Direct and Indirect Effects of Channel Catfish (Ictalurus punctatus) on Native Crayfishes (Carnbaridae) in Experimental Tanks

-FO~ the incised, sand-bed streams of northcentral Mississippi, USA, fish predation is one plausible mechanism to explain both relatively low crayfish densities and differences in stream size occupied by various native crayfishes. I conducted two mesocosm experiments to test effects of a fish predator (channel catfish, Ictalurus punctahls) on the survival and size structure of native crayfishes (primarily Pmcambams hayi and Orconectes chickasawae) in the presence and absence of shelter. I used predominantly the larger species, P. hayi, in the first experiment and the smaller species, 0. chickasawae, in the second. Experiments lasted 19-21 d, and each consisted of four replicated treatments: crayfish without shelter, crayfish with shelter, crayfish and predator without shelter, crayfish and predator with shelter. In both experiments, catfish significantly reduced crayfish S U M V ~ . Shelter significantly reduced catfish predation on the smaller, but not the larger, crayfish species. Comparisons between experiments showed that in tanks containing catfish, P. hayi had higher survival than 0. chickasawae. In both experiments, the mean size of crayfish increased less in the presence than in the absence of catfish, and I argue that the effect is due largely to a reduction in crayfish growth. Channel catfish directly and indirectly influenced crayfish in experimental settings, with the degree of influence varying by crayf~sh species and presumably related to crayfish size. Thus, fish predation and shelter availability are likely important factors influencing densities of and macrohabitat use by these native crayfishes. In the incised, sand-bed streams common in northern Mississippi, USA, habitat is typically uniform, and shelter for fish or crayfish in the form of woody debris, rocks, vegetation or undercut banks is relatively scarce (Shields et aL, 1994, 1995). During electrofishing in local streams, I captured crayfish predominantly near cover, even small patches of leaves or debris, and rarely over open sand. Abundant shelter can reduce fish predation rates on crayfkh (Garvey el al., 1994), and even gravel and pebble substrates afford more protection from predation than does sand (Stein and Magnuson, 1976). Given choices of substrate and habitat complexity, crayfish tend to avoid open sand substrates, the effect being most pronounced for small crayfish when fish predators are present (Stein and Magnuson, 1976; Hill and Lodge, 1994; Kershner and Lodge, 1995). Size-selective fish predation within and among crayfish species has been studied extensively in the upper midwestern U.S.A. for a suite of crayfishes including the native Orcone&s virilis and invaders 0. propinquus and 0. rusticus, and appears to facilitate species replacement by the invaders. Fishes selectively prey on small or medium sized crayfish and seem to select crayfish based primarily on chelae size (Stein, 1976; Garvey et al., 1994; Keller and Moore, 2000; Roth and Kitchell, 2005). Crayfish behavior also influences crayfish predation risk and tends to be size-dependent as well (Garvey et aL., 1994; Keller and Moore, ' ~ e l e ~ h o n e : (662) 2342744 ext. 267, e-mail: [email protected] FIG. 1.-Reproductive form males of Osconectes chickasawe (left) and Procambanw hayi (right) from Mississippi, USA 2000). However, relative predation risk has received little attention as a potential mechanism determining crayfish distributions in assemblages of native, co-evolved species. In headwater streams of the Little Tallahatchie River drainage, Mississippi, five species of native crayfishes are common (Cambarus stnatus, Orconectes chickasawae, Procambarns hayi, P. uuachitae and P. uioscai), but abundances are fairly low (e.g., 0.09 mp2 in perennial, fishbearing stream reaches, Adams and Warren, 2005). Procambarus hayi is common in lakes and ponds (Payne, 1972) and reaches its highest stream abundances in deep, slow-flowing stream sites close to lakes. Procambarus hayi is one of the larger local species (post-orbital carapace length [POL] up to 37 mm) and is characterized by very long chelae (pincers), especially on mature males (Fig. 1). The smallest of the local species, 0. chickasawae (POL up to 26.5 mm), is common in a variety of small-stream habitats, even in intermittent streams, and its densities are inversely correlated with stream size. The chelae of adult 0. chickasawae are much smaller than those of P. hayi (Fig. 1). A shortage of shelter may contribute to the low crayfish densities in sand-bed streams in northern Mississippi, U.S.A. Also, if fish prey preferentially on certain crayfish species, then predation pressure in conjunction with limited shelter may contribute to the differences in stream sizes used by some native crayfish species. I conducted two experiments to examine whether shelter availability affects channel catfish predation on crayfish (primarily P. hayi and 0. chickasawae) and whether catfish influence crayfish size structure. Comparisons between experiments provided insights into the relative direct and indirect effects of channel catfish on the two species. 1 I used channel catfish (Ictalurus punctatus) as the predator for the experiments because it is a common, native fish that sometimes forages extensively on crayfish (Flotemersch and Jackson, 2003). To ensure that channel catfish would be appropriate predators under the experimental conditions, I conducted a brief pilot study. I placed six Orconectes chickasawae crayfish and one channel catfish in each of two outdoor tanks (without gravel substrate), one with shelter and one without. In tanks without and with shelter, all crayfish were eaten by the first and third mornings, respectively. The objectives of the two main experiments were to test: (1) whether shelter for crayfish reduced channel catfish predation rates on cra@sh, (2) whether shelter influenced crafish cannibalism, (3) whether catfish influenced crayfish size structure and (4) whether catfish had a greater influence on the survival and size structure of Orconectes chickasawae compared to Procambarus hayi.