ECOLOGICAL STUDIES OF CTENOSCIARA HAWAIIENSIS (HARDY) (Diptera: Sciaridae) 2

Transcription

1 Pacfc Insects 15 (1): May 1973 ECOLOGICAL STUDIES OF CTENOSCIARA HAWAIIENSIS (HARDY) (Dptera: Scardae) 2 By Wallace A. Steffan 1 Abstract: The seasonal fluctuatons of Ctenoscara hawaenss (Hardy) populatons n two dfferent ecosystems at the same elevaton on Mauna Loa on the sland of Hawa are compared and correlated wth fluctuatons n ranfall, humdty, and temperature. Ranfall appears to be the most mportant extrnsc factor affectng seasonal changes n ths speces. There was a marked dfference between the two populatons whch n part can be explaned by observed dfferences n the ecosystems. Ctenoscara hawaenss s closely assocated wth Acaca koa, and the larvae lve under the bark of dead branches. In culture, developmental tme from ovposton to adult emergence ranged from days at a constant temperature of 20 C. The Scardae or dark-wnged fungus gnats are a promnent element n many of Hawa's ecosystems. The adults are small, usually dark fles commonly found around decayng plant and anmal materals. The larvae are characterzed by the shny black head and the featureless, twelve-segmented, whte, translucent body. Larvae generally feed on decayng plant and anmal materals, anmal excrement, molds and fung; some feed on lvng plant tssues. Hardy (1960) revsed the Hawaan Scardae and provded an excellent base for future studes. Seventeen speces were recognzed by Hardy, and approxmately 40 speces are now known from ths sland chan. None of the genera are endemc ; however, many of the speces are. Because these fles are easly transported by man and hs actvtes, and Scardae n general are poorly known and dffcult to dentfy, determnaton of endemc versus ntroduced speces s an arduous task. Ctenoscara hawaenss was selected as one of the speces to be ntensvely nvestgated snce t s relatvely easy to dentfy, s common to all the stes beng nvestgated and appears to be closely assocated wth one of the domnant elements of the plant communty Acaca koa. The seasonal fluctuatons of C. hawaenss populatons n two dfferent ecosystems at the same elevaton on Mauna Loa are compared and correlated wth fluctuatons n ranfall, humdty, and temperature. Laboratory studes of C. hawaenss are reported and correlated wth feld observatons. The nvestgaton of Hawaan Scardae s beng contnued as part of ntegrated ecosystem studes under the Internatonal Bologcal Program (IBP). 1. Contrbuton No. 14. ISLAND ECOSYSTEMS IRP/IBP HAWAII. Partal results of grants to Bshop Museum from the Natonal Scence Foundaton (GB-6761, GB-23075). 2. Bshop Museum, Honolulu, Hawa

