gumus, 2001. Effects of different applications on grafting under controlled conditions of walnut. Acta Hort., 544: 515-520.

Rezaee, R. and K. Vahdati, 2008. Introducing of a simple and effective procedure for topworking Persian walnut tree. J. Amer. Pom. Sci., 62: 21-26.

Rongting, X. and D. Pinghai, 1993. A study on the uniting process of walnut grafting and the factors affecting. Acta Hort., 311: 160-170.

Solar, A., F. Stampar, M. Trost, J. Barbo and S. Avsec, 2001. Comparison of different propagation methods in walnut (Juglans regia L.) made in Slovenia. Acta Hort., 544: 527-530.

Tsurkan, L.P. 1990. Production technology of English walnut planting, materializing winter table grafting. Acta Hort., 286: 65-68.

Unal, A. 1995. Studies on the effects of different practices on grafting success in bark, whip and top-cleft grafting in walnuts. E.U.Z.F. Dergisi Izmir / Turkey 32: 85-90.

Vahdati, K. 2000. Walnut situation in Iran. Nusis Newsletter, 9: 32-33.Vahdati, K. 2003. Nursery Management and Grafting of Walnut. 2th ed.

Khaniran Publ,Tehran.

Fig. 3. Interlocking and quality of callus bridge formation between rootstock and scion in the studied grafting methods. A, D) Whip tongue method, B, E) Side stub method, C, F) Omega method

Table 3. Interaction of grafting methods × cultivar on callus quality, graft take, survival and scion growthGrafting × variety Callus

qualityAGraft take

(%)Graft

survival (%)

Scion growth (cm)

Omega × ‘Z53’ 2.76aB 64.44ab 96.67a 14.66aOmega × ‘Hartley’ 2.71a 88.44a 80abc 14.34aOmega × ‘Pedro’ 2.07ab 57.78bc 81.67abc 11.07abOmega × ‘Serr’ 2.57a 64.44ab 79.00abc 10.91abcSide stub × ‘Z53’ 1.36cb 13.33d 11.11c 3.067bcdSide stub × ‘Hartley’ 2.33a 60.00ab 30.00cd 7.44abcdSide stub × ‘Pedro’ 2.75a 68.89ab 90.91ab 8.82abcSide stub × ‘Serr’ 2.71a 44.44bcd 33.33bcd 2.22abcWhip tongue × ‘Z53’ 0.00d 0.00f 0.00d 0.00dWhiptongue ×‘Hartley’ 1.00c 13.33ef 16.67d 3.93bcdWhip tongue × ‘Pedro’ 1.43bc 35.56cb 41.22abcd 3.98bcdWhip tongue × ‘Serr’ 1.45bc 28.89ed 33.33bcd 4.33bcdAValues are the means of callus scoring ratings from 1 (low callus) to 4 (very good callus). BMeans with different letters in each column are signifi cantly different at P<0.05.

58 Persian walnut grafting as infl uenced by different bench grafting methods and scion cultivars


Extraction and determination of α-solanine in eggplant fruits

Zhiwen Li1, Baoli Zhou1*, Yuwen Ding1 and Xiang Liu2 1College of Horticulture, Shenyang Agricultural University, Shenyang 110161, China; 2College of Agriculture and Bioengineering, Tianjin University, Tianjin 300072, China. *E-mail: [email protected]

AbstractA simple and effective high-performance liquid chromatographic (HPLC) method for determination of α-solanine in eggplant fruits is described in our study. A new extraction method is established for extracting α-solanine in eggplant fruits. Single and orthogonal tests were designed to analyze the effect of different extraction methods and ultrasonic wave extraction condition on extraction of α-solanine in eggplant fruits. HPLC separation was achieved on a Waters Nova-pak C18 column with the mobile phase acetonitrile-0.05N potassium dihydrogen phosphate (55:45, V/V). The fl ow rate was 0.7mL min-1 and the UV absorbance was monitored at 202 nm. The optimal extraction method was ultrasonic wave extraction in 70% methanol for 60 minutes at 50oC, and with material to liquid ratio of 1:10. Under the optimal extraction conditions, the average content of α-solanine in skins and fl esh of dried eggplant fruits was 0.107±0.006 and 0.626±0.004mg g-1, respectively. The average recovery effi ciency was 97.97%.

Key words: Eggplant, α-solanine, HPLC, extraction, ultrasonic wave

IntroductionEggplants (Solanum melomgena L.) produce natural active substances called glycoalkaloids with biological activity. Glycoalkaloids, like many secondary metabolites, are thought to function in plant chemical defense, acting as nonspecifi c protectors or repellents against potential pest predators (Osman, 1980). Eggplants contain a type of glycoalkaloids like other nightshade species called α-solanine (Manuchair, 2006), which is an important active substance with many biological activities including anti-tumor (Tai, 2002, Ji et al., 2005), lowering blood pressure and heart stimulation (Yang, 2004). In addition, α-solanine has been used in the treatment of asthma and epilepsy (Zhang et al., 2002), and also has shown the effects of antifeedant (Beier, 1990), fungicide (Allen and Kúc, 1968), and pesticide (Birch et al., 2002).

Consideration of the above demonstrates the importance of accurate and reliable analytical methods for α-solanine. Literatures suggest that α-solanine could be extracted with methanol (Kobayashi et al., 1989; Saito et al., 1990), methanol-chloroform (2:l v/v) (Bushway and Ponnampalam, 1981), heptanesulfonic and acetic acids (Carman et al., 1986), 1.1% acetic acid (Filadelfi and Zitnak, 1982), and tetrahydrofuran (THF)-water-acetonitrile with 1% acetic acid (Bushway et al., 1980a, 1983, 1986). However, the common soaking extraction procedure usually requires long time, high temperature and more energy. As a novel technique for sample pretreatment, ultrasound-assisted extraction has attracted more attention in recent years. Compared with other leaching techniques such as Soxhlet extraction, microwave-assisted extraction and supercritical fl uid extraction, ultrasound-assisted extraction shows many potential advantages in various aspects, such as faster extracting rate, matter quality reservation and lower time and energy cost. It has been used in extracting anthraquinones from Rheum palmatum L, and cyanuric acid from pet food (Chen et al., 2008; Wang et al., 2008)

Many analytical methods for α-solanine are reported in literature,

these methods include isotachyphoresis (Kvasnicka et al., 1994), thin layer chromatographic scanning (Ferreira et al., 1993), various colorimetric methods, gas chromatography (Lawson et al., 1992), countercurrent chromatography (Fukuhara and Kubo, 1991), and high-pressure liquid chromatography (HPLC) (Everard et al., 1996; Hitoshi et al., 2005), and enzyme immunoassays (Plhak and Sporns, 1992). Each method has advantages and disadvantages. For example, the colorimetry and thin layer chromatographic scanning are not proper to quantitative analysis of glycoalkaloids because of low sensitivity and recovery; the gas chromatography and enzyme immunoassays treatment has complicated process, bad colour stability, low sensitivity. Compared to other methods, HPLC is a sensitive, simple, rapid, and relatively cheap detection method.

The objective of this study was to establish a concise, reproducible method for the routine extraction and quantifi cation of α-solanine in eggplants.

Materials and methodsMaterials: The seeds of eggplant ‘Liaoqie-7’ (purple eggplant) were bought from Fuyou Seed Company in Liaoning Province, China. The seeds were sown in soil and transplanted on 60th days in the greenhouse in Shenyang Agricultural University. Eggplant fruit samples were harvested at mature stage. In order to separate skin from fl esh, a knife was used to peel fruits. Then scrape the inside of skin to remove the left fl esh. So the content of α-solanine in skin didn’t contain the fl esh. And the content of α-solanine in fl esh didn’t contain the skin. The skin and fl esh were porphyrized to powder after drying in the air. The rate of dried weight to fresh weight in samples was 1:10. The powder was then passed through a 40-mesh screen, and stored in the extractor at 4oC.

The α-solanine stock solution (3 mg mL-1) was prepared through dissolving α-solanine (Sigma-Aldrich chemical Co., USA) in methanol. All solutions were prepared with ultrapure water (Barnstead Diamind, USA). All chemicals were AnalaR grade except for special notifi cations.

Journal

Appl

Extraction procedure: The extraction technique was adopted from Bushway et al. (1980b) with some modifi cations. 25g of dried powder was stirred with 250 mL of 70% methanol and 1mg mL-1 sodium bisulfi tein in a mixer (Corning PC-420D, USA) for 15 min. The mixture was treated by different treatments which were ultrasonic wave extraction, microwave extraction, soxhlet extraction, oscillation extraction and soaking extraction at 50o for 60 min, and then vacuum fi ltrated. The fi lter was collected and concentrated to approximately 10-15 mL by rotary evaporator (Eyela N-1000, Japan). The concentrate was then added with 10-15 mL of 0.2 NHCl while being stirred. The extraction was centrifuged at 8000 rpm at 20o for 15 min (HERMLE Z 36 HK, German), and the pH of the suspension was adjusted to 10-11 (Horiba F-53, Japan) with concentrated ammonium hydroxide before the sample was placed into a water bath at 70o for 30min. The solution was then cooled in an ice-water bath for 30 min and centrifuged at 10000 rpm at 20o for 30 min. The precipitate was washed with 5 mL of 1% ammonium hydroxide and centrifuged at 10000 rpm at 20o for 15 min, and the pellet was collected. The air-dried pellet was then dissolved in 10 mL of methanol (HPLC grade, Merck, German) and treated in the ultrasonic treatment unit at 50o for 30min. The suspension was fi ltered through a 0.45 μm fi lter membrane. The fi ltrate was used for HPLC. In this experiment we adopted different factors which were three solvent (70% Methanol, 5% Acetic acid and Methanol: chloroform = 2:1), three extraction time (40, 50 and 60 min), three temperature (30, 40 and 50oC) and three material to liquid ratio (1:8, 1:9 and 1:10) to fi nd the best condition of extracting α-solanine from eggplant fruit with ultrasonic wave by orthogonal testa.

HPLC Procedure. The HPLC procedure was adopted from Friedman and Dao (1992) with some modifi cations. Spectra of α-solanine in the adopted mobile phase showed maxima absorbance at 202 nm. Therefore, 202 nm was used for quantifi cation of α- solanine. A Waters 600 controller HPLC system with a 600 pump and degasser, an automatic injector with a fi nal volume loop of 10 μL, a Nova-pak Cl8 coupled cartridges column (250×4.6mm, 5μm) and a Waters 2487 UV-vis variable-wavelength detector was used. System management and hardware interface for data acquisition were performed by the Millennium32 computer software package from Waters. The mobile phase was acetonitrile (HPLC grade, Merck, German)-0.05 N potassium dihydrogen phosphates (55:45 v/v) adjusted to pH 4.5 with 1% phosphoric acid. This was passed through a 0.45 μm fi lter and degassed for 20 min under reduced pressure. The HPLC fl ow rate was 0.7 mL min-1, and the UV absorbance was measured at 202 nm.

Spiking Experiments: A series of spiking experiments were carried out to establish the extent of recovery of added α-solanine from eggplants. Specifi cally, 20 g dried eggplant fruit powder was added in different experiments with 1, 2, 3, 4 and 5 mg α-solanine, respectively. The samples were thoroughly mixed, extracted, and analyzed by HPLC for recovery of the added alkaloids.

TLC Procedure: Preparative thin-layer chromatography (TLC) was performed on silica gel precoated plates, 0.25 mm×20cm×20 cm (Merck, Darmstadt, Germany). A 25-μL methanol concentrated extract was spotted on the plate along with α-solanine standards. Then the plate was developed with a bottom layer of benzene-methanol-1 % ammonium hydroxide (10:2 v/v). After that, the dried plate was sprayed with 1% bismuth potassium iodide with visible violet red spot.

ResultsThe methanol concentrated extracts were spotted on a TLC plate along with α-solanine standards. And the skin and fl esh extracts were analyzed. After development, all plates showed only one spot corresponding to α-solanine (R = 0.35).

Fig. 1 shows linear relationships in the range 0.15-3 mg mL-1 between concentrations of α-solanine (r = 0.995) and the peak height on HPLC chromatograms. Fig. 2 illustrates the separation of α-solanine on the HPLC columns in standard solution, extract and extract added with standard of eggplant fruit, with the retention time of 2.8 min. These data suggested that minimal error was related to HPLC.

The highest extracting rates of α-solanine in both skin and fl esh were obtained using ultrasonic wave (Table 1), with virtually 0.107 and 0.626 mg g-1, whereas common extraction gave only 0.0114 and 0.192 mg g-1, respectively. None of the other extraction means used gave extracting rates as high as those obtained with the ultrasonic wave extraction.

