Behavioral Responses of the Endemic Shrimp Halocaridina Rubra

Behavioral Responses of the Endemic Shrimp Halocaridina rubra (Malacostraca: Atyidae) to an Introduced Fish, Gambusia affinis (Actinopterygii: Poeciliidae) and Implications for the Trophic Structure of Hawaiian Anchialine Ponds1 Krista A. Capps,2 Caroline B. Turner,2 Michael T. Booth,2 Danica L. Lombardozzi,2 Scott H. McArt,4 David Chai,3 and Nelson G. Hairston Jr.2,5 Abstract: In the Hawaiian Islands, intentionally introduced exotic fishes have been linked to changes in native biodiversity and community composition. In 1905, the mosquito fish Gambusia affinis was introduced to control mosquitoes. Subsequently, G. affinis spread throughout the Islands and into coastal anchialine ponds. Previous studies suggest that presence of invasive fishes in anchialine ponds may eliminate native species, including the endemic shrimp Halocaridina rubra. We examined effects of G. affinis on H. rubra populations in anchialine ponds on the Kona-Kohala coast of the island of Hawai‘i. In the presence of G. affinis, H. rubra exhibited a diel activity pattern that was not seen in fishless ponds. Shrimp in ponds with fish were active only at night. This pattern was ev- ident in anchialine ponds and in laboratory experiments. In laboratory predation experiments, G. affinis preferentially consumed smaller H. rubra, and in the field the H. rubra collected from invaded sites were larger than those from fishless ponds. Analysis of trophic position using stable isotope analyses showed that feeding of H. rubra was not significantly distinct from that of snails, assumed to feed at trophic level 2.0 on epilithic algae, but G. affinis was slightly omnivo- rous, feeding at tropic level 2.2. The mosquito fish diet was apparently com- posed primarily of algae when the defensive behavior of H. rubra made them substantially unavailable as prey. The effect of successful establishment of G. affinis on shrimp behavior has the potential to alter abundance of benthic algae and processing and recycling of nutrients in anchialine pond ecosystems. The introduction of nonnative species is and the introduction of nonnative species one of the greatest threats to biotic diversity will likely continue due to both intentional and ecosystem functioning throughout the and unintentional release by humans and in- world (Mack et al. 2000, Silliman and Bert- effective regulations (Crivelli 1995). Species ness 2004). Invasion events are typically invasions cause major environmental damage linked to human activities (Rejmanek 1996), and economic losses amounting to billions of dollars annually in the United States alone (Pimentel et al. 2005); therefore, the conse- 1 Manuscript accepted 1 April 2008. quences of invasion are of substantial interest 2 Department of Ecology and Evolutionary Biology, to scientists and concern to policy-makers Cornell University, Ithaca, New York 14853. and the general public (Vitousek et al. 1997). 3 Four Seasons Resort, 72-100 Ka‘u¯pu¯lehu Drive, Kailua-Kona, Hawai‘i 96740. Fishes have been intentionally introduced 4 Department of Entomology, Cornell University, throughout the world for sport, as ‘‘ornamen- Ithaca, New York 14853. tal’’ taxa, and for biological control of insect 5 Corresponding author (e-mail: [email protected]). species (Leyse et al. 2004). The U.S. Geolog- ical Survey (2006) has reported that a total of Pacific Science (2009), vol. 63, no. 1:27–37 653 species of fishes have been introduced : 2009 by University of Hawai‘i Press into aquatic habitats in the United States. In- All rights reserved troduced fishes have been linked to declines 27 28 PACIFIC SCIENCE . January 2009 in native invertebrate species, especially in chemistry in other systems (Hurlbert et al. previously fishless water bodies (Englund 1972, Hurlbert and Mulla 1981). In 1905, 1999, Townsend 2003, Leyse et al. 2004, Vi- 150 G. affinis were brought to the Hawaiian tule et al. 2006), and to changes in trophic Islands from Texas to promote mosquito con- relationships (Cardona 2006). For example, trol (Stearns 1983). Subsequently, G. affinis brown trout (Salmo truta) introductions in spread throughout the Islands and is now New Zealand have caused local fish and in- present in many anchialine ponds. vertebrate extinctions and changes in the The introduction of G. affinis to Hawaiian behavior of native fishes and macroinverte- anchialine ponds could directly affect native brates (Townsend 1996). shrimp populations in at least two ways: by As with other inland water habitats in the eliminating the H. rubra through consump- United States, fish introductions have been tion and by modifying H. rubra behavior, frequent in coastal Hawaiian anchialine making them less active during the day when ponds—small brackish-water ecosystems vulnerability to visual predation by the fish found in areas with porous basins, typically is greatest. Other studies have shown that of pumice (Brock and Kam 1997). These predators can affect their prey through con- ponds are hydrologically linked to both sumption or by instigating