Our Liviiiii RfsoiitVi's — Hiiwiiii 



.Ul 



Population lieiids are suggested by analysis 

 of guilds of picture-wing tlies that breed on spe- 

 cific host plants. One example is the increase in 

 the relative frequency of observations of D. 

 sproati, a species that appears to breed exclu- 

 sively in rotting bark of one of the most com- 

 mon trees in this rain forest, the endemic mem- 

 ber of the Alalia family. Olapa {ClicinHlciulrun 

 trigyniiw). In contrast, a long-term decline is 

 evident in the guild of four species that breed 

 primarily in rotting bark of native lobelioids in 

 the genus CIcnnontia. There has been a concor- 

 dant reduction in frequency of all four species 

 over the last two decades and two of the four 

 species are now missing from this site (Fig. 4). 



There is historical evidence that the decline 

 in the latter group is a consequence of the 

 reduction of host plant populations. For exam- 

 ple, an important host of the two picture-wings 

 that appear locally extinct. Clenmmtia luurai- 

 ic'iisis. has been extirpated from at least one 

 nearby forest with a long history of disturbance 

 by feral pigs and cattle. 



Another factor may have been the invasion of 

 Olaa Forest by alien western yellowjackets 

 {Vt'spiila pensylvanica) in the early I980"s. The 

 wasps have become dominant predators of other 

 insects and may have contributed to the decline 

 of picture-wing Drosophila by feeding on larvae 

 that are particularly exposed on Clcnuoiuia 

 (Carson 1986; Foote and Carson, unpublished 

 data). These and other potential causative agents 

 of the changes in the community structure of 

 Drosophila need further investigation. 



Drosophila as a Focal Taxon To Monitor 

 Ecosystem Change 



Because of the extensive data that exist on 

 the population genetics and evolutionary rela- 

 tionships within the picture-wing Drosophila. 

 the potential consequences of disruption of the 

 Drosophila community can be examined at sev- 

 eral levels. For instance, the loss of a local pop- 

 ulation of Drosophila silvestris (one of the four 

 species that breeds in Clermontia) occurs at the 

 lower end of an altitudinal cline in inversion fre- 

 quencies among populations that extend above 

 Kilauea Volcano (Carson et al. 1990). A contin- 

 ued decline of D. silvestris populations at lower 

 elevations has the potential to change inversion 

 frequencies. Such long-term data may prove 

 useful in evaluating why different populations 

 or species may not respond similarly to green- 

 house stresses associated with global climate 

 change (Hoffman and Blows 1993). 



The potential influence of species extinc- 

 tions of picture-wing Drosophila on specific 

 lineages of subgroups (Fig. 2) can also be eval- 

 uated. For example, the guild of lobelioid-asso- 

 ciated picture-wing tlies undergoing a decline 

 in Olaa Forest is made up of species from three 



separate lineages of Hawaiian Drosophila. 

 Among fossil records of Hawaiian birds, an 

 analogous situation may have occuncd with the 

 extinction of several groups of flightless geese 

 and geeselike terrestrial herbivores (Olson and 

 James 1991 ). The monitoring of focal groups of 

 Hawaiian insects may well complement our 

 understanding gained from vertebrates on how 

 changes in Hawaiian ecosystems selectively 

 favor certain taxa that make up the contempo- 

 rary species diversity found in the islands. 



Lastly, data from active resource manage- 

 ment programs, such as the construction iif feral 

 pig exclosures at Hawaii Volcanoes National 

 Park, suggest that the population declines of 

 certain Hawaiian Drosophila are reversible, 

 even in the case of local extinction. Many host 

 populations are recovering in Olaa Forest fol- 

 lowing removal of pigs, and nearby populations 

 of many of the missing picture-wing species 

 persist. This is a rain forest that has twice been 

 devastated by volcanic eruptions and has recov- 

 ered. We are testing an old evolutionary tradi- 

 tion on these islands as we encourage the recol- 

 onization of this protected habitat by its former 

 occupants from nearby populations. 



The 30 years of intensive research on 

 Hawaiian Drosophila will not be readily repeat- 

 ed for even a small fraction of the remaining 

 species that make up the biological diversity pre- 

 sent in Hawaii. The fact that Hawaiian 

 Drosophila flies have received so much attention 

 is due in part to the fact that Drosophila are read- 

 ily observed and sampled (at baits) in native 

 forests. The sensitivity of such groups to a wide 

 range of human impacts needs to be evaluated. 

 Taxa, such as Hawaiian Drosophila. that can be 

 monitored in a cost-effective manner and yield 

 statistically reliable data need to be exploited as 

 potential indicators of the impact of environmen- 

 tal change on the vast majority of species about 

 which we know too little to manage intelligently. 



References 



Carson, H.L. 1967. Genetics and evolution of Hawaiian 



Drosopliilidae: preliminary report of collections made in 



Kipahulu Valley. Maui, cooperatively with the Kipahulu 



Valley Expedition, August, 1967. In R.E. Warner, ed. 



Scientific report of the Kipahulu Valley Expedition: 



Maui. Hawaii. 2 August-31 August. 1967. The Nature 



Conservancy, Honolulu. HI. 

 Carson. H.L. 1986. Drosophihi populations in the Ola'a 



tract. Hawaii Volcanoes National Park. 1971-1986. 



Proceedings Hawaii Volcanoes National Park Natural 



Science Conference 6:3-9. 

 Carson. H.L. 1992. Inversions in Hawaiian Dwsophihi. 



Pages 407-439 in C.B. Krimhas and J.R. Powell, eds. 



Drosophila inversion polymoiphism. CRC Press, Ann 



Arbor. MI. 

 Carson, H.L.. E.M. Craddock. W.E. Johnson. L.J. Newman. 



Y.K. Paik. W.W.M. Steiner. and K.C. Sung. 1981. 



Genetic studies of natural populations. Pages 438-470 in 



D, Mueller-Dombois, K.W. Bndges, and H.L. Carson. 



eds. Island ecosystems. Hutchinson-Ross, Stroudsburg. 



PA. 



— D. heteroneura 

 40 (presently missing) 



D. silvestris 



(presently missing) 

 D. murptiyi 

 D. setosimentum 



1971-72 1981-86 1992-93 



Survey period 



Fig. 4. Long-term trends within 

 one host-specific guild of picture- 

 wing Drosophila that breed in rot- 

 ting bark of native lobelioids in 

 the genus Clermontia. 