The effect of extracting α-solanine from eggplant fruit with ultrasonic wave was dependent on solvent, extraction time, extraction temperature and material to liquid ratio. For both skin and fl esh, the effect of these four factors on extracting rate was solvent (A) > material to liquid ratio (D) > time (B) > temperature (C) (Table 2). The best treatment was A1B3C3D3, with extraction

0

100

200

300

400

500

600

700

0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3 3.3Concentration ( mg mL )-1

Pea

k he

ight

resp

onse

(mV

)

Fig. 1. Standard α-solanine peak height response with HPLC

Table 1. Extracting rates of different extraction means on skin and fl esh of eggplanta

Extraction methodα-solanine extracting rate (mg g-1)

Skin FleshUltrasonic wave extraction 0.107±0.006 a A b 0.626±0.004 a A

Microwave extraction 0.0764±0.002 b B 0.432±0.003 b B

Soxhlet extraction 0.0422±0.0004 c C 0.358±0.006 c C

Oscillation extraction 0.0405±0.001 c C 0.343±0.007 d D

Common Extraction 0.0114±0.001 d D 0.192±0.003 e Ea Results values are means ± standard errors (n=3)Different lowercase letters in the same column indicate signifi cant difference at P=0.05 and different capital letters indicate signifi cant difference at level of P=0.01

60 Extraction and determination of α-solanine in eggplant fruits

rates 0.107 and 0.626 mg g-1, extracted at 50o for 60 minutes in 70% methanol with the material to liquid 1:10.

Spiking experiments (Table 3) revealed that the modifi ed method could recover a 96-99% of added α-solanine. In the range of 1- 5 mg of α- solanine / 20 g dried eggplant fruit powder (fl esh and skin), the extent of recovery increased with the amount added before extraction. The average recovery was 97.97%.

DiscussionSince eggplant fruits contain approximately 93-94% water, the air-dried samples provide more than 10 times concentration of α-solanine. Air-dried samples can be easily grinded and stored for longer periods of time prior to measurements. Considering these advantages, air-dried fruits were used in this study.

The recovery of added and native α-solanine in skin and fl esh of eggplant fruts increased by approximately 2% in the condition of

without washing the precipitate with 1% ammonium hydroxide. Washing aimed to remove most brown pigments in the fi nal extraction product when eggplants fl esh was used as starting materials. The brown pigment does not seem to interfere in the analysis except for turning into dark. To minimize possible interference of pigment in skins, the purple precipitate formed during evaporation of the extracts was first removed by centrifugation and fi ltration. The partially evaporated clear fi ltrate was then analyzed for alkaloid content.

Separation and retention of the α-solanine increased by raising the buffer pH. However, this resulted in reduced sensitivity of the system due to low solubility of α-solanine. No alkaloid was detected above pH 7 because of precipitation upon injection onto the column. The optimum pH was found to be 4-5. Reducing the amount of acetonitrile in the mobile phase also improved separation. Using 55% acetonitrile, 45% buffer gave maximum separation. α-solanine was insoluble at low acetonitrile ratios.

mV

0

100

200

300

400

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Minutes

mV

0

100

200

300

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4. 0 4.5 5.0 5.5 6.0Minutes

0

100

200

300

400

500

600

Minutes0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

a b c

1

1

1

Pea

k he

ight

(mV

)

Retention time (min)Fig. 2. Chromatograms of α-solanine standard (a), α-solanine in eggplant fruit (b) and α-solanine in mixture of eggplant fruit extracting solution added with α-solanine standard (c). Peak: 1, α-solanine

Table 2. Results of orthogonal testa of α-solanine extracted from eggplant skin and fl esh with ultrasonic waveb

Treatment ASolvent

BTime(min)

CTemperature

(°C)

DMaterial to liquid

(g:mL)

Extracting rate in skins (mg g-1)

Extracting rate in fl esh (mg g-1)

1 A1 70% Methanol B1 (40) C1 (30) D1 (1:8) 0.0867±0.004 0.524±0.002

2 A1 B2 (50) C2 (40) D2 (1:9) 0.0958±0.003 0.554±0.003

3 A1 B3 (60) C3 (50) D3 (1:10) 0.1070±0.006 0.626±0.004

4 A2 5% Acetic acid B1 C2 D3 0.0776±0.003 0.386±0.006

5 A2 B2 C3 D1 0.0535±0.0003 0.147±0.004

6 A2 B3 C1 D2 0.0547±0.0003 0.158±0.005

7A3 Methanol:chloroform 2:1

B1 C3 D2 0.0395±0.001 0.139±0.007

8 A3 B2 C1 D3 0.0458±0.001 0.151±0.006

9 A3 B3 C2 D1 0.0220±0.001 0.0635±0.004

Rα 0.4542 0.0708 0.0528 0.1390 — —

Rα 0.0608 0.0067 0.0044 0.0228 — —aL9(3)4 design was adopted in this orthogonal test bResults are the mean extracting rate of standard ± standard errors, n=3

Extraction and determination of α-solanine in eggplant fruits 61

The composition of the mobile phase used for HPLC was important in ensuring full separation of α-solanine. A number of published methods were conducted with Tris-HCl at near-neutral pH (Jonker et al., 1992). However, under such conditions the HPLC stability was not good and the linear relationships did not perform well.

Many of extraction solvents were utilized in the published methods and most were based on some solvents or salts in a weak solution of acetic acid. Heptanesulfonic acid and aqueous acetic acid (Everard et al., 1996) or 5% acetic acid (Hitoshi et al., 2005) were commonly added. However, as α-solanine is probably stored within the aqueous phase of the eggplant cell and is readily soluble in solvents like methanol with the addition of water through some special extraction means like ultrasonic wave extraction, the use of acid is unnecessary. In addition, the solution of acetic acid under ultrasonic wave condition may lead to hydrolyzation of α-solanine resulting in a reduced recovery. In this experiment, the extracting rate of 70% methanol as solvent was higher than that of 5% acetic acid. Sodium bisulfi te was used to reduce oxidation of the extract (Hellenas, 1986), and also used in these investigations.

Ultrasound is an effi cient non-thermal alternative (Kim und Zayas, 1989; Yan et al., 2002; Knorr et al., 2002). Ultrasonic cavitation creates shear forces that break cell walls mechanically and increase material transfer. In this experiment, ultrasonic wave extraction with methanol as extraction solvent gave the highest extracting rate and optimal recovery.

Some researches showed that during ultrasonic extraction, the relationships between extraction time and extracting rate are as follows (Guo, 1995; Guo, 1997): (1) The extracting rate increased as time prolongs; (2) The extracting rate increasment slowed as time prolonging to certain extent; (3) The extracting rate decreased as time prolonging to critical stage. The reason for extracting rate decrease is probably due to the fact that prolonged time may decompose the active components and increase their impurity, leading to low extracting rate (Duan and Feng, 1992). In a preparatory experiment, the extracting rate increased rapidly within 60 minutes extraction, slowly in 60-90 minutes extraction, and decreased after 90 minutes.

The above method aim at a number of problems with HPLC analysis of α-solanine, and utilizes a new extracting way to improve sample recoveries and got good results. It can be used in eggplant breeding, production and medical care research. Laboratory also can use this method to investigate the response of eggplant alkaloid concentrations to environmental conditions during cultivation and some biological activity of eggplant.

ReferencesAllen, E.H. and J. Kúc, 1968. α-solanine and α-chaconine as fungitoxic

compounds in extracts of Irish potato tubers. J. Phytopathol., 58: 776-781.

Beier, R.C. 1990. Natural pesticides and bioactive components in foods. Rev. Environ. Contam. Toxicol., 113: 47-137.

Birch, N.E., I.E. Geoghegan, D.W. Griffi ths and J.W. Mcnicol, 2002. The effect of genetic transformations for pest resistance on foliar solanidine-based glycoalkaloids of potato (Solatium tuberosuni). J. Annals Appl. Biol., 14: 143-149.

Bushway, R.J. and R. Ponnampalam, 1981. α-chaconine and α-solanine content of potato products and their stability during several modes of cooking. J. Agr. Food Chem., 29: 814-817.

Bushway, R.J., E.S. Barden, A.W. Bushway and A.A. Bushway, 1980a. The mass extraction of potato glycoalkaloids from blossoms. Am. Potato J., 157: 175-180.

Bushway, R.J., E.S. Barden, A.M. Wilson and A.A. Bushway, 1980b. Analysis of potato glycoalkaloids by high-performance liquid chromatography. J. Food Sci., 45: 1088-1089.

Bushway, R.J., J.L. Bureau and J. King, 1986. Modifi cation of the rapid high-performance liquid chromatographic method for the determination of potato glycoalkaloids. J. Agr. Food Chem., 34: 277-279.

Bushway, R.J., J.L. Bureau and D.F. McGann, 1983. α-Chaconine and α-solanine content of potato peels and potato peel products. J. Food Sci., 48: 84-86.

Carman, A.S., S.S. Kuan, G.M. Ware, O.J.J. Francis and G.P. Kirschenheuter, 1986. Rapid HPLC determination of the potato glycoalkaloids α-solanine and α-chaconine. J. Agr. Food Chem., 34: 279-282.

Chen Y., L. Zhua, J. Xiaoa, H. Tanga, G. Guob, Q. Zengb and X.Wang, 2009. Ultrasonic extraction and determination of cyanuric acid in pet food. Fd. Control, 20(3): 205-208.

Duan, G.M. and C.Q. Feng, 1992. Glycoalkaloids in potatoes. J. Plant Physiol. Commun., 28: 457-461.

Everard, J., Edwards and H.Cobb. Andrew, 1996. Improved high-performance liquid chromatographic method for the analysis of potato (Solanum tuberosum) glycoalkaloids. J. Agr. Food Chem., 44: 2705-2709.

Ferreira, F., P. Moyna, S. Soule and A. Vazquez, 1993. Rapid determination of solanum glycoalkaloids by thin-layer chromatographic scanning. J. Chromatogr., 653(2): 380-384.

Filadelfi M.A. and A. Zitnak, 1982. Preparation of chaconines by enzymic hydrolysis of potato berry alkaloids. J. Phytochemistry, 21: 250-251.

Friedman, M. and L. Dao, 1992. Distribution of glycoalkaloids in potato plants and commercial potato products. J. Agric. Food Chem., 40: 419-423.

Fukuhara, K. and I. Kubo, 1991. Isolation of steroidal glycoalkaloids from Solanum incanum by two countercurrent chromatographic methods. J. Phytochemistry, 30: 685-687.

Guo, X.W. 1995. Effect on extraction yield of berberine by ultrasound-assisted extraction. J. Chin. Materia, 20: 673-675.

Guo, X.W., 1997. Comparison of effect on rutin component by ultrasound-assisted and hot alkali extraction. J. Chin. Traditional Herbal Drugs, 28: 88-89.

Hellenas, K. 1986. A simplifi ed procedure for the quantifi cation of potato glycoalkaloids in tuber extracts by HPLC, comparison with ELISA and a colorimetric method. J. Sci. Food Agr., 37: 776-782.

Hitoshi, K., S. Keiitsu, N. Nobumitsu, Y. Shigeo and T. Youichi, 2005. Simple and sensitive method for determination of glycoalkaloids in potato tubers by high-performance liquid chromatography with chemiluminescence detection. J. Chromatogr. A, 1100(1): 26-31.

Table 3. Results of recoveries of α-solanine added to dried samplea

Sample (mg) Added (mg) Detected (mg) Recoveryb (%)15.09±0.05 1 15.90±0.05 98.82±1.114.27±0.05 2 16.18±0.05 99.45±0.914.25±0.05 3 16.70±0.05 96.81±1.314.65±0.05 4 18.24±0.05 97.80±0.814.89±0.05 5 19.29±0.05 96.98±1.2Mean±SD 97.97±1.1

a20 g powder was added the known amount of α-solanine. The samples were mixed, extracted, and analyzed with HPLC.bResults are the mean recoveries of standard ± standard errors, n=3

62 Extraction and determination of α-solanine in eggplant fruits

Ji, Y.B., H.L. Wang and S.Y. Gao, 2005. Effect of solanine on DNA and RNA in tumor cell of tumor-bearing mice. J. Chin. Traditional Herbal Drugs, 36: 1200.

Jonker, H.H., A.J. Koops and J.C. Hoogendoorn, 1992. A rapid method for the quantifi cation of steroidal glycoalkaloids by reversed phase HPLC. J. Potato Res., 35: 451-455.

Kim, S.M. and J.F. Zayas, 1989. Processing parameter of chymosin extraction by ultrasound. J. Food Sci., 54: 700.

Knorr, D., B.I.O. Ade-Omowaye and V. Heinz, 2002. Nutritional improvement of plant foods by non-thermal processing. J. Proceedings Nutr. Soc., 61: 311-318.

Kobayashi, K., A.D. Powell, M. Toyoda and Y. Saito, 1989. HPLC method for the simultaneous analysis of α-solanine and α-chaconine in potato plants cultured in vitro. J. Chromatogr., 462: 357-364.

Kvasnicka, F., K.R. Price, K. Ng and G.R. Fenwick, 1994. Determination of potato glycoalkaloids using isotachophoresis and comparison with a HPLC method. J. Liquid Chromatogr. Related Technol., 17: 1941-1951.

