OpnRlrroNlr, AND ScrENTrFrc NorES r09

EGG HATCHING OF AEDES MOSQUITOES DURING SUCCESSIVEFLOODINGS IN A RIFT VALLEY FEVER ENDEMIC AREA IN

KENYA

r. M. LoGAN, K. J. LrNrHrcuM,l ".:JHtil3f,r{.r WAGATEH, c. o. NELSoN'eNp

United Stdtes Army Medical Research Unit-Kenya, Box 401, APO New Yorh,09675-5000

ABSTRACT. Floodwater Aedes breeding habitats in central Kenya were sequentially flooded todetermine the numbers of mosquito eggs hatching during each flooding. Approximately 90% of tbe larvaesampled during 4 floodings emerged during the initial flooding. The number of Aedes eggs hatchingduring the second flooding was lowest of all 4 floodings, and no significant differences in the amount ofegg hatching during floodings 3 and 4 were seen. Unhatched Aedes eggs were present in soil samplescollected after the final flooding. The possible implications of these findings with regard to Rift Valleyfever virus control are discussed.

Rift Valley fever (RVF) virus causes diseasein domestic animals and humans. During inter-epizootic periods, the virus is thought to survivein Kenya in transovarially infected A edes mcin-toshi Huang eggs (Linthicum et al. 1985). Nat-ural RVF viral infections have been found in theadults of many other mosquito species; the epi-demiology of RVF has been recently reviewed(Meegan and Bailey 1988).

Floodwater Aedes spp. eggs enter a state ofdiapause that is broken when the embryonatedeggs are submerged in water of reduced oxygencontent (Gjullin et al. 1941); however, not alleggs hatch uniformly in response to submersionin deoxygenated water, and some require re-peated exposures before hatching. Most AedesDittatus (Bigot) eggs from mud samples collectedin Nigeria hatched during the second and thirdflooding, and none hatched after a sixth flooding(Service 1970). More than 50% of Aedes detritus(Haliday) eggs hatched during floodings 4-6when mud samples containing these eggs weresoaked 13 times (Service 1968). Cooney et al.(1981) obtained93-99% hatching of Aedes uex-@ns eggs in laboratory tests in which knownnumbers of eggs were placed in beakers, coveredwith various depths of soil and flooded to stim-ulate hatching; only the initial flooding was per-formed. Tree hole debris collected by Buxtonand Breland (1952) was flooded 13 times, andfrom this, Aedes triseriatus (Say) was collected12 times, and, Aedes zoosophus Dyar and Knabwag recovered 5 times.

Mosquito control efforts directed against im-mature stages of floodwater Aedes spp. in Kenyamay provide an effective vector-reduction strat-egy for limiting RVF virus introduction intosusceptible animals, thus curtailing RVF out-

t United States Army Medical Research Institute ofInfectious Diseases, Fort Detrick, Frederick, MD,2r702-50rr.

breaks. The potential for use of a sustained-release methoprene formulation to interruptboth the enzootic and epizootic cycles of thisdisease was demonstrated (Linthicum et al.1989). However, specific information regardingthe frequency and occurrence of expected egghatch of floodwatet Aedes spp. during the initialand subsequent flooding in these areas is un-known. The main breeding habitats for flood-water Aedes spp. associated with RVF virus are"dambos" or shallow, streamless, grasslanddepressions associated with river drainage sys-tems subject to seasonal flooding (Mackel 1974).To implement vector-control methods effec-tively against immature mosquito stages, thenumber of expected egg hatching during succes-sive floodings of dambos must be known. Theobjective of this study was to determine thenumber of Aedes spp. eggs hatching in floodwa-tet Aedes spp. after floodings of dambos in aRVF endemic zone.

The study was conducted during the dry sea-son from January to March 1989 at a dambosystem located on the west bank of the KiuRiver approximately 8 km SSE of Ruiru, ThikaDistrict, Central Province, Kenya (1' 13'S; 36"58'E, altitude 1,500 m). During 3 days of thestudy period, there was an accumulation of 109mm of rainfall and no natural flooding of thedambo occurred. Two distinct areas, 180 m2each, within the dambo system were formed byconstructing a trenched barrier lined with poly-urethane sheets to prevent water flow in or outofthese areas (Fig. 1). The 2 areas, separated byca. 25 m, were similar in terms of plant species,primarily the sedge Cyperus immensus (C. B.Clarke), and the grass Digitaria abyssinica (A.Richard) Stapf. The sites were flooded with riverwater and maintained at a constant level ofapproximately 0.3 m for 5 days. Dip collections(0.47-liter dippers) were made daily for 4 days,starting day 1 after flooding. Fifty samples were

MARCH 1991 OpnnerroNel nNo ScrpNrrFrc NorES

Table 1. Mean daily number of Aedes larvae collected per dip in each site during 4 sequential floodings.

Site 1 Site 2

FloodingMean -rSE

no. larvaelV o ftotal2

Mean -rSE

no. Iarvae'% o ftotal'

150 + 168 + 26 - f 14 1 - 1

t Number of dips : 100.2 Percents within a column followed by different letters are significantly different based on Tukey's

Studentized range test (P < 0.05).

Table 2. Number of Aedes spp. eggs recovered from soil samples collected in study areas sequentially flooded.