2 86 Pacfc Insects Vol. 15, no. 1 METHODS A modfed Malase trap (Gresstt & Gresstt 1962) was used to sample the scard populatons. Ths s an ntercepton trap desgned to trap the nsects as they fly. Presumably wth ths type of trap, t s easer to obtan ndces of absolute populaton snce t ntercepts the nsects more or less randomly. Jullet (1963) compared a Malase trap wth others and suggests that for larger Hymenoptera and some Dptera ths trap s unbased, but he found t unsatsfactory for Coleoptera and Hemptera. Malase traps were set up n the Klauea Forest Reserve (Transect-Profle 2, segment ll) and near the IBP weather staton on the Mauna Loa Transect (Transect-Profle 1, segment 7). These two stes are beng ntensvely studed by other researchers nvolved n the Islands Ecosystems IRP. Both of the traps are near IBP weather statons and are at approxmately the same alttude, 1660 m. The ecosystems, however, are dfferent. Klauea Forest Reserve s a montane ran forest composed prmarly of Acaca koa, Metrosderos and tree ferns (Cbotum speces). The Mauna Loa ste s n a mountan parkland ecosystem formed by Acaca koa tree colones and Styphela-Dodonaea tall-shrub communtes, both n a matrx of subalpne grassland. Durng the frst sx months, samples were taken contnuously and collected weekly. After July 5, 1971, the stes were sampled on alternate weeks only, because t was not feasble to analyze the volume of materal collected on a weekly bass. Laboratory cultures were started from gravd females or larvae collected n the feld and from adults reared from rottng wood. The adults or larvae were placed n glass shell vals (25 x 95 mm) contanng an agar substrate. Ths s a modfcaton of a culture method used by Metz and hs students (Smth-Stockng 1936) for genetc studes of Scardae and s descrbed below. It was ntally reported n Steffan (1966). Agar substrate : Mx 4.2 g of Bacto-Corn Meal Agar and 2.0 g of Bactoagar n 200 ml of dstlled water. If excess fungal growth s detrmental to the scard culture, a plan Bacto-agar medum can be used. Ths s prepared by mxng 8 g of Bacto-agar n 200 ml of water. Heat mxture n a pan of bolng water for 10 mnutes. Pour approxmately 3 cm of ths mxture nto each val, plug wth cotton and autoclave for 15 mnutes. Prepare slants by coolng vals n a dagonal poston. Slant cultures are preferred as they provde more surface area for the larvae and adults so that they are easer to observe and manpulate. Sprnkle chopped, sterlzed straw over the agar surface pror to the ntroducton of adults or feld-collected larvae to remove excess surface mosture and provde a sutable envronment for ovposton and larval feedng. Eggs generally hatch wthn 2-6 days. Larvae are subsequently fed a mxture of fnely chopped sterlzed straw and Brewer's yeast every 2-3 days. The cultures are mantaned n a constant temperature cabnet at 20 C. RESULTS Laboratory studes. Durng the past 5 years only 8 of the 28 laboratory cultures of C. hawaenss attempted were successfully carred through all stages and of these, only one yelded an F 2 generaton. The F 3 generaton ded out. Other Hawaan Scar-

3 1973 Steffan: Ecologcal studes of Ctenoscara hawaenss (Hardy) 87 dae have been reared and mantaned easly on the same medum. The dffculty n establshng cultures of C. hawaenss may be due to the habtat preferences of the larvae and ovpostng females. Larvae are found commonly under the bark of dead branches of Acaca koa and several other natve trees, ether fallen trees or broken branches on the ground. The larvae are most often n branches that stll have the bark farly tght aganst the wood wth lttle organc debrs. Ths habtat undoubtedly protects them from predators, parastes, and pathogens that would be abundant n the more normal scard larval habtat,. e., loosely packed decayng organc substances. The female prefers to ovpost n deep crevces or under the bark. Straw was apparently not a sutable substrate snce gravd females frequently de wthout ovpostng. Larvae tended to burrow nto the agar whereas most other Hawaan scard larvae reman on the surface. Duraton of the lfe cycle of C. hawaenss n the laboratory ranged from days from ovposton to adult emergence. The preovposton perod was usually 3 days. The ncubaton perod ranged from 4-5 days. Duraton of the 4 larval nstars ranged from days ana the pupal perod ranged from 4-6 days. Ths speces s apparently unsexual,.e., the progeny from one female are ether all males or all females. Development does not appear to be synchronous as adults from one batch of eggs emerged over a 10-day perod. Both n the laboratory and n the feld, the larvae tended to be scattered throughout the substrate, contrastng wth the gregarous behavor of some other speces. Seasonal fluctuatons, Klauea Ste : Numbers of adult Ctenoscara hawaenss collected n the Malase trap were relatvely low throughout 1971, rangng from 0 to 26 per 7- day collecton perod (Fg. 1). The greatest adult actvty occurred between January 25 and Aprl 26 wth two peak perods between January 25-March 1 and March 8-Aprl 26. The numbers trapped subsequently were relatvely low, ncreasng to a hgh of 20 durng the week of August 30-September 6. There was also a hgh of 13 durng the week of September 27-October 4 and agan durng the week of November Seasonal fluctuatons, Mauna Loa Ste : Numbers of adults collected showed very pronounced fluctuaton (Fg. 2). The frst and greatest ncrease began durng the week of January 25-February 1 clmbng from 13 to 131. The followng week, 258 adults were captured. After a decrease to 219 durng the week of February 1, the catch rose to 281, the hghest of the year. Durng the next four weeks, the numbers of adults collected dropped to 46. Two smlar cycles occurred n md-aprl and early June but wth a lesser magntude. Durng the summer dry season, the catch dropped to a weekly low of 7. After July 5, the collectons were made on alternate weeks only and ths undoubtedly affected the results. Three addtonal cycles were observed n early September, md-october, and late December. Comparson of Klauea and Mauna Loa populatons : C. hawaenss adults were always more abundant at the Mauna Loa ste than at Klauea (Fg. 3). Durng the 55-week perod, 3,084 adults were trapped at Mauna Loa versus 353 at Klauea. At the perod of peak adult actvty durng the week of February 15-22, 281 adults were trapped at the Mauna Loa ste versus 23 at Klauea. The two populatons approached each other n relatve abundance only durng the dry seasons.