Lawson, D.R., W.A. Erb and A.R. Millar, 1992. Analysis of Solanum alkaloids using internal standardization and capillary gas chromatography. J. Agric. Food Chem., 40: 2186-2191.

Manuchair, S.E., 2006. Pharmacodynamic basis of herbal medicine, US: CRC Press, 309.

Osman, S.F., 1980. Glycoalkaloids of the Solanaceae. Recent Adv. Phytochem., 14: 75-96.

Plhak, L.C. and P. Sporns, 1992. Enzyme immunoassay for potato glycoalkaloids. J. Agric. Food Chem., 40: 2533-2540.

Saito, K., M. Horie, Y. Hoshino, N. Nose and H. Nakazawa, 1990. HPLC determination of glycoalkaloids in potato products. J. Chromatogr., 508: 141-147.

Tai, J.R. 2002. The effect of cancer prevention and resistance of main vegetables. J. Southwest Hort., 30: 28-29.

Wang, L., D. Li, C. Bao, Z. Wang, Y. Shi and H. Zhang, 2008. Ultrasonic extraction and separation of anthraquinones from Rheum palmatum L. Ultrason. Sonochem., 15: 738-746.

Yan, W., S.F. Li and S.J. Tian, 2002. Ultrasound-assisted extraction technology. J. Chem. Ind. Eng. Progress, 21: 649-651.

Yang, X.W. 2004. Alkaloids. Beijing, Chemical Industry Press, 393.Zhang, C.Y., S. Li, X. Chen and C.M. Wang, 2002. Studies on the anti-

asthma action mechanism of nightshade alkaloids. J. Theory and Practice of Chinese Medicine, 2002: 488-489.

Extraction and determination of α-solanine in eggplant fruits 63


Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress and normal conditions

M. Altaf Wani1, G.R. Lawania2, R.A. Bhat3, Iffi at Fayaz1, A. Nanda3 and Gazenfar Gani3

1Division of Plant Breeding and Genetics, SK University of Agricultural Sciences and Technology of Kashmir, Srinagar, India.2Division of Plant Breeding and Genetics, Allahabad Agricultural Institute-Deemed University, Allahabad, India.3Division of Floriculture, SK University of Agricultural Sciences and Technology of Kashmir, Srinagar, India.

AbstractThree different strains of Hypericum perforatum viz. HP-1, HP-2, and HP-3 were subjected to different levels of saline stress (0.25, 0.5, 0.75 and 1.0% with NaCl) and high pH regime (8.5, 9.0, 9.5 and 10.0 with NaOH). Gradual loss in callus growth was observed in all the three strains in response to both kinds of stress. However, high pH showed more drastic effect than saline stress. All the three strains showed higher content of pseudohypercin than hypercin. Change in hypercin production was negligible, however remarkable change was observed in pseudohypercin production in response to both kinds of stress. HP-2 strain produced higher content of hypercin than HP-1 and HP-3 strains under normal as well as under stressfull regime. Proteins were affected qualitatively as well as quantitatively. Maximum numbers of proteins were isolated from control cultures at the retention time of fi ve minutes. Among the three strains maximum numbers of proteins were isolated from HP-3 strain. High pH reduced number of proteins to 12 and 3 while salinity increased number of proteins to 42 and 52 in HP-1 and HP-2, respectively due to accumulation of low molecular weight proteins in response to saline stress.

Key words: Hypericum perforatum, hypercin, pseudohypercin, hyperforin, HPLC, saline stress

Abbreviations: PMSF- Phenylmethylsulphonylfl uoride, LC/MS-Liquid chromatography/Mass spectrometry

IntroductionAmong the medicinal plants which recieved great attention during the recent years from pharmaceutical industry at global level, Hypericum perforatum ranks very high due to its secondary metabolites (hypercin and pseudohypercin) which are found to be anti depressant in action (Wentworth and Agostini, 2000). Besides, anti depressant activity H. perforatum is gaining popularity due to its antiviral (Lopez and Hudson, 1991), anti retroviral (Lavie et al., 1989) and antibiotic action (Fritz and Schleicher, 1993).

Although vast information is available on cell cultures of H. perforatum, however, investigations are required to understand the biosynthetic pathway of hypercin, pseudohypercin and hyperforin production, enhancement of production by media manipulations and induction of gene/ protein alterations by environmental stress (Ramgopal and Carr, 1991; Sairam and Aruna, 2004). Abiotic stress like salinity, pH changes and drought conditions highly infl uence biomass production, accumulation of secondary metabolites and yield of the most plants. The fast expanding research in plant molecular biology has given several clues in understanding how plant responds under stressful regime (Cushman et al., 1990). Great deal of success has been achieved in unveiling gene / protein alterations associated with preparation of plant against the abiotic stress.

Owing to the medicinal value of H. perforatum an experiment was carried out at IIIM (CSIR), Jammu to fi nd out the morphogenetic and biosynthetic potential of H. perforatum under saline and alkaline stress and to unveil the alternations in protein profi les in response to stress.

Materials and methodsThe calli resulted from shoot tip cultures were transferred to stress induced (saline and alkaline) solid MS culture media and cultivated on a constant temperature of 25+2 oC with 16 hour photoperiod. High pH stress was induced by adjusting pH of culture media to 8.5 (T-1), 9.0 (T-2), 9.5 (T-3) and 10.0 (T-4) with NaOH and saline stress was induced by incorporating different concentrations of NaCl2 0.25% (T-6), 0.5% (T-7), 0.75% (T-8) and 1.0% (T-9). The calli developed on this high pH and saline stress media were analyzed for presence and quantifi cation of hypercin, pseudohypercin and hyperforin.

Morphological analysis of callus cultures: Growth was measured in terms of fresh callus weight. All the replicates of all the treatments were harvested at the same time. All cultures were also visually screened for red colour formation and for friability at different stages of growth to assess pigment formation.

Biochemical analysis of callus culturesSample extraction for biochemical analysis: Air-dried callus of H. perforatum was extracted with chloroform and then fi ltered. The powder from fi lter paper was dissolved in acetone and concentrated on rotawave to dryness. This dried material was then dissolved in methanol and analyzed by HPLC after fi ltration.

HPLC analysis: Presence and quantifi cation of hypercins and pseudohypercins was determined by an HPLC using Gilson system (Gilson Medical Electronics, France) equipped with 321-model pump, operated at room temperature (25+2oC). Separations were performed on Lichosphere RPC-18 (5 μm, 4.6 x 150 mm) and detection was achieved with a UV detector set at 589 nm.

Journal

Appl

The chromatographic data was recorded and processed on uni-point system software. Isocratic mobile phase was used with methanol, ethyl acetate and water (pH 2.5) in the ratio of 67:16:17 respectively. pH of water was adjusted to 2.5 with O-phosphoric acid. The fl ow rate was kept constant at 1 mL min-1 and injection volume 25 μL using six point calibration curve generated with authentic hypercin and pseudohypercin.

Sample extraction for protein analysis by LC/MS: Fresh callus was homogenized with 2-3 volume of 0.1M buffer carrying 5mM EDTA. A pinch of PMSF was added during grinding. The homogenized sample was centrifuged at 4oC @ 10, 000 rpm for 20 minutes. The protein from supernatant was precipitated with 5 volumes of chilled (-20oC) acetone and all the samples were kept at -20oC for 1-2 hours. The samples were again centrifuged at 4oC @ 10000 rpm for 10 minutes to separate the precipitated proteins from the solution. The acetone was evaporated from the sample and the precipitate was dissolved in 500μL of HPLC grade water. The protein solution was then fi ltered through 0.2 mesh micro fi lter.

LC/MS analysis: The extracted samples were analyzed before LC/MS using AGLLENT 1100 series HPLC and squin 300 (Bruker/Mass spectrometer) connected by ESI (Electro Spray Ionization) interface using DAD (Diode Array Detector) at wavelength of 280.8 nm. The mobile phase was used with 0.05% TFA in acetonitrile and 0.05% TFA in water. RP-8 column with diameter of 5μm and length of 4x250 mm and fl ow rate of 0.5 mL min-1 was used. The temperature of column was kept at 300C.

Results Response of callus to high pH and NaCl stress: Calli of all three strains investigated showed similar response in terms of callus weight however gradual decrease in callus growth was observed in stress cultures (saline and high pH) as compared to control cultures (devoid of stress). Among the two stresses high pH stress showed more drastic effect than saline stress. No callus growth was observed at pH 9.5 (T3) and pH 10 (T4). Further callus showed slight increase in compactness towards the higher concentration of NaCl and decrease in compactness towards the increase in pH.

The visual examination of callus revealed dark brown to reddish brown callus and the intensity of red clour formation showed a slight increase with time. Among the three strains HP-2 strain showed more tendency towards red colour formation. No remarkable effect was shown by high pH on colour formation but NaCl showed slight increase in red colour formation. (Table 1)

Accumulation of hypercin and pseudo hypercin in response to stress: Calli of all the three strains investigated showed

production of hypercin and pseudohypercin under normal (devoid of stress) as well as under stress conditions (saline and high pH). In the present study, HP-2 strain produced higher content of hypercin (0.272) than HP-1 (0.188) and HP-3 (0.196) under control conditions. There was no remarkable effect of stress on accumulation of hypercin; however pseudohypercin production was highly infl uenced by both kinds of stress indicating that abiotic stresses has differential effect on biosynthesis of different metabolites. (Table 1).

Inability of undifferentiated cultures to produce fl avonoids under both normal and stress conditions: HPLC analysis of all the three strains revealed no signs of hyperforin production from the undifferentiated cultures where as fi eld grown plants showed presence of hyperforin particularly at the fl owering stage.

Effect of stress on protein profi le: Nine selected samples analyzed by LC/MS showed proteins of varied molecular weights (1,000-12,000 Daltons). Maximum numbers of protein were isolated from saline cultures (NaCl stress) within the retention time of 5 minutes. Among the three strains, maximum number of protein was isolated from HP-3 strain. High pH and salinity showed signifi cant effect on the number of proteins. High pH reduced the number of protein to 12 and 3 while salinity increased the number of protein to 42 and 52 in HP-1 and HP-2 strain, respectively. The qualitative change in proteins was also indicated by change in polarity and difference in retention time. Saline stress in present study altered the protein profi le by formation of proteins at different retention times and accumulation of low molecular weight protein in response to saline stress. (Table 3).

DiscussionOur results showed that the calli of all three strains used in this study had potential for accumulating hypercins and pseudohypercins under normal (devoid of stress) as well as under stress conditions (saline and high pH). Mosen et al. (1993) reported the formation of

Table 1. Effect of saline and high pH on fresh callus weight/fl ask (Mean value ± Standard error)Treatment HP-1 (g) HP-2 (g) HP-3 (g)T1 3.68±0.30 3.89±0.60 3.19 ±0.08T2 1.55±0.18 1.88±0.04 1.03±0.25T3 - - -

T4 - - -T5 5.62±0.22 5.49±0.27 5.27±0.33T6 4.45±0.24 3.87±0.32 3.61±0.33T7 3.89±0.13 2.69±0.28 3.24±0.30T8 3.15±0.35 3.15±0.15 3.77±0.46T9 - - -

Table 2. Effect of saline and high pH stress on total hypercin and Pseudohypercin contentTreatment HP-1 (%) HP-2 (%) HP-3 (%)

Hypercin Pseudohypercin Hypercin Pseudohypercin Hypercin PseudohypercinT1 0.180 0.213 0.176 0.278 0.180 0.310T2 0.172 0.229 0.184 0.246 0.176 0.164T3 - - - - - -T4 - - - - - -T5 0.188 0.574 0.272 0.590 0.196 0.573T6 0.178 0.413 0.184 0.451 0.176 0.459T7 0.176 0.203 0.178 0.369 0.192 0.328T8 0.184 0.178 0.176 0.196 0.180 0.164T9 - - - - -

Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress 65

hypercin and other metabolites in undifferentiated cultures of H. perforatum. The calli of HP-2 strain accumulated highest content of hypercin than HP-1 and HP-3 strain under control conditions. Cambell and May (1992 ) also confi rm the difference in hypercin content of different strains of H. perforatum.

Karting et al. (1996) reported that fi eld grown plants of H. perforatum produce hyperforin particularly at the fl owering stage but the production of hyperforin from undifferentiated cultures of H. perforatum has not been documented so for. Similar results were revealed during the present investigations thus confi rms the inability of undifferentiated cultures of H. perforatum to biosynthesize hyperforin. This inability of undifferentiated cultures to synthesize particular metabolite is due to loss of gene or mutation of gene involved in particular steps of biosynthetic pathway of the metabolite or unavailability of substrate and enzymes or lack of specialized cells involved in metabolite formation (Bohm, 1982).

Bohnert et al. (1995) reported that biosynthetic pathways are altered by abiotic stresses. During the present investigations saline and high pH stress showed negligible effect on hypercin production, however remarkable change was observed in pseudohypercin production (Table 2). Thus the fi ndings are in agreement with the previous fi ndings of Cushman et al. (1990) who revealed that salt and dehydration stress shows biochemical effects along with morphological, physiological and molecular effect.