changes in prey groundwater sources and the ocean. Anchia- behavior and life history traits (Flecker 1992, line ponds, restricted in the United States to Allan 1995). For example, many zooplankton the Hawaiian Islands (Brock 1977), support in lakes migrate vertically during the day to endemic communities of native organisms, dark bottom waters to avoid predation by including two shrimp species, Halocaridina visually foraging fishes (Zaret and Suffern rubra (‘o¯pae‘ula) and Metabetaeus lohena. His- 1976), a behavior that is in many cases in- torically, these ponds were important to the duced by fish kairomones (Lampert and coastal fishing culture of native Hawaiians Loose 1992, De Meester et al. 1998). This and as aquaculture sites (Santos 2006). Cur- modification of behavior may incur costs in rently, populations of H. rubra and other terms of lost foraging time in illuminated sur- anchialine-pond species are threatened by face waters where algae are more abundant habitat destruction and the introduction of or in the metabolic cost of swimming (Loose nonnative predatory fishes and shrimps. Pre- and Dawidowicz 1994). vious research suggests that the presence of The direct effects of predator introduc- invasive fishes may eliminate H. rubra from tion, through the reduction or elimination ponds (Bailey-Brock and Brock 1993), but of prey and the alteration of prey behavior, the mechanism for their eradication has been can impact the trophic dynamics of aquatic unclear. In addition to being biologically im- ecosystems (Nystrom et al. 2001). Gambusia portant, the ‘o¯pae‘ula are culturally significant introductions would be expected to add a tro- to native Hawaiians and have become impor- phic link to naturally fishless Hawaiian an- tant in the aquarium trade (Santos 2006). chialine ponds that typically lack vertebrate One of the fish invaders most commonly predators of H. rubra (Brock and Kam 1997). found in anchailine ponds is the western mos- In fishless ponds, the food chain should have quito fish, Gambusia affinis (Brock and Kam just two trophic levels (and one trophic link): 1997, Englund 1999). These fish are diurnal epilithic algae consumed by ‘o¯pae‘ula. Ponds consumers and, although they have been in- with Gambusia might then be expected to troduced throughout the world to control have three trophic levels (two links): algae mosquito populations (Pyke 2005), they are consumed by ‘o¯pae‘ula, which are in turn also known to feed on other insects, algae, consumed by fish. However, if the ‘o¯pae‘ula crustaceans, and small vertebrates (Leyse were effective at evading predation by alter- et al. 2004, Pyke 2005). Introductions of ing their behavior, the ominivorous Gambusia mosquito fish have led to reductions in in- would be forced to feed in part or entirely on vertebrate populations, changes in phyto- benthic algae, and food chain length short- plankton abundance, and changes in aquatic ened to an intermediate position of between Effects of an Introduced Fish on an Endemic Shrimp . Capps et al. 29 one and two trophic links. Here, we report to differences in pond surface area, sampling the results of a study of Hawaiian anchialine methods differed among ponds. Ho‘onanea ponds to determine: (1) if G. affinis is capable and Waiiki were each sampled using a 30 cm of preying on H. rubra under laboratory wide net (straight, flat bottom edge) using 10 conditions; (2) if H. rubra persists in anchia- sweeps of 20 sec each. Because of its small line ponds after G. affinis introduction, and size, Wahi pana could only be sampled with whether its behavior changes so as to facili- a small net (10 cm bottom edge) using brief tate its persistence; and (3) the actual trophic (5 sec) sweeps moved along the bottom five position of G. affinis in ponds to which it has times. All sampling was carried out on flat been introduced. lava substrate containing small crevices and holes. In Wahi pana the lava substrate had materials and methods numerous sharp protrusions around which the sampling net had to be moved. To stan- Study Sites dardize for net size and sampling time, we report catch per unit effort (CPUE), defined as All sampling and collection of organisms the number of H. rubra caught per centime- were conducted in three anchialine ponds at ter of net width per minute. This measure, al- the Four Seasons Huala¯lai Resort on the though somewhat qualitative for comparisons Kona-Kohala coast of the island of Hawai‘i of shrimp density among ponds, provided a 2 (Figure 1). Two ponds, Ho‘onanea (270 m , quantitative estimate of diel changes in 2 50 cm average depth) and Waiiki (48 m ,25 shrimp abundance on the substrate surface cm average depth), contained large popula- or in the water column within each pond. tions of the introduced mosquito fish, Gam- In addition to diel changes in abundance busia affinis. Ho‘onanea also supported a patterns in the ponds, we assessed activity small population of introduced guppies, Poeci- patterns in the laboratory using experimental lia reticulata, and glass shrimps, Paleomonetes tanks.

Load more