S.Afr.J.Bot., 1993, 59(1): 9 - 13 9

Seed structure and germination of Dichrostachys cinerea

W.E. Bell and J. van Staden* UN/FRO Research Unit for Plant Growth and Development, Department of Botany, University of Natal, Pietermaritzburg, 3200

Republic of South Africa

Received 28 January 1992; revised 10 August 1992

Dichrostachys cinerea (L.) Wight et Arn. (Legume: Mimosoideae) is a major contributor to the bush encroachment problem in Mkuzi Game Reserve. Seed characteristics were examined so that ecological and managerial concepts could be evaluated. Dormancy is imposed by the smooth, hard and impermeable testa. Scarification or cracks in the testa allow imbibition through the lens, which is followed by germination. The percentage germination of Dichrostachys cinerea is higher in light than in dark and there is a loss of temperature sensitivity with time. The behaviour of Dichrostachys cinerea seeds is not as that typically reported for mimosoid seeds, since germination improved with storage, and light enhanced germination of scarified seeds. This has important ecological implications, since seeds would germinate easier where the canopy has been removed or the soil has beeh disturbed.

Dichrostachys cinerea (L.) Wight et Arn. (Legume: Mimosoideae) is 'n belangrike bydraer tot die bos­indringerprobleem in die Mkuzi Wildreservaat. Saadkaraktertrekke is bestudeer sod at ekologiese en bestuurskonsepte geevalueer kon word. Saadrus word veroorsaak deur die gladde, harde en waterdigte testa. Skarifikasie en barste in die testa laat imbibisie deur die lens toe sodat ontkieming kan plaasvind. Die persentasie ontkieming van Dichrostachys cinerea is hoer in die lig as in die donker en daar is 'n verlies van temperatuursensitiwiteit met tyd. Dichrostachys cinerea sade reageer nie soos verwag word van mimosoid­sade nie, aangesien ontkieming met opberging verbeter het, en lig die ontkieming van geskarifiseerde sade verhoog het. Hierdie aspek het belangrik ekologiese implikasies omdat sade beter sal ontkiem waar die plantegroei verwyder of die grond versteur is.

Keywords: Dichrostachys cinerea, legume, Mimosoideae, seed, dormancy, seed structure, germination .

-To whom correspondence should be addressed.

Introduction Bush encroachment is a major problem in Mkuzi Game Reserve (Goodman 1977). Dichrostachys cinerea (L.) Wight et Am. grows on a variety of soils within the reserve and is one of the major contributors to the bush encroachment problem. It is a thombush belonging to the Leguminosae subfamily Mimosoideae. Due to climate, stocking densities and management practices, dense thickets of this plant have formed, preventing grazing animals from reaching available grass. While this plant can make a valuable contribution as feed for browsing stock, it nevertheless reduces the carrying capacity of the veld. It is therefore necessary to determine the physiological and ecological factors which contribute to the spread of D. cinerea in its natural habitat if control programmes are to be instituted.

The occurrence of hardseededness in members of the Leguminosae is well established (Comer 1951; Bewley & Black 1982; Van Staden et al. 1989). At present, little is known about the seed characteristics of D. cinerea. This must be remedied before ecological and managerial con­cepts and strategies can be fully evaluated. In this paper we report in detail on the structure ~nd hardseeded character­istics of the seed.

Materials and Methods All the seed used in this study was collected in Mkuzi Game Reserve in northern Natal during May and June 1989. Only dry mature pods were collected from the trees. Unctamaged

seeds were removed from the pods and these were stored in sealed glass bottles at 25°C in the dark. To minimize damage by bruchids (Coleoptera: Bruchidae), Captab (N-tri­chloromethylthio-cyclohex-4-ene-1,2-dicarboximide) was in­cluded with the seeds.