LC/MS analysis revealed that all the three strains varied from one another in their protein content as was evident from the variation in the number as well as in molecular weight of proteins. Changes

in gene expression at transcriptional and post-transcriptional level are initially demonstrated by analysis of protein profi les elicited in plant using salt treatment. These studies revealed both qualitative and quantitative changes in pattern of polypeptides synthesized following salt treatment (Ericson et al., 1984; Chen et al., 1991). In the present study, high pH reduced the number of proteins perhaps due to creation of free amino acid pool due to inhibition of peptide bond formation; while salinity increased the number of proteins due to accumulation of large number of low molecular weight proteins. This coincides with Lopez et al. (1994) who reported accumulation of protein in leaves of Raphanus sativus in response to salt stress or water defi ciency. The qualitative change in proteins was also indicated by change in polarity and difference in retention time. Saline stress in present study altered the protein profi le by formation of proteins at different retention times. The change in protein profi le in response to stress is also reported by Ramgopal et al. (1991) and Sairam et al. (2004).

We conclude from our results that among the three strains used in this study HP-2 strain showed best potential for accumulating hypercins and pseudohypercins (antidepressant in function). Hypercin production did not show any change in response to stress, however pseudohypercin production was highly infl uenced by both kinds of stress. High pH stress showed more drastic effect on both morphogenetic as well as biosynthetic potential of the plant as compared to saline stress. Change in protein profi le was observed in response to both kinds of stress however this change could not be correlated with the change in accumulation of secondary metabolites due to stress.

66 Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress

Table 3: Effect of saline and high pH stress on protein profi le

Treatment Strain Retention time

Molecular weight of proteins (Daltons x 1000)12 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11

HP-l 5 min - - 8 - 1 5 1 1 4 2 T5 25min - - - - - - - - - -

60min - - - - - - - - - - HP-2 5 min 2 1 - 2 - 1 4 4 2 6

25min - - - - - - - - - -60min - - - - - - I - - -

HP-3 5 min 1 2 2 4 9 6 9 6 8 525min - 1 - - - - - - 1 ‘. -60min - - - - - - - - - -

T2 HP-l 5 min 2 - 1 .- - 1 I 3 3 -25min - - - - - - - - 1 -60min - - - I- - - - - - -

HP-2 5 min - 1 - - 1 1 - - - -25min - - - - - - - - - -60min - - - - - - - - - -

HP-3 5 min - 2 - 3 3 1 1 4 2 225min - - - - - - - - - -60min - - - - - - - - - -

T8 HP-I 5 min - - - - - 3 - - - -25min - - - - - 1 - - - -60min 3 1 1 15 3 - 8 5 6 4

HP-2 5 min - - - - - - - - - -25min 8 - - 1 2 - 7 11 15 560min - - - - - 6 - - - -

HP-3 5 min - - - 1 1 - - - - -25min - - - - - 6 - - - -60min - - - - - - - - - -

ReferencesBohm, H.1982. The inability of plant cell cultures to produce secondary

metabolites. In: Plant tissue culture, Fujiwara, A. (Ed.) Japanese Assoc. for Plant tissue culture, Tokyo. 325-326.

Bohnert, H.J., D.W. Nebson and R.G. Jonsen, 1995. Adaptation to environmental stress. Plant Cell, 7: 1099-1111.

Cambell, M.H. and C.E. May, 1992. Variation and varietal determination in H. perforatum. Plant Protection Quarterly, 2: 43-45.

Chen, R.D. and Z. Tabaeizadeh, 1991. Alteration of gene expression in tomato plants by drought and salt stress. Genome, 35: 385-391.

Cushman, J.C., E.J. De Rocher and H.J. Bohnert, 1990. Gene expression during adaptation to salt stress. Environmental inquiry of plants. Ed. F. Kalterman, Academic Press, Bandigo; 177-203.

Ericson, M.C. and S.H. Alfi nito, 1984. Protein produced during salt stress in tobacco cell culture. Plant Physiology, 74: 506-509.

Fritz, Z. and A. Schleicher, 1993. Effect of substituting milfoil, St. Johns Wort and lovage for antibiotics on chicken performance and meat quality. J. Anim. Free Sci., 2(4): 189-195.

Karting, T., I. Goebel and B. Heydel, 1996. Production of hypercin and pseudohypercin and fl avanoids in cell cultures of various Hypericum species and their chemotypes. Planta Medica, 62 (1): 51-53.

Lavie, G., F. Valentine, B. Levin, Y. Mazur, G. Gallo and D. Weiner, 1989. Studies of microorganism of action of antiretroviral agents hypercin and pseudohypercin. Proc.Natl. Acad. Sci, 86: 5963-5967.

Lopez, B.I. and J.B. Hudson, 1991. Antiviral activity of the photoactive plant pigment hypercin. Photochem and Photobio., 54 (1): 95-98.

Lopez, F., G. Yansuyt, P. Fourcroy and F. Cassedelbart, 1994. Accumulation of a 22 Kda protein and its mRNA in the leaves of Raphanus sativus in response to salt stress or water defi cit. Plant Physiol., 91: 605-614.

Mosen, B.V., K. Zdunik, H.J. Woerdenbag, N. Pras and A.W. Alfermann, 1993. Production of natural products by shoots in cultures cultivated SW in vitro. J. Pharma. World, 15: 6.

Ramgopal, S. and J.B. Carr, 1991. Sugarcane proteins and messenger RNA regulated by salt in suspension cells. Plant Cell Environ., 14: 46-47.

Sairam R.K. and T. Aruna, 2004. Physiology and molecular biology of salinity stress in plants. Curr. Sci., 86: 407-422.

Wentworth J.M. and M. Agostini, 2000. St. John’s Wort a herbal antidepressant activates the steroid X receptor. J. Endocrino.,166(3): R11-R16.

Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress 67


Growth interference of invasive Russian knapweed on “valcatorce INTA” onion

Carlos R. Bezic, Armando A. Dall Armellina, Omar A. Gajardo, Lucrecia M. Avilés and Silvia L. Cañón

Weed Ecology and Control Research Group, CURZA-University of Comahue (8500) Viedma, Río Negro province, Argentina. E-mail: [email protected]

Abstract Russian knapweed is an invasive creeping perennial herb which affects crops by competition and allelopathy. Herbicides available for use in onion are not able to control Russian knapweed in a crop context. Conversely, recommended products for Russian knapweed are not selective for the crop. The aims of this work were to study Russian knapweed biomass production and propagation for a range of increasing densities in an experimental onion culture and to characterize the productive response of onion plants under these conditions. A partial additive experiment was carried out to study Russian knapweed interference (variable density, 0-64 ramet m-2) on onion transplants (constant density, 40 pl m-2) under greenhouse conditions in Viedma, Argentina (40º 03’ S; 62º 48’ O). Although no differences among treatments were found for weed fi nal aboveground biomass, low density treatments (0, 2 ramet m-2) were lower than 64 ramet m-2 for belowground biomass. Final weed density was proportional to initial conditions. For onion, total (-54%) and commercial bulb yield (- 56 %) were reduced by weed competition with ≥ 32 ramet m-2. While size 3 bulbs (50-70 mm eq. diam.) were less represented at weed densities higher than 16 ramets m-2, size 4 ones (70-90 mm eq. diam.) were not present in this condition. For A. repens, traits such as the rate of vegetative propagation, high competitive ability, mainly belowground, and high propagule pressure support its high invasive potential.

Key words: Acroptilon repens, Allium cepa, plant competition, partial additive experiment, plant invasion, irrigated agriculture.

IntroductionAs a consequence of weed interference, vegetable yield and quality are subjected to severe losses in most irrigated areas in Argentina. Instead of rational management programs, weed control is normally done by herbicide application in a clearly reactive approach.

There are several irrigated valleys in southern Argentina whose areas are mostly dedicated to fruit and vegetable production for external markets. The Lower Valley of Rio Negro (LVRN), in particular, is located at the southeast of the Rio Negro province, near the Negro river outlet, at 40º S. It comprises 18,500 hectares of irrigated lands with a high value irrigation infrastructure of both concrete lined and unlined channels and also an effi cient drainage system. There are almost 500 properties among 30-120 hectares each.

The invasive weed, Russian knapweed (Acroptilon repens L.), is increasing in incidence at the LVRN. While 2,000 hectares were directly and indirectly affected by 1980 (Dall Armellina and Iglesias, 1984) almost 50% of the properties reported patches of different size (10 m2-100 hectares) in 2005. The affected area was actually estimated as 6,000 hectares approximately (Bezic et al., 2005).

Russian knapweed is a creeping perennial herb which affects crops by competition and allelopathy (Watson, 1980; Whitson, 1987; Dall Armellina and Zimdahl, 1988; Gajardo et al., 2004). It reproduces vegetatively by sprouting from roots. Each new ramet forms an early season rosette by the end of August

which elongates in September, the time at which onion must be transplanted. The fi rst fl owers appear in November (Bezic et al., 2005). Clonal populations of Russian knapweed may be 30-100 ramets m-2 in density and 40-60 cm in height, dense enough to be a strong competitor for any vegetable crop (Kearney et al., 1960; Whitson, 1987; Panter, 1991).

Onion (Allium cepa L.) is the 3rd place vegetable crop in the LVRN with 500-1500 hectares annually sowed (Pozzo Ardizzi et al., 2005). The species is considered a poor competitor against weeds because of its slow initial growth associated to few narrow and erect foliage leaves. If remain uncontrolled, weeds can reduce onion yield by 36-96 % (Khan et al., 2003; Williams et al., 2005).

Herbicides available for use in onion are not able to control Russian knapweed in a crop context. Conversely, recommended products for Russian knapweed are not selective for the crop.

The aims of this work were to study Russian knapweed biomass production and propagation for a range of increasing densities in an experimental onion culture and to characterize the productive response of onion plants under these conditions.

Materials and methodsSite and experimental methods: The experiment was carried out in a plastic covered 120 m2 greenhouse, located at the Universidad Nacional del Comahue (CURZA) campus in Viedma, Río Negro province, Argentina (40º 03’ S; 62º 48’ O). All lateral windows remained open over the entire experimental period (26 October, 2004-20 March, 2005) in order to avoid

Journal

Appl

excessive heating during summer. Twenty one 0.49 m2 (0.70 m by 0.70 m) wooden boxes (35 cm deep), internally lined with a 200 microns black polyethylene sheet were used. Each box was fi lled with sandy-loam soil (OM 3 %; SAR 1.9) free of Russian knapweed propagule. The soil was nutrient enriched with 0.1 kg m-3 of 15-15-15 (N:P:K) at the beginning of the experiment. All boxes were separated 20 cm from the soil line by placing them on wooden blocks and the bottom of each one was perforated for drainage purposes.

Plant material: Several clonal plants of Russian knapweed were obtained by vegetative propagation of sprouting root pieces which were selected for the presence of one visible bud preformed in the original habitat. This plant material was collected in winter 2004 from a unique population colonizing an irrigated fi eld. They were maintained under refrigeration at 5 ºC in plastic bags until September 15, 04 when they were individually placed in 368 cm3 plastic pots (64 cm2) to induce sprouting. Additionally, late onion cv. Valcatorce INTA transplants were produced in an experimental nursery until the two true leaves stage when they were transplanted in a regular 10 by 13 cm grid over the entire box surface (20 transplants per box; 408,000 plant ha-1) on 26 October, 2004.

Experimental design: A partial additive experimental design was applied to study Russian knapweed interference (variable density) on onion transplants (constant density) under greenhouse conditions. The boxes were arranged in a complete randomized block design (n=3) with 7 treatments for A. repens density (AR): AR0 (crop check without weeds); AR2 (2 ramet m-2); AR4 (4 ramet m-2); AR8 (8 ramet m-2); AR16 (16 ramet m-2); AR32 (32 ramet m-2): AR64 (64 ramet m-2). Both species were transplanted the same day and they coexisted for a period of 145 days until onion harvest. The variables evaluated for onion plants included i) total and commercial fresh bulb yield (kg m-2), ii) non commercial bulb proportion by size (bulbs <35 mm diameter) in terms of biomass and total number percentages, iii) onion plant survival (percentage of dead plants) and iv) bulb size distribution by commercial categories. Weed variables observed were: i) ramet density at harvest (fi nal density, ramet m-2), ii) above and belowground biomass at harvest (g DW m-2), iii) vegetative propagation rate (VPR, fi nal ramet number/ initial ramet number). Dry biomass was obtained by desiccation of fresh samples in an electrical oven at 65 ºC for 72 h.

Data analysis: While biomass data were logarithmically transformed (natural log, ln) to achieve homoscedasticity and

normality, percentages were transformed by the usual arcsine transformation p’ = arcsin(SQRT(p)), where p is the proportion. Data were subjected to ANOVA and the multiple comparison test SNK was applied. The LSD test was also employed in specifi c cases.

