housing for many species. When used, enrichment should elicit species-appropriate behaviors and be evaluated for safety and utility.

Generally, schooling fish species are housed with conspecifics, and many amphibians, especially anuran species, may be group housed. Aggression in aquatic animals does occur (van de Nieuwegiessen et al. 2008; Speedie and Gerlai 2008) and, as for terrestrial animals, appropriate monitoring and intervention may be necessary (Matthews et al. 2002; Torreilles and Green 2007). Some species need appropriate substrate (e.g., gravel) to reproduce or need substrate variety to express basic behaviors and maintain health (Overstreet et al. 2000). Improved breeding success in enriched environments has been reported but further research in this area is needed (Carfagnini et al. 2009). For many species (including, e.g., X. laevis), visual barriers, hides, and shading are appropriate (Alworth and Vasquez 2009; Torreilles and Green 2007). Most semiaquatic reptiles spend some time on land (basking, feeding, digesting, and ovipositing) and terrestrial areas should be provided as appropriate.

Animals used in aquaculture are often housed in situations that mimic agricultural rearing and may be in outdoor and/or sheltered raceways, ponds, or pens with high population densities. In these settings, where natural predation and mortalities occur, it may be appropriate to measure animal “numbers” by using standard aquaculture techniques such as final production biomass (Borski and Hodson 2003).

Space recommendations and housing density vary extensively with the species, age/size of the animals, life support system, and type of research (Browne et al. 2003; Green 2009; Gresens 2004; Hilken et al. 1995; Matthews et al. 2002). In the United States, for example, adult zebrafish (Danio rerio) in typical biomedical research settings are generally housed 5 adult fish per liter of water (Matthews et al. 2002), but this housing density varies when breeding and for housing younger animals (Matthews et al. 2002). This guidance is not necessarily relevant for other species of fish, and may change as research advances (Lawrence 2007). X. laevis adults may be housed at 2 liters of water per frog (NRC 1974), but a wide variety of housing systems are currently used in research settings (Green 2009). Institutions, investigators, and IACUC members should evaluate the appropriate needs of each species during program evaluations and facility inspections and continue to review ongoing research in these areas.