4 Pacfc Insects Vol. 15, no. 1 UJ > > H- ELATI UMID % or x LU a: 3 \- < Q: O UIO Q- S LU s cr 100 z LU o LU a. oo cr LU ao JAN APR JULY OCT JAN TIME IN WEEKS Fg. 1. Fluctuatons of a Ctenoscara hawaenss populaton relatve to ranfall, temperature and relatve humdty, Klauea Forest Reserve. DISCUSSION Seasonal fluctuatons, Mauna Loa Ste: Ranfall appears to be the most mportant extrnsc factor affectng seasonal changes n C. hawaenss populatons. There were three cycles of adult actvty durng the frst half of the year. The frst cycle, January 25-March 15, began about three weeks after a perod of hgh ranfall n January, The followng week the populaton dropped slghtly, correspondng to a decrease n ranfall three weeks earler. The populaton reached a hgh dur-

5 1973 Steffan: Ecologcal studes of Ctenoscara hawaenss (Hardy) 89 >t IOO JS 60 S 50 <> "V. \/-\/ /N/W/S/ - '"' '-\A J W - \ -x/\/ A^\./..'V AV- ~.y v \ y\ /"\...A... V \ RAINFALL SPECIMENS J ^ (/) LL. O cr Ld CQ Jan apr July TIME IN WEEKS oct Jan Fg. 2. Fluctuatons of a Ctenoscara hawaenss populaton relatve to ranfall, temperature and relatve humdty, Mauna Loa Strp Road.

6 90 Pacfc Insects Vol. 15, no I ll >',1 «, l/l 1 Al ICA FMLAUtA MAUNA LOA 250! '».,III (M f) z LJ Q. to OC UJ CD O Z IOO :' > \ l». 1 \ ' I 1 I' 1 ' 1 ; / 1 '. j». I H ' h l '» H ; lt 50 A/ 1 ;V 1 ' KA->AA 0 5 JAN APR * '* * l-v^*-"" JULY / * I ' / / \ 1 \ / / V «. ' OCT JAN NUMBER OF WEEKS Fg. 3. Comparson of Ctenoscara hawaenss populaton fluctuatons, Mauna Loa Strp Road and Klauea Forest Reserve. ng the week of February when the catch rose to 281, the greatest of the year. Agan ths can be correlated to the hghest ranfall of the year whch occurred three weeks earler. The rapd decrease n adult actvty the followng week was preceded three weeks earler by a smlar but greater decrease n ranfall. The next two cycles n md-aprl and early June were almost dentcal but about half the magntude. Agan the fluctuatons could be correlated wth fluctuatons n ranall three weeks earler, the only dscrepancy beng n the last cycle. The frst peak occurred about 4 weeks after the week of peak ranfall. Ranfall then decreased sharply and remaned at a relatvely low level. However, the actvty of the adult C. hawaenss after droppng only slghtly, ncreased sharply. Snce the populaton of C. hawa-

7 1973 Steffan: Ecologcal studes of Ctenoscara hawaenss (Hardy) 91 enss was stll farly hgh, the relatvely low level of ranfall probably was suffcent to provde another perod of favorable growth. Ths speces s very senstve to ranfall fluctuatons. The usual three-week delay n response,.e., actvty of adults as expressed n Malase trap captures, possbly represents the developmental tme of C. hawaenss. Ths s about one week shorter than ts developmental tme n the laboratory, but the dffculty n colonzng ths speces ndcates that our laboratory condtons are far from optmal and ths may well be reflected n a lengthened developmental perod. There were agan three cycles of adult actvty durng the second half of the year. The frst and second were responses to ncreases n ranfall three weeks pror to each - KlLAUEA MAUNA LOA < 2 z JAN APR JULY OCT JAN TIME IN WEEKS Fg. 4. Comparson of weekly ranfall, Mauna Loa Strp Road and Klauea Forest Reserve.