To determine the degree of hardseededness, ten batches of five seeds each were maintained at 25°C in the light and in the dark on moist filter paper in Petri dishes. At 12-hourly intervals the seeds were blotted dry using absorbent paper towelling and then weighed. Mechanical scarification was effected by nicking the seeds at the chalazal end with a sharp scalpel blade. These seeds were incubated at 25°C in the light and in the dark. Seeds were incubated in controlled incubators at 20, 25, 30 and 35°C in light and the rate of imbibition and germination was recorded every two days. The experiment was repeated 2 and 3 months later on the same seed batch. For the germination experiments, radicle emergence was recorded daily. The route of water uptake was studied by immersing the seeds in a coloured solution (aqueous fast green 0.1 % w/v) and tracing the path of water entry by examining the seed under a light microscope after various time intervals. Two methods of seed preparation for scanning electron microscopy (SEM) were compared. Seeds were sputter-coated and viewed as previously described (Manning & Van Staden 1987), or the E.M. SP 2000 sputter cryo low-temperature system was used where the tempera­ture in the microscope was lowered to -175°C using liquid nitrogen, and the specimens were then examined.

10

Results

Seed structure

The dry mature seed of D. cinerea is a typical mimosoid (Gunn 1981) with a pleurogram, hilum, lens and micropyle (Figure la). The surface of the testa was smooth following sputter-cryo-coating (Figure la) and cracked after sputter­coating (Figure lb). The sputter cryo method was adopted

200 tim

S.-Afr.Tydskr.Plantk., 1993,59(1)

for all further SEM work. The seed coat is composed of the

cuticle, macrosclereids with a clearly visible light-line,

osteosclereids and parenchyma (Figure lc). The vascular

bundle is visible below the macrosclereids towards the lens

where it lies beneath the epidermal palisade (Figure ld). The

macrosclereid cells in the lens region are shorter than those

surrounding the hilum (Figure Id).

Figure 1 The lens end of the seed of Dichrostachys cinerea showing the lens (L), micropyle (M) and hilum (H). These scanning electron micrographs were prepared using: (a) sputter cryo, and (b) sputter-coating. The latter method of coating results in cracks (arrow) in the testa. (c) A transverse section through the testa of Dichrostachys cinerea showing the cuticle (C), the macrosclereids, apparently divided by a light line (LL), osteosclereids (0) and the parenchyma (P). (d) The vascular system (arrow) in the seed of Dichrostachys cinerea at the lens end. It starts at the hilum (H), where it enters the seed and extends into the mesophyll tissue before continuing towards the lens (L).

S.AfrJ.Bot.,1993,59(1)

Imbibition Fresh undamaged seeds of D. cinerea did not imbibe or germinate unless scarified. Once the seed had been mechani­cally scarified, imbibition and germination took place providing that water was available. Water uptake was rapid during the first 24 h (Figure 2). No difference was noted between seeds imbibed in the light or dark (Figure 2). Seeds which were unscarified did not germinate in the light or dark (Figure 3). Once seeds were scarified, germination was significantly higher in the light than in the dark (Figure 3).

Effect of temperature on the imbibition and germination of Dichrostachys cinerea seeds

The initial rate of imbibition of scarified seeds incubated at 20, 25 and 30°C was similar (Figure 4). Incubation of the scarified seeds at 35°C reduced the rate of water uptake significantly (Figure 4d). Unscarified seeds did not imbibe or germinate. Germination of the scarified seeds occurred more rapidly at 30°C than at 20, 25 or 35°C (Figure 5). Repetition of this experiment on seed stored for two and three months, respectively, resulted in greater rates of germination for seeds incubated at 25 and 35°C (Figures 5b,d), while seeds stored for three months and incubated at 30°C resulted in similar rates and final percentage germ ina-

100

0, 80

.5 ~ 60 « ~

Z 40 « UJ ~

20

o

® light @ dark

o 2 4 6 8 jo 0 2 4 6 8 10

TIME (DAYS)

Figure 2 The rate of water uptake of -seeds stored at 2SOC and subsequently incubated at a temperature of 2SOC in the light (a) and dark (b). Seeds were unscarified (0) or scarified (e) . The vertical bars represent ± 2 S.E.

100

Z o ~ 80 Z ~ ffi 60 (!)

UJ (!) 40 « t-Z ~ 20 0: UJ a. 0

® light @ dark

o 2 4 6 8 10 12 0 2 4 6 8 10 12

TIME (DAYS)

Figure 3 The percentage germination of seeds which received no pretreatment and which were subsequently incubated at 25°C in the light (a) or dark (b). Seeds were unscarified (0) or scarified

(e). The vertical bars represent ± 2 S.E.

11

tion (Figure 5c). Germination of scarified seeds at all temp­eratures was approximately 80 - 90%.