ResultsWeed growth and propagation: As season was in progress new ramets were formed by sprouting from roots. This increased the box weed density for every treatment. The statistical analysis revealed that AR2 and AR4 were lower in fi nal weed density in comparison with AR32 and AR64. Intermediate treatments were intermediate in response and no differences were detected in relation to the previous reports on this aspects(Table 1).

Vegetative propagation rate (VPR, number of fi nal ramets / number of initial ramets) was decreasing with initial density increase (Fig. 1). The highest VPR was calculated as 12.3 for AR2 and the lowest as 2.6 for AR64.

No differences among weed densities were observed for aboveground weed biomass at harvest (P = 0.13). The average value was calculated in 1.06 ± 0.11 ton DW ha-1 (Table 1).

The less conservative LSD test, however, indicated signifi cant differences between AR2 and AR32, AR64. Same results were obtained for the aboveground biomass in elongated plants (rosettes excluded) as shown in Table 1, were AR2 plots evidence a biomass lower than AR32 and AR64, which in turn were not different between them. Intermediate densities also showed an intermediate response.

Elongated ramet rate or proportion (elongated ramets/ total ramets) was the same among treatments, comprising the 62.8% of the complete ramet population. The mean elongated plant height was 34.4 cm and no differences among treatments were observed (Table 1).

Belowground biomass increased with the initial plant density in a direct proportional way. Differences among treatments were observed for AR2 and AR4 with respect to AR64. Intermediate densities also showed an intermediate response (Table 1).

Onion bulb yield: As initial weed density increased, a decreasing tendency was observed for both total and commercial yield (Fig. 2). Bulb production was not reduced under low weed density in AR2 (p>0.05). In this case the reference values were calculated

Table 1. Growth parameters of Acroptilon repens for initial densities of 0-64 ramets m-2 in onion transplanting

Initial density (ramet m-2 )

Final density (ramet m-2)

Rate of elongated plants (%)

Total aboveground biomass

(g DW m-2)

Aboveground biomass of elongated

plants (g DW m-2)

Elongated plants height

(cm plant-1)

Belowground biomass

(g DW m-2)2 25.2(6.7)a 76.7(6.2) 57.9(20.1)a 53.6(19.5)a 31.28(1.80) 25.8(8.8)a4 47.6(6.7)a 60.5(5.9) 103.4(5.0)ab 101.3(11.6)ab 32.62(1.36) 47.1(7.1)a8 87.8(30.9)ab 64.8(6.9) 117.5(34.9)ab 96.4(29.1)ab 36.57(1.48) 74.6(28.0)ab16 81.0(3.6)ab 55.2(3.3) 96.6(7.3)ab 82.0(5.3)ab 38.00(3.39) 77.3(8.4)ab32 164.6(52.8)b 61.6(4.7) 118.1(27.2)b 102.7(25.4)b 32.60(2.12) 105.6(44.4)ab64 172.1(14.9)b 58.2(3.8) 142.5(12.1)b 120.7(10.8)b 35.13(4.09) 146.8(2.4)bAverage 62.8 106.0 92.8 34.37Statistical signifi cance

P= 0.0089 P = 0.16 P = 0.13 LSD 0.05 = 59.90

P = 0.28LSD 0.05 = 58.85

P = 0.42 P = 0.0219

Data between brackets correspond to standard error (SE) of the treatment mean value. Average values calculated when no differences were found.

Growth and interference of invasive Russian knapweed on onion 69

by averaging AR0 and AR2, observing 42.3 and 41.1 ton ha-1 for total and commercial bulb yield, respectively. In AR32 and AR64 the bulb production was statistically low in comparison with AR0 and AR2. An average of 19.6 ton ha-1 of total bulb yield (-54%) and 18.1 ton ha-1 of commercial bulb yield (-56%) were calculated. Intermediate treatments were intermediate in response as were cited for the other variables studied.

No bulb losses (discard) were observed at harvest due to phytopathological disorders. Low size (bulb equatorial diameter lower than 35 mm) was the unique cause for bulb discarding. There were no differences among treatments for this variable.

Mean bulb loss was 13.7 ± 3.28 % of the total bulb number and 5 ± 1.19 % in weight, equivalent to 1.2 ton ha-1 (Table 2).

As observed yield differences can not be explained by differential bulb loss among treatments. Plant mortality was 5.7 ± 1.4 % in average without differences among treatments (Table 2).

The average size for commercial bulbs decreased as initial weed density increased. The pattern of statistical differences was the same as observed for total and commercial yield. While the mean bulb diameter was 5.6 ± 0.1 cm for AR0, the value for AR32 and AR64 was 20 % low (Table 2). The same grouping was observed

0

4

8

12

16

20

0 20 40 60 80Initial weed density (ramet m- )2

VP

R (r

amet

ram

et m

-)

2

0

15

30

45

60

0 2 4 8 16 32 64Initial weed density (ramet m- )2

Bul

b yi

eld

(ton

ha-

1)

0

5

10

15

20

25

30

35

40

0 20 40 60 80

< 35 mm35-50 mm50-70 mm70-90 mm

Initial weed density (ramet m- )2

Bul

b w

eigh

t (to

n ha

-1)

0

20

40

60

0 20 40 60 800

60

120

180

240


Final weed density (ram

et m-

) 2

Tota

l bul

b yi

eld

(ton

ha-1

)

0.0

0.5

1.0

1.5

2.0

0 10 20 30 40 50 60 700

10

20

30

40

50

AR2AR4

AR64


Wee

d dr

y w

eigh

t (to

n ha

-1)

Bulb yield (ton ha -1

)

Fig. 1. Vegetative propagation rate (VPR) of A. repens in a partial additive competition experiment with onion cv “Valcatorce INTA”. VPR = fi nal ramet number per initial ramet.

Fig. 2. Bulb yield in “Valcatorce INTA” onion under Russian knapweed competition for a range of 0-64 ramet m-2 of weed density at transplant. Bars indicate total (white) and commercial (black) yield (ton FW ha-1). While statistical analysis was performed independently for each data series, differences presented (SNK, P<0.05) are valid for both.

Fig. 3. Onion bulb yield for the most common commercial categories (discard, < 35 mm; size 2, 35-50 mm; size 3, 50-70 mm; size 4, 70-90 mm of eq. diameter) under competition with increasing densities of the invasive weed A. repens.

Fig. 4. Onion bulb yield (dot line) and fi nal density of A. repens (full line) in a partial additive competition experiment with an initial weed density range of 0-64 ramet m-2.

Fig. 5. Total Onion bulb yield (full line-thick) and fi nal A. repens aboveground (dot line) and belowground biomass (full line-thin) for increasing A. repens initial density among 0-64 ramet m-2 in a partial additive competition experiment.

Table 2. Onion yield parameters under A. repens competition (0-64 ramet m-2 at transplant time)A. repens initial density (ramet m-2)

Discard bulbs rate (number

based) %

Discard bulbs rate (weight

based) %

Onion plant mortality

rate %

Commercial bulbs

diameter (cm)0 6.7(4.4) 1.6(1.1) 1.7(1.7) 5.6(0.1)a2 4.9(2.9) 5.1(4.6) 0.0(0) 5.5(0.3)a4 12.2(6.5) 3.0(1.7) 6.7(3.3) 5.3(0.2)a8 6.7(6.7) 2.3(2.3) 5.0(0) 5.1(0.2)ab16 18.3(11.7) 7.0(3.6) 8.3(8.3) 5.0(0.2)ab32 24.1(13.0) 8.5(4.3) 10.0(0) 4.6(0.1)b64 23.0(11.7) 7.5(3.9) 8.3(4.4) 4.7(0.1)bAverage 13.7 5.00 5.71P value 0.34 0.47 0.35 0.0344 *Data between brackets correspond to standard error (SE) of the treatment mean value. Average values calculated when no differences were found.

70 Growth and interference of invasive Russian knapweed on onion

for mean bulb fresh weigh (FW) at harvest. In AR32 and AR64 the calculated value was 52.97 ± 3.31 g FW bulb-1, 47.3 % lower than AR0 (100.60 ± 4.11 g FW bulb-1). Intermediate treatments were also intermediate in response (Table 2). Fig. 3 demonstrates the incidence of weed density on bulb size for the most common categories. While size 3 bulbs (50-70 mm eq. diam.) were less represented at weed densities higher than 16 ramets m-2, size 4 ones (70-90 mm eq. diam) were not present. Low size categories remained constant in proportion for the entire weed density experimental range.

DiscussionOnion is a poor competitor against weeds. Soares et al. (2003) reported a 95 % yield loss and 91 % of decrease in mean bulb weight for a 98 days period of weed competition. The complexity increases when perennial weeds are dominant. Creeping perennials normally have a higher relative growth rate than onion, with biomass accumulation, both above and belowground, suffi ciently high to strongly compete with the crop.

For instance, onion yield losses due to volunteer potato (Solanum tuberosum) interference occur at densities commonly observed in the fi eld. One plant per 14.9 m-2 resulted in 10% crop yield loss while 100% yield loss was achieved with 4 volunteer potato plants m-2. Volunteer potato competition limits onion bulb size, resulting in a lower quality, less valuable crop (Williams et al., 2004).

Because Russian knapweed is not controlled by herbicides in a crop context, yield losses due to weed interference could be severe. For 64 ramets m-2 Watson (1980) reported 75 % grain yield loss in wheat and 88 % in corn. Other weed control operations, like mechanical or hand weeding, are also ineffective because the weed regrow and extend from creeping roots. In this experiment we observed a 56 % average reduction in the onion commercial yield for 32-64 ramets m-2 at transplant.

While an increasing tendency was observed as weed density increased, non commercial bulb proportion was statistically constant for the complete range of Russian knapweed initial density. The observations reported a high coeffi cient of variation (CV = 1.10) as consequence of which no signifi cant differences were detected among treatments. This variability was higher than that observed for total bulb yield (CV = 0.37) and also for commercial bulb yield (CV = 0.40).

Most reports indicate that certain plant traits are responsible for invasive weed dominance. As cited by Mashhadi and Radosevich (2004) and Rejmánek et al (2005) the list include i) vegetative propagation, ii) a high competitive ability in resource foraging, iii) high propagule pressure, iv) high seedling relative growth rate and specifi c leaf area, v) seed dispersal by vertebrates, vi) high phenotypic plasticity, and vii) allelopathy.

We suppose that Russian knapweed probably evidence various of these traits. In this experiment we observed at least three of these:

a) Vegetative propagation: Weed density increased along the experimental period as a consequence of vegetative propagation and interference relationships between species resulted which is a normal behaviour of creeping herbs. If we extrapolate the experimental VPR, it is possible to infer that low density patches would reach a density enough to strongly compete with onions in

a second year. A similar behaviour would be expected under fi eld conditions, at least in newly infected sites where belowground biomass tends to increase as a function of time.

From the statistical analysis applied to fi nal ramet density no differences were found between the two lowest densities, AR2 and AR4, whose average value was 36.4 ramet m-2. Additionally, no differences were observed between the two highest densities, AR32 and AR64, for which the mean value was 168.4 ramet m-2 at harvest. Both groups were, in effect, different from each other. As in all variables under study, intermediate treatments had an intermediate response.

Total onion yield correlated well with both ramet density (Fig. 4, r =-0,75) and weed belowground biomass (Fig. 5, r =-0,68). The last two variables were clearly interdependent because, at least under our experimental conditions, ramet production was directly associated with root growth (r = 0.87).

The relation among belowground biomass-bulb yield and ramet density-bulb yield could not be observed for aboveground biomass, which in turn had a similar value for the entire range of weed density. This could be possible as a consequence of intra-specifi c canopy limitation in the Russian knapweed population.

b) Competitive ability: The best predictors for competition coeffi cients for pairs of species are: i) maximum plant size, ii) root biomass allocation, iii) time of emergence and iv) seed size (Freckleton and Watkinson, 2001). Russian knapweed is clearly a dominant competitor in relation to onion plants when the fi rst three attributes from the list are considered.

It is generally assumed that root foraging traits are size symmetric in coexistent species. This implies that the intensity of resource acquisition is proportional to the size of the root system (Schwinning and Weiner, 1998). In this context, we can argue that the intensity of Russian knapweed plant competition would be a consequence of a higher size than onion plants.

However, Rajaniemi and Reynolds (2004) working with several herbs including Centaurea maculosa, which is similar to Russian knapweed, demonstrate size asymmetry in weed root foraging due to an anticipated use of limited soil nutrients. We also support this hypothesis with respect to Russian knapweed competition. In established patches of this weed there is a clonal root system that is both extensive and pre-existent as compared with any direct seeded or transplanted vegetable crop. Except for water, that is not limited under furrow irrigation, nitrogen (N) fertilization for instance is, both spatially and temporally concentrated. Conventional N applications are normally localized near the crop rows in two times along the crop cycle. Then, asymmetric competition resulted as consequence of the extension of the weed root system and also by an anticipated access to soil nutrients with respect to crop plants.

c) Propagule pressure: The persistence of an important bank of vegetative propagule allows us to understand the importance of controlling the progress of Russian knapweed invasion in agricultural fi elds at the LVRN. We must bear in mind the high cost of land and also the loss of agricultural aptitude in invaded soils. As cited by Soukup et al. (2004), crop losses and land degradation are common in both natural and agricultural systems affected by invasive weeds.

