.vo 



Hiiwuu — Our Living Resources 



1,0- 



Picture-wings 



Other endemics 



Cosmopolitans 



Inside 

 exclosure 



Fig. 2. A conipanson of the a\'er- 

 age relative frequencies of 

 Hawaiian picture-wing tlies, other 

 endemic Dnisuplula. and recently 

 introduced cosmopolitan species 

 inside and outside of a 7-year-old 

 24()-ha (593 acre) fenced pig 

 exclosure in Hawaii Volcanoes 

 National Park. Frequencies are 

 based on observations at 837 bait 

 stations set up over five survey 

 periods in 1992 and 1993 along 

 four 2.400-m (7.875-ft) transects. 

 The data are e.xpressed as a per- 

 centage of total observations with- 

 in a survey period. 



silvestris above Kilauea Volcani) and lenects 

 the recolonization of habitat destroyed by two 

 explosive emptions within the last 2.100 years. 

 The surfaces of Kilauea Volcano are covered by 

 new lava flows at a rate of about 90*^ per 1 .000 

 years. The population biology of Hawaiian 

 Diosopliiki and other endemic insects has been 

 one of continual local extinction of and recolo- 

 nization by populations on single volcanoes 

 over thousands of years (Carson et ai. 1990). 



Dominance of Cosmopolitan Drosophila in 

 Disturbed Habitat 



The dominance of non-native plant and ani- 

 mal species in Hawaii associated with human 

 activity and the subsequent loss of endemic 

 species have long been recognized (Perkins 

 1913). Rain fores^ts above 1.000-m (3.280-ft) 

 elevation provide habitat for much of the 

 remaining native biota. Most of these wet 

 forests occur on the "Big Island" of Hawaii 

 where about 30% of the 5()().0()0 ha ( 1 .2 million 

 acres) of upland native woodland is rain forest 

 (Jacobi and Scott 1985). While this wet forest 

 vegetation is among the most intact in the state, 

 invasions by alien species have seriously 

 degraded components of the understory. 



Non-native ungulates (cattle, goats, pigs, 

 etc.) cause major problems. Feral pigs, in par- 

 ticular, feed upon and uproot native tree ferns, 

 shrubs, and herbs. They also actively consume 

 tleshy fruits of non-native plants and thereby 

 spread their seeds. As a consequence, feral pigs 

 help establish non-native plants that can perma- 

 nently alter native communities (Cuddihy and 

 Stone 1990: Stone et al. 1992). 



Twenty-five years ago fenced feral ungulate 

 exclosures were first tested by Hawaii 

 Volcanoes National Park as a method of pro- 

 moting natural wet forest restoration. In the 

 park, including approximately one-third of the 

 Olaa Forest, exclosures encompassing as much 

 as 800 ha ( 1,980 acres) are now used to manage 

 tracts of montane wet forest (Stone et al. 1992). 



When intensive research on Hawaii 

 Drosophila was initiated over 30 years ago. it 

 was apparent that habitats altered by human 

 activity had greatly reduced populations of 

 endemic Drosophila (Carson 1967). The con- 

 struction of large fenced feral pig exclosures 

 has provided an opportunity to measure changes 

 in Drosophila community composition associ- 

 ated with this one particular agent of distur- 

 bance in wet forests. A significant increase in 

 the frequency of cosmopolitan species of 

 Drosophila has been measured in wet forest 

 habitat disturbed by feral pigs and associated 

 non-native plants (Fig. 2). In areas with high pig 

 densities, many host plants for endemic 

 Drosophila are reduced to those few individuals 

 growing as epiphytes above the reach of pigs. 



In contrast, many alien plants that thrive in 

 pig-disturbed areas are species that produce 

 tleshy fruits eaten by the pigs. These fruit-bear- 

 ing non-native plants, such as banana poka 

 iPdssiflora iiiollissiiiia) and yellow Himalayan 

 raspbeny {Riihus t'llipticiis). also support large 

 populations of introduced cosmopolitan 

 Drosophila. such as D. iniinigrans. D. siimtlans. 

 and D. suzukii (known collectively to geneti- 

 cists as "yellow flies"), that breed primarily on 

 rotting fruit (Foote et al.. unpublished data). 



Long-term Changes Among Populations of 

 Picture-wing Drosophila 



Changes in the relative proportions of differ- 

 ent species of endemic picture-wing flies in 

 Olaa Forest between 1971 and 1993 are shown 

 in Fig. 3. These proportions are based upon 

 observations of individual picture-wings, total- 

 ing 1.222 in 1971, 1,467 in 1981, and 1,294 in 

 1992. A general decline in overall picture-wing 

 diversity is suggested by the observation that 4 

 species out of 14 were missing from one or 

 more of the more recent surveys. There has also 

 been a change in the relative abundances of 

 species within the group. For example, two of 

 the most common species of picture-wing flies 

 from the original survey, D. murphyi and D. 

 setosimentiwu are now much less common. 



15 - 



10 



,Jki- 



fcp ■Jj m 



J^ 



I 



t 



E 



Fig. 3. A comparison of the relative frequencies of 14 

 endemic species of picture-wing DrosophiUi over three 

 periods of observation from 1971 to 1993 In Olaa Rain 

 Forest in Hawaii Volcanoes National Park. The data are 

 expressed as a percentage of total observations within a 

 suney period. 