100

80

60

40

C) 20 E ~

~ 0 « ~

Z100 « UJ ~

80

60

40

20

o

® 20·C @ 2S·C

/ r © 30·C @ 3S·C

r I o 2 4 6 8 10 0 2 4 6 8 10

TIME (DAYS)

Figure 4 The rate of water uptake of scarified seeds incubated at 20 (a), 25 (b), 30 (c) and 35°C (d), respectively (n = 100 for

each trial). The vertical bars represent ± 2 S.E.

100 ® 20.C @ 2S·C

80

60

40

@ 3S·C

Jl

( ,

o 2 4 6 8 10 12 0 2 4 6 8 10 12

TIME (DAYS)

Figure 5 The percentage germination of scarified seeds incu­bated at 20 (a), 25 (b), 30 (c) and 35°C (d), respectively (n = 100

for each trial). Broken lines represent repetition of this experiment at a later date: the results for 25°C were obtained after two months and for 30 and 35°C after three months. The vertical bars repre­sent ± 2 S.E.

12

Site of water entry

Only a few of the intact, unsorted seeds which had been soaked in the fast green dye for 6 h imbibed. In these seeds the green dye was localized in patches on the seed coat and in the pleurogram (Figure 6a). Those seeds which had not imbibed were stained in the hilar region (Figure 6b) when compared with those seeds which had not been placed in the

S.-Afr.Tydskr.Plantk.,1993, 59(1)

dye (Figure 6c). Permeable seeds which were left for 12 h became swollen and green at the lens (Figure 6d). The tissue split slightly at the lens to reveal a cavity which became visible once the cap was dislodged (Figure 6e). Transverse sections of the permeable seeds showed that the dye entered the testa at the lens (Figure 6f), and accumulated in the vascular tissue between the lens and the hilum.

Figure 6 Untreated, unsorted seeds were used in experiments to trace the path of water entry into the seeds. (a) A seed of Dichro­stachys cinerea left in fast green for 12 h which imbibed. The dye penetrated portions of the testa and pleurogram (P) (X437.5). (b) Seed of Dichroslachys cinerea which was soaked in fast green dye for 12 h but did not imbibe. Dye was absorbed at the hilum region (H), but not at the lens (L) or micropyle (M) regions of the seed (X875). (c) Dichroslachys cinerea seed which was not placed in dye (control). Lens (L), hilum (H), micropyle (M) (X 875) . (d) : Seed of Dichroslachys cinerea, left in fast green dye for 12 h. The seeds which the dye had penetrated were selected for observation. The dye had penetrated the lens (L) which was swollen around the edges (arrow), leaving a small split in the centre. Hilum (H), mi<..1opyle (M) (X875). (e) Seed of Dichroslachys cinerea left in fast green dye for 12 h. The swollen tissue around the lens (L) of permeable seed was scraped away with a scalpel to reveal a sub-lens area (arrow). Hilum (H), micropyle (M) (X 875) . (f) Transverse section through the lens end of a permeable, imbibed Dichroslachys cinerea seed showing the vascular trace (arrow). Lens (L), hilum (H), micropyle (M) (X525).

S.AfrJ.Bot.,1993,59(1)

Discussion Many authors have described the testa surface of the mimosoid seed as cracked (Dell 1980; Manning 1987; Bhat­tacharya & Saha 1990). The testa surface of D. cinerea is smooth and the cracks that were observed apparently re­sulted from the method of seed preparation for scanning electron microscopy. The use of the sputter cryo procedure ensured that no artifacts occurred during the preparation of seed for scanning electron microscopy. No anomalies were noted in the structure of the lens, micropyle and hilum.

The shorter macrosclereid cells that were observed at the lens (Hamly 1932; Aitken 1939; Ballard 1973), and the close contact between the vascular trace and lens (Newman 1934; Tran 1979; Dell 1980), suggests that this is the natural site for water entry. The fact that the vasculature between the lens and the hilum became stained would suggest that the xylem is functional (Dell 1980) and not blocked as was the case for Acacia kempeana (Muell.) (Hanna 1984). The precise role of the hilum in imbibition of D. cinerea seeds was not established.