Growth and interference of invasive Russian knapweed on onion 71

In the irrigated LVRN we have demonstrated the competitive ability of Russian knapweed over one of the most important vegetable alternatives in the south of Argentina and 2nd national vegetable crop with 25,000 ha cultivated area and 700,000 ton annual production (FAO, 2006).

Characterization of the vegetative propagation ability of Russian knapweed explained the constantly modifi ed crop-weed interference relationship. It is important to gain knowledge about the importance of plant invasions and their impact in order to persuade the public and private sectors about the urgent need to limit the invasive weed dispersal and, additionally implement concrete plans for controlling established populations in both private agricultural and public sites.

AcknowledgementsThis research was funded by the Universidad Nacional del Comahue throughout its Secretaría de Investigación and it is part of Carlos Bezic Doctoral thesis at Universidad Nacional del Sur, Bahía Blanca, Argentina. We specially thank undergraduate student Sergio Vazquez for fi eld and lab support.

ReferencesBezic, C.R., O.A. Gajardo, S. Polo, L. Avilés, S. Cañón, W. Benitez,

S. Vazquez, S. Iribarne, D. Schwindt and A. Dall Armellina, 2005. Distribución y abundancia del yuyo moro (Acroptilon repens L.) en el Valle Inferior del río Negro. Actas Jornadas Interdisciplinarias de Estudios Agrarios, Bs. As., Argentina, 8 p.

Dall Armellina, A. and H. Iglesias, 1984. Estrategias para el control de “yuyo moro” Centaurea repens L. en parcelas hortícolas del Valle Inferior del Río Negro. Actas X Reunión Nacional sobre malezas y su control. Tucumán, Argentina, Tomo II: 43-46.

Dall Armellina, A.A. and R.L. Zimdahl, 1988. Effect of light on growth and development of fi eld bindweed (Convolvulus arvensis) and Russian knapweed (Centaurea repens). Weed Sci., 36: 779-783.

FAO. 2006. FAOstat Prodstat module. On-line serch at http: //faostat.fao.org.

Freckleton, R.P. and A.R. Watkinson, 2001. Predicting competition coefficients for plant mixtures: reciprocity, transitivity and correlations with life-history traits. Ecol. Lett., 4: 348-357.

Gajardo, O., C. Bezic, L. Avilés, S. Cañón and A. Dall Armellina, 2004. Alelopatía del yuyo moro (Acroptilon repens L.) sobre maíz dulce. Revista Pilquén, 6: 9-18.

Kearney, T.H., R.H. Peebles, J.T. Howell and E. Mc Clintock, 1960. Arizona Flora. 2nd ed., University of California Press, Berkeley, CA, 1085 p.

Khan, M.H., N. Khan and N. Badshah, 2003. Effect of weedicides and hand weedings on the yield of onion (Allium cepa L.). Asian J. of Plant Sci., 2(6): 464-466.

Mashhadi, H.R. and S.R. Radosevich, 2004. Invasive plants: Ecol. and management. In: Weed Biology and Management. Inderjit (ed.) Dordrecht, The Netherlands. Kluwer, p 1-28.

Panter, K.E. 1991. Neurotoxicity of the knapweeds (Centaurea spp.) in horses. In: Noxious range weeds. James and Lynn F. (ed.) Boulder, CO. Westview Press. p 316-324.

Pozzo Ardizzi, M.C., M. Abrameto, G. Pellejero, G. Aschkar, M.I. Gil, and A. van Konijnemburg, 2005. Efecto del período de conservación sobre algunas propiedades nutracéuticas y organolépticas en los bulbos de cultivares nacionales de cebollas (Allium cepa L.) en el Valle Inferior de Río Negro. RIA (Argentina), 34(3): 115-130.

Rajaniemi, T.K. and H.L. Reynolds, 2004. Root foraging for patchy resources in eight herbaceous plant species. Oecologia, 141: 519-525.

Rejmánek M., D.M. Richardson and P. Pyšek, 2005. Plant invasions and invasibility of plant communities. In: Vegetation Ecology. Van der Maarel E. (ed.). Blackwell Sci., Oxford. p-332–355.

Schwinning, S. and J. Weiner, 1998. Mechanisms determining the degree of size asymmetry in competition among plants. Oecologia, 113: 447-455.

Soares, D.J., R.A. Pitelli, L.T. Braz, R. Gravena and R.E.B. Toledo, 2003. Periods of weed interference with transplanted onion crop (Allium cepa). Planta daninha, 21(3): 387-396.

Soukup, J., J. Holec, P. Hamouz and L. Tyšer, 2004. Aliens on arable land. Proceedings of the Scientifi c Colloquium Weed Sci. on the Go. Universität Hohenheim, 11-22.

Watson, A.K. 1980. The biology of Can. weeds. 43. Acroptilon (Centaurea) repens (L.) DC. Can. J. Plant Sci., 60: 993-1004.

Whitson, T.D. 1987. Weeds and poisonous plants of Wyoming and Utah. Res. Rep. 116-USU. Laramie, WY: University of Wyoming, College of Agriculture, Cooperative Extension Service. 281 p.

Williams, M.M. II, C.V. Ransom and W.M. Thompson, 2004. Effect of volunteer potato density on bulb onion yield and quality. Weed Sci., 52: 754-758.

Williams, M.M., C.V. Ransom and W.M. Thompson, 2005. Duration of volunteer potato (Solanum tuberosum) interference in bulb onion. Weed Sci., 53: 62-68.

72 Growth and interference of invasive Russian knapweed on onion


Caulifl ower hybrids for spring production in southern mediterranean area

M. Sciortino1 and G. Iapichino2

1Dipartimento di Scienze e Tecnologie Agroambientali, Università degli Studi di Bologna, Via Fanin 44, 40127, Bologna, Italy. 2Dipartimento di Agronomia Ambientale e Territoriale, Sezione di Orticoltura e Floricoltura, Università degli Studi di Palermo, Viale delle Scienze, 90128, Palermo, Italy. 2E-mail: [email protected]

AbstractNew cultivars (F1 hybrids) of caulifl ower (Brassica olearacea L. var. botrytis L.) were evaluated at four crop densities (1.3, 1.7, 2.2 or 3.3 plants m-2) for spring harvest crop in a Southern Mediterranean area (western coast of Sicily). The F1 hybrids (‘White-Flash’, ‘Milky-Way’ and ‘White Excel’) having white head, usually cultivated in Northern Italy and Europe in the autumn, were used. The aim was the introduction of new varieties which can fi ll the gap from mid May to mid July, now existing in the Sicilian caulifl ower production, which is based on autochthonous ecotypes of green head varieties, e.g. ‘Cavolfi ore Verde di Palermo’. Crop density signifi cantly infl uenced the growth and the phenology of the new hybrids. It was positively correlated to earliness, total marketable yield and inversely to unmarketable product percentage and head size. The best crop density was found to be 2.2 plants m-2. Among the cultivar tested ‘White Flash’ and ‘Milky Way’ appeared particularly suited for a spring harvest in the experimental environment. They gave high yields with a minimum discard and uniform heads of approximately 1 kg of weight each.

Key words: Brassica olearacea L. var. botrytis L., planting density, cultivars, quality

IntroductionAccording to FAO statistics (2000-2005), Italy is the third producer (500,000 t) of caulifl ower (Brassica olearacea L. var. botrytis L.) in the world, after China and India (7,250,000 and 4,800,000 t, respectively), followed by Spain, USA and France. Italian production has always been commercialized also abroad, but in the last thirty years exports gradually decreased (Liguori,1985; Franca, 1985; Renzoni 2004). France and Spain are particularly strong competitors. They are offering caulifl owers of better and more constant quality standards. Moreover their farms are generally larger than in Italy and this helps in reducing management costs. Nowadays even traditional European importers, i.e. The Netherlands, Great Britain and Germany, are becoming new exporters, particularly with the introduction of cold resistant hybrids (Ranco, 1985; Franca, 1985; Renzoni, 2004).

In the past, Italian production was based on ecotypes with loose corymbs. Generally they lacked colour and size uniformity and their ripening period was prolonged, aspects that heavily affected commercialization and reduced customer satisfaction (Baldoni, 1982; Liguori, 1985; Acciarri et al., 2004). Exceptions were rare. For example ‘Romanesco’ and ‘Verde Maceratese’ were appreciated in the North European market for the typical shape and for the green colour of their heads, respectively. In the last few years Italian ecotypes lost importance, although many of them have been genetically improved, maintaining their typical organoleptic properties and particular taste. In Italy, most of the caulifl ower production comes from central and southern regions (ISTAT, 2005). Campania is the fi rst producer, followed by Puglia, Calabria, Abruzzo, Marche and Sicily (Table 1). Marche’s crops has always been focused towards the export. On the contrary, most of the Sicilian products are commercialized on the home market. Sicilian farmers generally use local varieties which

are less appreciated abroad. Their heads are usually coloured, too heavy (about 2.5-3 kg head-1) and often scarcely uniform. In western Sicily, the typical cultivar is ‘Cavolfi ore Verde di Palermo’ which has green, compact and medium-large corymbs, while in the oriental part of the island there is a predominance of ‘Cavolfi ore Violetto di Catania’ that is characterized by large and purple heads (Branca and Iapichino, 1997; Acciarri et al., 2004). The ‘Cavolfi ore Verde di Palermo’ has several ecotypes with different earliness of production (Incalcaterra and Iapichino, 2000). These types, selected by growers during the centuries, are now well identifi ed and are called with the name of the harvest month. They allow a continuous production sequence from August to April, with a unique gap that goes from the middle of May to mid July (Branca and Iapichino 1997, Incalcaterra and Iapichino 2000). The goal of this research was the introduction of new caulifl ower cultivars in the occidental part of Sicily for a spring harvest in order to cover this productive gap. Today most of the new available varieties of caulifl owers are F1 hybrid that were bred in localities quite different from the Sicilian environments. The adaptability of these species to many climates is confi rmed by several studies. For example, according to Wurr et al. (1993) head formation requires a vernalization period at 9- 21°C. Grevsen and Oelsen (1994) indicated a range between 0 and 26°C, while Wiebe (1990) suggests maximum vernalization stimulus at 16 to 30°C. A further proof is provided by the caulifl ower crops in tropical regions, with about 25°C as yearly mean temperature. However, the fi tness of a new caulifl ower variety to a particular environment (and market) should always be thoroughly evaluated. In this instance, a good cultivar for a spring cultivation in the Southern Mediterranean area like Sicily should produce smaller heads than the indigenous ecotypes in order to be appreciated on the international market. Moreover, it should be precocious and tolerate hot temperatures during head formation. Only thus

it can give new chances to the Sicilian caulifl ower production and marketing. The aim of the study was to explore the hybrid caulifl ower production during mid may to mid July in Northern Italy and Europe.

Material and MethodsThe research was carried out at the experimental farm of the Horticulture and Floriculture branch of the Department of Agronomy of the University of Palermo in the north-western coast of Sicily (38°7’ N; 13°22’ E), Italy . The fi eld experiment was conducted on a typical Mediterranean litosoil. It consists of a 50 cm deep layer of sandy-clayey textured soil, lying on dolomitic limestone rock, with the following main characteristics: sand 53.7%, lime 30.3%, clay 16.0%, pH 7.9, active limestone (Ca) 4.3%, Kjeldahl nitrogen (N) 1.2%, organic matter 3.1% (Lotti Method), Olsen phosphorous (P2O5) 0.3% and exchangeable potassium (K2O) 0.1%. Climatic parameters were continuously recorded by a meteorological station situated at the experimental farm. The experimental fi eld was previously cropped with winter melon (Cucumis melo L. var inodorus Naud.). In February 2000, a ploughing to 0.35 m was carried out, followed by fl exible harrowing. Before transplanting, mineral fertilizer (11-22-16 NPK) was applied at 0.6 t ha-1 as a basal dressing. On March 19th 2000 the caulifl ower seeds were sown in plastic trays with 45 mm diameter holes (Dufault and Waters, 1985a) and fi lled with dark peat; the plantlets were placed fi rst in a nursery and then outdoor for a slow acclimatizing. Transplanting was made on April 14th with 4-5 leaves/plant. The crop was hand weeded and was irrigated throughout the season according to its need: one irrigation per week starting from the middle of May.

The experiment consisted of a factorial comparison between three caulifl ower cultivars and four crop densities. The cultivars were commercial F1 hybrids: ‘White-Flash’, ‘Milky-Way’ and ‘White Excel’, obtained by Sakata Co. (Japan). All of them have white head and are widely used for autumn harvest in Northern Italy and all over Europe. Compared densities were 1.7, 1.3, 2.2 or 3.3 plants m-2. They were obtained by maintaining constant distance (1 m) between row and changing that along the row: 0.30- 0.45- 0.60 and 0.75 m.