Site 1 Site 2

1

oA

230 t 2l5 + 18 + 18 + 1

89a5b4b2b

92a2b3c3c

FloodingMean +SD

no. eggstV o ftotal2

Mean +SDno. eggst

% o ftotal2

Prior to 1Prior to 2Prior to 3Prior to 4After 4

555 + 383I / = r t

2 7 + 2 7Q R + Q O

1 3 + 2 4

86a3b4b,c5b2c

104 + 841 0 + 94 I + 2 41 2 t I 72 - r 4

61a6b

25b

lct Number of samples were 12 prior to the first and fourth flooding, 6 prior to second and third flooding and

50 after the fourth flooding.'Percents within a column followed by different letters are significantly different (P < 0.05) based on a

rank-sum test.

Ae. unidentatus Mcintosh, Ae. quasiuniuittatus(Theobald) and Ae. sudanensis (Theobald). Themost common species in this area are Ae. mcin-toshi and,Ae. dentatus, representing 5l and46%,respectively, of the total Aedes species (Linthi-cum et al. 1988). Selected rearings during thepresent study produced all the above speciesexcept Ae. unid,entarfus and rearings during thefourth flooding yielded Ae. mcintoshi, Ae. den-tatus and Ae. circumluteolus from both sites.Species identification of Aedes eggs was notpossible and thus could not be used to determinethe relative abundance.

In the present study, significantly more flood-wat"er Aedes eggs hatched during the first flood-ing in this study than during subsequent flood-ings. This is in contrast to the rock-pool breed-ing of Ae. uittatus in which few eggs hatchedduring the first soaking (Service 1970) or thesalt marsh breeding Ae. detritus that yielded lessthan 5% hatch during the first soaking (Service1968). The egg-hatch characteristic ofthe flood-water Aedes spp. Iocated in the dambos (i.e.,>90% hatch during the first flooding) may pro-vide an opportunity for control measures to beapplied effectively against the immature stages.One application of a sustained-release formula-tion of methoprene in Altosid@ pellets was com-pletely effective in blocking both Aedes and Cu-ler adult emergence for at Ieast 2 wk after flood-

ing in a similar dambo system (Linthicum et al.1989). The same formulation can be used as apretreatment in mosquito-breeding areas severalweeks in advance of flooding with similar effec-tiveness, allowing the applicator to treat theseflood-prone areas before access roads becomeimpassable due to the onset of rains. Control ofthe initial hatch of Aedes spp. in RVF endemicareas may reduce the population of transovari-ally infected vectors to a Ievel low enough toprevent the initiation of RVF epizootics.

The authors thank C. W. Kamau for technicalassistance, J. P. Kondig for assistance with sta-tistical analysis and M. J. Turell and S. W.Gordon for critically reviewing the manuscript,and R. A. Ward for editorial assistance. Theviews of the authors do not purport to reflectthe position of the U.S. Department of theArmy, the U.S. Department of Defense or theGovernment of Kenya. Published with the ap-proval of the Director of Veterinary Services,Government of Kenya.

REFERENCES CITED

Buxton, J. A. and O. P. Breland. 1952. Some speciesof mosquitoes reared from dry materials. Mosq.News 12:209-214.

Cooney, J. C., E. Pickard, J. W. Upton and B. R.Mcduff. 1981. Tillage-a nonchemical method forthe control of floodwatef mosquitoes. Mosq. News

7r2 JouRuer, oF THE AMERTcAN Mosqurro CoNrRol Assocrertotr VoL. 7, No. 1

4l:642-649.Gjullin, C. M., C. P. Hegarty and W. B. Bollen. 1941.

The necessity of a low oxygen concentration for thehatching of Aedes mosquito eggs. J. Cell Comp.Physiol. l7:793-202.

Horsfall, W. R. 1956. A method for making a surueyof floodwater mosquitoes. Mosq. News 16:66-71.

Linthicum, K. J., C. L. Bailey, F. G. Davies, A. Kairoand T. M. Logan. 1988. The horizontal distributionof Aedes pupae and their subsequent adults withina flooded dambo in Kenya: implications for RiftValley fever virus control. J. Am. Mosq. ControlAssoc. 4:551-554.

Linthicum, K. J., F. G. Davies, A. Kairo and C. L.Bailey. 1985. Rift Valley fever virus (family Bun-yaviridae, genus Phlebouinrs). Isolations from Dip-tera collected during an inter-epizootic period inKenya. J. Hyg. 95:197-209.

Linthicum, K. J., T. M. Logan, P. C. Thande, J. N.

Wagateh, C. W. Kamau, C. L. Bailey, F. G. Daviesand J. P. Kondig. 1989. Efficacy of a sustained-release methoprene formulation on potential vectorsof Rift Valley fever virus in field studies in Kenya.J. Am. Mosq. Control Assoc. 5:603-605.

Mackel, R. 1974. Dambos: a study in morphodynamicactivity on the plateau regions of Zambia. Catenal:327-365.

Meegan, J. M. and C. L. Bailey. 1988. Rift Valleyfever. pp. 57-76. In: T. P. Monath (ed.), The arbov-iruses: epidemiology and ecology. Vol. IV. CRCPress, Boca Raton, FL.

Service, M. W. 1968. The ecology of the immaturestages ofAedes detritus (Diptera: Culicidae). J. Appl.Ecol. 5:613 630.

Service, M. W. 1970. Studies on the biology and tax-onorny of Aed.es (Stegornyia) uittatus (Bigot) (Dip-tera: Culicidae) in Northern Nigeria. Trans. R. En-tomol. Soc. Lond. 121:101-143.