No reports have been published which indicate a sensi­tivity of legume seeds to light. Preece (1971) and Scifres (1974) have shown that for Acacia aneura and A. farne­siana, gennination results were identical in the light and the dark. D. cinerea appears to have a degree of light sensitivi­ty, as the percentage gennination was greater in the light than in the dark. This would suggest that there may be more than one donnancy mechanism operative and that donnancy may be broken in response to a number of environmental cues. Although light increased final percentage gennination, damage to the seed coat remains a prerequisite for imbibi­tion and germination. A sensitivity to light would ensure that seeds closer to the soil surface genninated preferentially compared to those buried in the soil. There also appears to be a loss of sensitivity to germination temperature with time which would allow for a number of seeds to germinate later in the summer season. Once again this type of behaviour has not been reported for a legume seed.

It would appear, therefore, that D. cinerea is not behaving exactly as other mimosoid seeds, since the removal of the characteristic water impermeable condition did not result in 100% gennination. Instead, the germination of D. cinerea improved with storage, while light appears to promote germination once seeds have been scarified. The improved response to light may be an important factor for seedling establishment in areas where fire is of frequent occurrence or used as part of the veld management procedure. Canopy removal results in higher light intensities and may well increase seedling establishment and subsequent bush en­croachment.

Acknowledgements The Foundation for Research Development and the Univer-

13

sity Research Fund are thanked for financial support.

References AITKEN, Y. 1939. The problem of hard seeds in subterranean

clover. Proc. R . Soc. Vict . 51 : 187 - 213.

BALLARD, L.A.T. 1973. Physical barriers to germination. Seed

Sci. Technol . 1: 285 - 303.

BEWLEY, J.D. & BLACK, M. 1982. The Physiology and Bio­

chemistry of Seeds in Relation to Germination, Vo!. 2. Springer-Verlag, Berlin.

BHATTACHARYA, A. & SAHA, P.K. 1990. Ultrastructure of

seed coat and water uptake pattern of seeds during germination

of Cassia sp. Seed Sci. Technol. 18: 97 - 103.

CORNER, EJ.H. 1951. The leguminous seed. Phytorrwrphology

1: 117 - 121.

DELL, B. 1980. Structure and function of the strophiolar plug in seeds of Albizia lophantha. Am. J. Bot. 67: 556 - 563.

GOODMAN, P.S. 1977. Soil types prone to bush encroachment. In: Proceedings of Veld and Bush Management for Beef and

Game Production, pp. 61 - 71. Hlabisa Soil Conservation

Committee. GUNN, J.R. 1981. Seeds of Leguminosae. In: Advances in

Legume Systematics (11), eds. R.M. Polhill & P.H. Raven, pp.

913 - 925. Royal Botanic Gardens, Kew.

HAMLY, D.H. 1932. Softening of the seeds of Melilotus alba.

Bot. Gaz. 93: 345 - 375. HANNA, PJ. 1984. Anatomical features of the seed coat of

Acacia kempeana (Mue1!.) which relate to increased germina­

tion rate induced by heat treatment. New Phytol. 96: 23 - 29. MANNING, J.C. 1987. The legume seed as a multifunctional

organism (with particular reference to the Papilionoideae).

Doctoral thesis, Department of Botany, University of Natal,

Pietermaritzburg.

MANNING, J.C. & VAN' STADEN, J. 1987. The role of the lens in seed imbibition and seedling vigour of Sesbania punicea

(Cav.) Benth. (Leguminosae: Papilionoideae) . Ann. Bot. 59:

705 - 713. NEWMAN, LV. 1934. Studies in the Australian Acacias: 4. The

life history of Acacia baileyana F.V.M. Proc. Linn. Soc N.S .w.

59: 277 - 313. PREECE, P.B . 1971. Contributions to the biology of mulga 2.

Aust.1. Bot. 19: 39 - 49. SCIFRES, C.J. 1974. Salient aspects of Huisache seed germina­

tion. SW Naturalist 18: 383 - 392. TRAN, V.N. 1979. Effects of microwave energy on the strophiole,

seed coat and germination of Acacia seeds. Aust. J. PI. Physiol .

6: 277 - 287.

VAN STADEN, J., MANNING, J.C. & KELLY, K.M. 1989.

Legume seeds - the structure : function equation. In: Advances

in Legume Biology, eds. C.H. Stirton & J.L. Zarucchi (eds).

Bot. Monogr. Syst . Mo . bot. Gdn. 29: 417 - 450.