8 92 Pacfc Insects Vol. 15, no. 5 peak. For the fnal cycle, the ntaton was later and the magntude was much less than expected. To some extent the fluctuatons may have been dampened by the alternate-week samplng method, but even takng that nto consderaton, the numbers of adult C. hawaenss were lower than expected. Fluctuatons of relatve humdty roughly corresponded to those of ranfall but the actvty of adults, as reflected n the Malase trap collectons, was not as closely correlated as t was to fluctuatons of ranfall. Snce most of the lfe of C. hawaenss s n a confned, protected area, the avalablty of mosture under the bark of dead koa branches would have been a more approprate ndcator of mosture avalablty. Temperature undoubtedly nfluences the development and actvty of C. hawaenss but s not the lmtng factor t would be n more temperate areas. A closer correla- -KlLAUEA MAUNA LOA 20 ce \- LJJO 15 CL LJ JAN APR JULY OCT JAN TIME IN WEEKS Fg. 5. Comparson of mean weekly temperatures, Mauna Loa Strp Road and Klauea Forest Reserve.

9 1973 Steffan: Ecologcal studes of Ctenoscara hawaenss (Hardy) 93 ton of adult actvty would be expected f the temperatures of the mcroenvronment under the bark were recorded. Seasonal fluctuatons, Klauea Forest Reserve Ste : Fluctuatons of C. hawaenss populatons were about 1/10 the magntude of those at the Mauna Loa ste ; however, they dd reflect smlar responses. Ranfall, relatve humdty, and mean weekly temperatures were generally greater at the Klauea ste. It therefore appears that some factors were suppressng the populaton of C. hawaenss at Klauea. The dfferences n the magntude of the flucutatons are shown n Fg. 3. Comparson of Klauea and Mauna Loa populatons : As explaned earler, the two stes, a Z> X UJ CC IO 15 APR OCT JAN TIME IN WEEKS Fg. 6. Comparson of mean weekly relatve humdty, Mauna Loa Strp Road and Klauea Forest Reserve.

10 94 Pacfc Insects Vol. 15, no. 1 although at the same elevaton and only a short dstance apart, are very dfferent ecologcally. Klauea s a ran forest and usually wet, whle Mauna Loa s a mountan parkland ecosystem wth more dry perods. Comparson of ranfall and relatve humdty between the two stes (Fgs. 4 and 6) reveals that both are usually hgher at the Klauea ste. Scardae reared n the laboratory are very susceptble to excess mosture although ths aspect has not been nvestgated quanttatvely. Excess mosture may favor the spread of epzootcs of fungal or bacteral dsease. Mean weekly temperatures are usually lower at the Mauna Loa ste (Fg. 5) and ths may ndcate that lower temperatures are more favorable for development of C. hawaenss. Avalablty of food and shelter may be an mportant factor n the dfferences n the relatve abundance of C. hawaenss at the two stes. C. hawaenss larvae are usually found under the bark of fallen koa trees and branches and seem to prefer branches where the bark s stll tghtly aganst the wood. Ths probably provdes them some protecton from predators and parastes and provdes a more sutable mcroclmatc nche. Fallen koa trees wth a great number of dead branches are very typcal of the Mauna Loa ste, but they are not promnent at the Klauea ste. LITERATURE CITED Gresstt, J. L. & M. K. Gresstt An mproved Malase trap. Pacf. Ins. 4: Hardy, D. E New Hawaan Scardae (Dptera). Proc. Haw. Ent. Soc. 16: Dptera: Nematocera-Brachycera. Vol p., 120 fgs. In: Zmmerman, E. C, ed. Insects of Hawa (q. v.). Honolulu. Jullet, J. A A comparson of four types of traps used for capturng flyng nsects. Canad. J. Zool. 41: Smth-Stockng, H Genetc studes on selectve segregaton of chromosomes n Scara coprophla Lntner. Genetcs 21: Steffan, W. A A generc revson of the famly Scardae of Amerca North of Mexco. Unv. Calf. Publ. Ent. 44: 1-77, 22 fgs.