Experimental plots were 18 m2 of area and were randomized according to a complete randomized block design with three replicates. At harvest, border plants within each plot were discarded to avoid interferences between treatments. During the growing season the rate of leaf appearance was assessed by leaf counts on 15 plants for each plot at 15, 30 and 45 days after transplanting. Moreover the time of head formation was recorded when the change of the apices became visible and heading earliness was expressed as number of days from transplanting. At harvest, caulifl ower heads were hand collected as soon they were reached marketable stage i.e. suffi ciently large, but still compact. Head collection implied three passages in each plot, throughout June. Collected corymbs were sorted into 5 diameter classes: Ø<0.11m; 0.11-0.13m; 0.14-0.155m; 0.156-0.18m and Ø>0.18m. This classifi cation is linked to the number of heads that can be placed in a standard commercial box and represents the base of the selling price of caulifl owers in the Italian market. Each box contains 24 corymbs of the 0.11-0.13m class, 18 pieces of the 0.14-0.155m class, 12 pieces of the 0.156-0.18m class and

9 heads with Ø>0.18m. Ø<0.11m caulifl owers are considered not marketable.

Data were analysed by a Two-Way ANOVA (cultivar x crop density) procedure and the differences between means were evaluated by Duncan’s multiple range test (SAS®, ProcAnova/DuncanTest).

ResultsDuring the experiment climate was very favourable to caulifl ower growth (Fig. 1). Minimum temperature varied from 3.4°C in the fi rst part of March to 19.5°C in last days of June, while maximum temperature ranged between 17.0 and 34.6°C in the same periods, respectively. Total rainfall was about 50 mm distributed during the growing season. In particular, 26.65 mm fell soon after transplanting (April 10 and 20) and two rainfall (8 and 6 mm) happened during fertile stem elongation and head ripening, respectively.

Leaf appearance rate: Leaf appearance rate was strongly reduced by increasing crop density (Table 2). This effect can be easily explained by considering the higher competition at root and shoot level of densely grown plants. The parameter tested signifi cantly varied among cultivars. Fifteen days after transplanting ‘White Flash’ showed more leaves per plant (1 or 3 more than the other cultivars) and this difference was particularly marked in denser crops. After 30 days ‘White Flash’ again had the maximum number of leaves. In this instance, however, the differences with the other cv. were signifi cant only at the lowest density. Regardless of the planting density, 45 days after transplanting, ‘Milky-Way’ produced the highest number of leaves per plant.

Head formation: The passage from vegetative to reproductive phase varied widely (from 36.0 to 55.5 d from transplanting) and depended mainly on genotype (Table 3). On average, ‘Milky Way’ was the fi rst hybrid to develop corymbs, followed by ‘White Excel’ and ‘White Flash’. Heading response to crop density was variable as well. The competition stress that is typical of a dense stand accelerated head formation. However this effect was marked in ‘White Excel’ (approximately one week between the highest

Table 1. Regional distribution of Italian caulifl ower production (ISTAT, 2005)Regions Area (ha) Production (t ha-1)Puglia 2692 15.30Campania 2858 29.61Calabria 2316 33.31Abruzzo 2117 23.73Marche 1943 23.25Sicilia 1804 21.86Italian 17950 24.10

Fig. 1. Air temperatures and rainfalls from March to July 2000.

74 Caulifl ower hybrids for spring production in southern mediterranean area

marketable yield was positively correlated to plant density. This implies that, even at the highest compared density, the plant competition was not so strong as to impair the high yield potential of the new hybrids. Only for ‘Milky Way’ the highest density slightly reduced yield, probably because its large plants are better suited to less dense plantings.

Head quality: As expected, the weight of single heads (Table 5) was negatively correlated to plant density, but genotypic infl uence was equally important. ‘Milky Way’ always produced the heaviest corymbs, which exceeded 1.3 kg head-1 at the lowest density. The head size was most sensitive to density in ‘White Excel’. From the lowest to the highest density the weight of its heads decreased more than half a kg.

Unmarketable heads were mainly due to bolting. Internode elongation, less leaves, thicker leaf lamina and thinner stem are all symptoms of competition stress that predispose the plant to bolting (Table 6). Thus this adversity was clearly associated to crop density. However genotype also had signifi cant effect. In particular, the discard of ‘White Flash’ was always inferior to 20%, even at the highest density (Table 6). On the contrary, in ‘White Excel’ it consistently represented more than a quarter of collected heads, with a slight variation in response to crop density. ‘Milky Way’ showed an intermediate behaviour. In general, the increase of crop density augmented the frequency of small caulifl owers reducing the number of the extra-large ones (Fig. 2). In ‘White Flash’ Ø<0.11m corymbs were very few (10.8 %) at 1.3 plants m-2 compared to 35.0% that were found at 3.3 plants m-2, while the opposite happened for Ø>0.18m (31.3 and 5.4 %, respectively). Also, in ‘Milky Way’ the extreme classes were markedly infl uenced by plant density: the percentages of corymbs belonging to the smallest class were 28.6 and 4.2% at 3.3 and 1.3 plants m-2 density, respectively, while those with Ø>0.18m were 0.9 and 35.2%, respectively. A similar pattern, even more

Table 2. Infl uence of cultivar and crop density on the leaf appearance assessed 15, 30 and 45 days after transplantingCrop density (plant m-2)

Number of leaves plant-1

‘White Flash’ ‘Milky Way’ ‘White Excel’15 days 3.3 8.7D 5.4G 6.7F

2.2 10.8B 8.6D 7.8E

1.7 9.6C 9.3C 9.5C

1.3 11.6A 11.5A 10.8B

30 days3.3 16.5BE 14.8DF 14.4DF

2.2 19.6AB 15.6CE 15.0DF

1.7 17.3BE 16.1BE 16.3BE

1.3 20.8A 18.7CE 18.2AD

45 days3.3 20.9D 24.3BC 18.4E

2.2 21.7C 24.8B 21.3D

1.7 24.8B 25.2A 23.6BC

1.3 27.3A 28.2A 24.4C

Means followed by different letters are signifi cantly different according to Duncan’s test at P<0.05Table 3. Caulifl ower heading time as affected by genotype and crop densityCrop density(plant m-2)

Heading time (Days after transplanting)

‘White Flash’ ‘Milky Way’ ‘White Excel’

3.3 36.0H 53.2B 41.8F

2.2 39.2G 54.6A 43.3E

1.7 39.3G 55.3A 44.8D

1.3 39.6G 55.5A 48.1C

Means followed by different letters are signifi cantly different according to Duncan’s test at P<0.05Table 4. Infl uence of genotype and crop density on the yield of marketable caulifl ower headsCrop density (plant m-2)

Marketable caulifl ower heads (t ha-1)‘White Flash’ ‘Milky Way’ ‘White Excel’

3.3 21.77A 15.40CDE 21.42A

2.2 16.07CD 18.28B 17.28BC

1.7 15.06DE 13.99DE 10.27G

1.3 13.24EF 11.54FG 8.20H

Means followed by different letters are signifi cantly different according to Duncan’s test at P<0.05

Table 5. Infl uence of genotype and crop density on the unit weight of collected caulifl ower headsCrop density (plant m-2)

Unit head weight (g head-1)

‘White Flash’ ‘Milky Way’ ‘White Excel’

3.3 774GH 893G 711H

2.2 993F 1023DE 796G

1.7 959EF 1186C 1015DE

1.3 1053D 1353A 1270B

Means followed by different letters are signifi cantly different according to Duncan’s test at P<0.05.Table 6. Percentage of unmarketable caulifl ower heads as affected by genotype and crop densityPlant density (plant m-2)

Percent unmarketable heads‘White Flash’ ‘Milky Way’ ‘White Excel’

3.3 19.3EF 24.5CD 32.1A

2.2 11.4G 22.1DE 28.3B

1.7 8.5H 19.1F 25.4BC

1.3 6.7H 17.5F 26.6BC

Means followed by different letters are signifi cantly different according to Duncan’s test at P<0.05

and lowest density) and less noticeable in the other two hybrids (2-3 days of difference, Table 3).

Head yield: For each hybrid, the harvest period lasted less than 15 days and all heads were collected in three harvests. ‘White Flash’ was the fi rst cultivar to give a marketable production (in the fi rst part of June), followed by ‘Milky Way’ and then by ‘White Excel’, which started to be collected at the end of June.

The highest yields of marketable heads (Table 4) were obtained with ‘White Flash’ and ‘White Excel’, both at the densest stand (21.8 and 21.4 t ha-1, respectively). ‘White Excel’ at the lowest density gave unsatisfactory production (8.2 t ha-1). On average,

Caulifl ower hybrids for spring production in southern mediterranean area 75

marked, was observed for ‘White Excel’. The percentage of its corymbs belonging to the smallest class ranged from 2.6 to 33.2% when density increased from 1.3 to 3.3 plants m-2, while the largest heads decreased from 30.5 to 0.7%. On the whole, the percentage of intermediate diameter classes was more stable over crop densities and hybrids.

DiscussionIn this study, the increasing plant density signifi cantly augmented the marketable yield as already observed by Salter and James (1975) in caulifl ower. Our results are also consistent with those obtained by Dufault and Waters (1985), who found that high plant density reduced head weight in caulifl ower cultivar Snow Crown. ‘White Excel’ appeared highly infl uenced by density, as at the highest stand it gave heads that were so small (less than 800 g head-1) as to be scarcely appreciated both by home and international customers. ‘White Excel’ gave the most suitable weights for the international market (1 kg head-1) at 1.7 plant m-2, whereas ‘Milky Way’ and ‘White Flash’ gave satisfactory head weight with high productivity even at 2.2 plant m-2. Higher percentages of too small (Ø<0.11m) and too large (Ø>0.18m) corymbs were obtained at 3.3 and 1.3 plants m-2, respectively.

Among the cultivar tested, ‘Milky Way’ can be considered the best responsive with medium sized head of even uniformity over a wide range of densities. ‘White Flash’ was the earliest hybrid, followed by ‘Milky Way’ and ‘White Excel’. This earliness was due to faster development, particularly during the initial stages of

growth, where it showed the fastest leaf formation. An increase of crop density sped up growth. This infl uence, combined with the genotype growth characteristic, can be used to achieve an earlier production at the end of spring, which is particularly imporant for the Sicilian caulifl ower market.

The amount of unmarketable product was affected by density through bolting; indeed at the highest density the discarded head percentages was high. But bolting effect was infl uenced by hybrid susceptibility as ‘White Excel’ (being later producing) was more prone than ‘White Flash’ (being earlier producing).

The results indicate that the tested hybrids can be successfully cropped in a Mediterranean climate area such as the western coast of Sicily. Indeed, they can almost completely fi ll the gap in the Sicilian caulifl ower production (mid May- mid July) with a high productivity, that was confi rmed in our agronomic environment. Both, the high productivity and the quality standard that the tested cultivar showed might be greatly appreciated on the international market, opening the export market to local growers.

Planting density infl uenced both plant growth and phenology: denser stands promoted earliness and augmented yield, but implied smaller corymbs and high proportion of discardable heads. Thus medium densities are recommended. On the whole, the better quantitative and qualitative results were obtained with ‘White Flash’ and ‘Milky Way’ transplanted at a density of 2.2 plants m-2: this density seems thus to be the best choice as reference for further experiments.

AcknowledgmentsWe are grateful to Mr. La Rosa and Mr. Galati for technical support.

ReferencesAcciarri, N., F. Branca, E. Sabatini, S. Argento and V. Magnifi co, 2004.

Miglioramento genetico dei cavolfi ori a corimbo bianco e colorato. L’Informatore Agrario, 24: 33-36.

Baldoni, G. 1982. La coltivazione del cavolfi ore. Edagricole, Bologna, pp. 83.

Branca F. and G. Iapichino, 1997. Some wild and cultivated Brassicaceae exploited in Sicily as vegetables. Plant Genetic Resources Newsletter, 110: 22-28.

Dufault, R.J. and L. Waters Jr.,1985a. Container size infl uences broccoli and caulifl ower transplant growth but not yield. HortScience, 20 (4): 682-684.

Dufault, R.J. and L. Waters, Jr., 1985b. Interaction of nitrogen fertility and plant populations on transplanted broccoli and caulifl ower yields. HortScience, 20(1): 127-128.

F.A.O. Food and Agriculture Organization of the United Nations. http://www.fao.org/es /ess/top/commodity.jsp?lang=EN.

Franca, F. 1985. Associazionismo per difendere e riallacciare la produzione del cavolfi ore. L’Informatore Agrario, 16: 43-44.

Goldberg, D., B. Cornat and Y. Bar, 1991. The distribution of roots, water, and minerals as a result of trickle irrigation. J. Amer. Soc. Hort. Sci., 96: 645–648.

Grevsen, K. and J.E Oelsen, 1994. Modelling caulifl ower development from transplanting to curd initiation. J. Hort. Sci., 69: 755-766.

Incalcaterra, G. and G. Iapichino, 2000. Sowing time influences caulifl ower seed production. Acta Hort. 533: 45-52.

ISTAT Istituto Nazionale di Statistica. http://www.istat.it/agricoltura/datiagri/ coltivazioni/anno2005.htm#italia.

Fig. 2. Distribution of collected heads into diameter classes (<11, 11-13, 14-15.5, 15.5-18, >18 cm) as affected by hybrid and planting density: (a) White fl ash (b) Milky Way, (c) White Excel.

a

b

c

76 Caulifl ower hybrids for spring production in southern mediterranean area

Liguori, A. 1985. Cavolfi ore: come riconquistare i mercati. L’Informatore Agrario, 22: 75-79.

Ranco, E. 1985. Diffi cile ma non impossibile riconquistare le posizioni perdute sui mercati esteri. L’Informatore Agrario, 16: 21-24.

Renzoni, F., 2004. Obiettivi qualità certifi cata e tipicità per il cavolfi ore nazionale. L’Informatore Agrario, 24: 37-39.

Salter, P.J. and J.M. James, 1975. The effect of plant density on the initiation, growth and maturity of curds of two caulifl ower varieties. J. Hort. Sci., 50 (3): 239-248.

SAS Institute, 1988. SAS/STAT User’s Guide, Release 6.03 Edition. Cary NC, SAS Institute, Inc.

Wiebe, H.J. 1990. Vernalization of vegetables crops- a review. Acta Hort., 267: 323-328.

Wien, H.C. and D.C.E. Wurr, 1997. Caulifl ower, Broccoli, Cabbage and Brussels Sprouts. In: The Physiology of Vegetable Crops. Wien H.C. (Ed.). CAB International, New York, USA.

Wurr, D.C.E., J.R. Fellows, K. Phelps and R.J. Reader, 1993. Vernalization in summer/autumm caulifl ower (Brassica olearacea var. botrytis L.). J. Exp. Bot., 44: 1507-1514.

Caulifl ower hybrids for spring production in southern mediterranean area 77


Effect of different sowing times on development and effi ciency of some chinese cabbage varieties (Brassica campestris sbsp. pekinensis)

Funda Eryilmaz Acikgoz

Faculty of Agriculture, Namik Kemal University, Tekirdag, Turkey. E-mail: [email protected]

AbstractThe aim of the study was to determine the effect of different sowing times on development and effi ciency of some Chinese cabbage varieties (Brassica campestris sbsp. pekinensis) under Corlu conditions. The study was conducted in Corlu County which has a tougher climate than its Province Tekirdag where a similar research had been done before. The research was conducted in 2000 and three different sowing times (15 August, 15 September and 15 October) and four domestic varieties (Tokat-2, Tokat-5, Tokat-29 and Tokat-89) were used. The variety, Tokat-89 and the sowing time of 15 September were found to be the most suitable variety and sowing time, respectively, The variety and time of sowing recorded superiority for head weight, level of hardness and head quality.

Key words: Chinese cabbage (Brassica campestris sbsp. pekinensis), sowing times, development, effi ciency

IntroductionChinese cabbage is a vegetable which is widely grown and pleasingly consumed in the East Asian Countries; especially in China, Japan, Korea and Taiwan, and which is also very well adapted to the cool climate conditions (Vural et al., 2000). When it is eaten fresh and raw, it is crisper, fresher and easily digested than the other cabbage sorts. For a working man, 100 g of Chinese cabbage supplies the total of daily need for calcium, magnesium, and vitamin C while it also provides some other minerals and elements important for human nutrition (Eryilmaz and Varis, 1996).

Since, Chinese cabbage has a short vegetation period as 2-2.5 months; it can be grown as the second crop after harvesting of cereals. By using the region’s advantage of being closer to a metropolis as Istanbul, Chinese cabbage can be marketed in the same season as lettuce for which its production can be done as an alternative. Moreover, its production is suggested in between the growing periods of wheat and sunfl ower when the fi elds remain unused. Whether the results of the research conducted in Tekirdag Province in 1993 (Eryilmaz and Varis, 1994) can also be used for Corlu County which has a tougher climate was the objective of the research.

Materials and methodsThe research was conducted in Corlu County of Tekirdag Province in 2000 under fi eld conditions. The varieties, Tokat-2 (V1), Tokat-5 (V2), Tokat-29 (V3) and Tokat-89 (V4) well adapted to Tokat Region of the country, were used in the experiment and the seeds were obtained from Faculty of Agriculture; Gazi Osman, Pasa University, Tokat, Turkey.

The seeds of the varieties were sown as 2 seeds for each PE bags of which the closed dimensions were 15 x 15 cm and the thickness was 0.15 mm, black coloured in order to check moss growth and had bellows in order to stand upward, and under

which there were drainage holes and which were fi lled in with peat. The sowings were done at the dates of 15 August (S1), 15 September (S2) and 15 October (S3). The seedlings were planted in double lines on rows at 40x40 cm distances in the rows and on the rows respectively.

Production plan of the experiment1st Sowing time : 15 August 1st Planting time : 16 SeptemberFertilizing : 30 September 1st Harvesting time : 22 OctoberDays from sowing to harvesting : 672nd Sowing time : 15 September 2nd Planting time : 9 OctoberFertilizing : 24 October 2nd Harvesting time : 14 DecemberDays from sowing to harvesting : 943rd Sowing time : 15 October 3rd Planting time : 8 NovemberFertilizing : 23 November 3rd Harvesting time : 18 JanuaryDays from sowing to harvesting : 94Irrigation was supplied with a fi ltered bucket in the seedling period and in addition it was provided by furrow irrigation and raining irrigation with a hose in the development period. (NH4)2SO4 2% N was used for fertilization and it was supplied as 7 g m-2 (Eryilmaz and Varis, 1996). The experiment was set in a double recurrence and according to the experimental design of divided parcels.

The average temperature was 17.7 oC, the average rainfall was 44.3 mm, the average proportional humidity was 68.2 % and the average wind speed was 2.1 m s-1 during the experimental period (July-October).

The observations and analysis methods were as per Opena and Lo (1980) and Eryilmaz and Varis (1996). Head length (cm) was measured from the bottom to the top of the head. Head width was measued in the middle part of the head cut at length (cm). For

Journal

Appl

number of leaves, all leaves were counted and averaged. For head weight plant head weight with outer leaves was measured after harvesting. Heading effi ciency was calculated by dividing mean head weight with non-wrapper leaf weight.

For head tightness, all plants were observed and evaluated after each harvesting and they were explained as tight head, loose head and not forming head by (%). Hardness was determined by using following formula: Hardness (g cc-1) = Average head weight head volume -1. Where, Head volume (cc) = 0.524(d1

2 d2-1), d1 = Average

Head width (cm), d2 = Average head length (cm).

Results and discussionHead weight: Analyis of variance for testing effect of variety and sowing time on head weigh revealed that the mean sowing effect was signifi cant at P=0.01 for head weight and sowing time of 15 September gave the plant with the heaviest head (Table 1).

Tight head: Analysis of variance indicated that the main sowing effect was significant at P=0.01 for the tightness of head. Consequently, the sowing time of 15 August gave the tightest head (Table 2). Tight head is an important factor for Chinese cabbage because of easy marketing (Gercekcioglu and Yazgan, 1989).

Loose head: Analysis of variance revealed that the sowing time was important for loose head formation. Thus, the sowing time of 15 October gave the plants having the loose heads.

Percentage of not forming head: The number of plants which did not form head was quite low. Hence, variance analysis was not realised. From the sowing time of 15 August, 7.4 % and from the sowing time 15 October 20 % of not forming head plants were

obtained among the sowing times with respect to the averages. There was no signifi cant difference within the varieties. However, Tokat-29 gave the highest percentage of not forming head plant (Table 4).

Hardness: The hardness of the varieties ranged from 0.39-0.42 (Table 3). The higher the hardness is, the tighter the head will be and this is a desirable property, because the paleness is attained later in plants with tight head. There was an excess weight in the plants which have smaller volume and they are comfortably and profi tably marketed (Yazgan and Edizer, 1987). Moreover, a high value of hardness is advantageous for the producer, carrier, marketer and consumer due to enduring (Yazgan and Ece, 1990). Table 3. Hardness in Chinese cabbage varieties

Variety Hardness (g cc-1)Tokat-2 0.39Tokat-5 0.40Tokat-29 0.40Tokat-89 0.42

The results of the research are given as a whole in Table 4. For Corlu conditions; the variety of Tokat-89 and the sowing time of 15 September were determined as the most convenient variety and sowing time, respectively, regarding: head weight, level of hardness and head quality.

ReferencesEryilmaz, F. and S. Varis, 1994. The effect of different sowing times

on development and effi ciency in some Chinese cabbage varieties. The J. Trakya Univ., 3(1-2): 1-8.

Eryilmaz, F. and S. Varis, 1996. The effect of different sowing times on

Table 1. The effect of sowing times on average head weight in some Chinese cabbage varieties (g)Sowing Times Tokat-2 Tokat-5 Tokat-29 Tokat-89 Main Sowing Effect15 August 1070.0 1475.0 1326.5 1375.0 1311.6AB15 September 1470.0 1532.0 1610.0 1600.0 1553.0A15 October 1500.0 1200.0 1100.0 1300.0 1275.0BMain variety effect 1346.6 1402.3 1345.1 1425.0 1379.8LSD for Main Sowing Effect: LSD (P=0.01)=192.7Table 2. The effect of sowing times on the ratio of tight head in some Chinese cabbage varieties (%)Sowing Times Tokat-2 Tokat-5 Tokat-29 Tokat-89 Main Sowing Effect15 August 63.0 71.0 56.5 68.0 64.6A15 September 62.0 58.0 57.5 43.0 55.1AB15 October 51.5 60.0 43.0 60.0 53.6BMain variety effect 58.8 63.0 52.3 57.0 57.7

LSD for Main Sowing Effect: LSD (P=0.01) = 11.2

Table 4. The results of the criteria analysed (S: sowing time; V: variety)Treatment Head length

(cm)Head width

(cm)Number of

leavesHead weight

(g)Seed forming

head (%)Tight head ratio (%)

Loose head ratio (%)

Not forming head (%)

Heading effi ciency

S1V1 23.8 15.8 36 1070.0 0 63 37 0 21400S1V2 22.0 18.3 30 1475.0 0 71 30 0 29500S1V3 24.0 16.7 33 1326.5 0 56.5 43 0 26530S1V4 23.2 15.9 30 1375.0 0 68 42 0 27500S2V1 26.2 16.1 33 1470.0 0 62 38 0 29400S2V2 24.3 16.7 35 1532.0 0 58 41 4 30640S2V3 27.4 15.2 33 1600.0 0 43 43 11 32000S2V4 28.2 16.4 34 1610.0 0 57.5 32 14 32200S3V1 25.5 16.3 40 1500.0 0 51.5 30 18 30000S3V2 22.8 15.8 42 1200.0 0 60 17 23 24000S3V3 23.4 16.1 33 1100.0 0 43 20 35 22000S3V4 25.2 15.2 38 1300.0 2 60 15 25 26000

The effect of different sowing times on development and effi ciency of some chinese cabbage varieties 79

development and effi ciency in some Chinese cabbage varieties. The Project of South-eastern Anatolia (GAP) 1st Agriculture of Vegetables Symposium: 106-112, 7-10 May, 1996, Sanlıurfa, Turkey.

Gercekcioglu, R. and A.Yazgan, 1989. A research on determining the convenient Chinese cabbage varieties (Brassica campestris L. Ssp. pekinensis Lour Olsson) for Tokat region’s fall season. Doga TU Tar. ve Or., 13(2): 255-264.

Opena, R.T. and S.H. Lo, 1980. Procedure for Chinese cabbage evaluation trials. International Co-operator’s Guide, AVRDC 80-144, 6.

Vural, H., D. Esiyok and I. Duman, 2000. Culture Vegetables (Vegetables Growing), Ege University Prints, Izmir, Turkey. p 23-129.

Yazgan, A. and Y. Edizer, 1987. The determination of Chinese cabbage varieties (Brassica campestris L. Ssp. pekinensis Lour Olsson) convenient to be grown in Tokat Province’s spring and summer. The Journal of Cumhuriyet University Faculty of Agriculture, 6(1): 127-149.

Yazgan, A. and A. Ece, 1990. The adaptation of Chinese cabbage variety nominees Tokat-2, Tokat-5, Tokat-29 ve Tokat-89 in Tokat conditions. The Journal of Cumhuriyet University Faculty of Agriculture, 6(1): 439-449.

80 The effect of different sowing times on development and effi ciency of some chinese cabbage varieties

appl hort horticulture of applied... · appl hort issn 0972-1045 horticulture applied journalournal...

Documents