Environ Biol Fish (2010) 88:143–152 DOI 10.1007/s10641-010-9624-y Relationships among habitat, ecomorphology and diets of cichlids in the Bladen River, Belize Jennifer L. Cochran-Biederman & Kirk O. Winemiller Received: 7 February 2009 /Accepted: 4 February 2010 /Published online: 2 March 2010 # Springer Science+Business Media B.V. 2010 Abstract The Neotropical Cichlidae is among the niche segregation reflects selective establishment of most species-rich and ecologically diverse groups of species from a more species-rich regional species pool freshwater fishes. This study investigated interspecific rather than in situ adaptive evolution. morphological and ecological relationships within an assemblage of six cichlids in the Upper Bladen River, Keywords Cichlidae . Central America . Neotropics . Belize. This portion of the river drains a nearly Niche relationships . Ordination . pristine watershed within a nature reserve, and thus Resource partitioning . Species assemblage should provide a natural ecological context for interpretation of ecological patterns. Species distribu- tions within morphological, habitat and dietary space Introduction yielded patterns consistent with a hypothesis of niche partitioning. Statistical analyses of the species assem- With an estimated 600 species, the Neotropical Cichlidae blage revealed relationships between two principal is among the most species-rich and ecologically diverse morphological gradients from multivariate analysis groups of freshwater fishes (Farias et al. 1999; López- with several diet and habitat variables, and these Fernández et al. 2005). Local cichlid assemblages that patterns were consistent with prior functional mor- contain from one to over 20 species occur in river phological interpretations. Given that this local cichlid basins from Texas to Argentina and provide numerous assemblage contains no congeneric species, it is opportunities to study community ecology, adaptive apparent that morphological divergence resulting in radiation and biogeography. The subclade Heroini includes about 30 genera and 150 species distributed in South and Central America, with a single genus in J. L. Cochran-Biederman : K. O. Winemiller Cuba and Hispañola (Stiassny 1991;Hulseyetal. Section of Ecology, Evolution and Systematic Biology, 2004). The Heroini is perhaps the most ecologically Department of Wildlife and Fisheries Sciences, diverse clade of Neotropical cichlids, with certain Texas A&M University, species that specialize on specific resources ranging College Station, TX 77843-2258, USA from algae and detritus to insects, crustaceans or fish Present Address: (Eaton 1943; Winemiller et al. 1995; Valtierra-Vega and J. L. Cochran-Biederman (*) Schmitter-Soto 2000; Waltzek and Wainwright 2003). Department of Biology, Winona State University, In addition to having diverse diets, Central American 215 Pasteur Hall, Winona, MN 55987, USA cichlids also reveal great variation in morphology, e-mail: [email protected] reproductive behavior and habitat use (Winemiller et 144 Environ Biol Fish (2010) 88:143–152 al. 1995;Hulseyetal.2004; Soria-Barreto and Rodiles- Methods Hernández 2008). Evolutionary relationships within the Neotropical Study area Cichlidae have been investigated (Stiassny 1991; Kullander 2003; Hulsey et al. 2004; López-Fernández The field study was conducted during December 2006 et al. 2005), and a few studies have examined morpho- and January 2007 in the mainstem of the Upper logical relationships within an evolutionary framework Bladen River within the southeast portion of the (Norton and Brainerd 1993; Winemiller et al. 1995). Bladen Nature Reserve in the Toledo District of Fishes reveal general and robust relationships between southern Belize (16°34′N, 88°43′W, ca. 45 m eleva- form and function involving many easily measured tion). The headwaters of the Bladen River originate traits, and this makes the comparative study of within the Bladen Reserve and are a part of the morphology useful for inferring ecological patterns Monkey River Basin. This system drains the south- among fishes (Wainwright 1991; Motta et al. 1995; eastern slope of the Maya Mountains and contains Norton 1995; Ruber and Adams 2001). Winemiller three branches (including the Trio and Swasey rivers), et al. (1995) concluded that local fluvial assemblages making it the fourth largest river basin in Belize of Central American cichlids show greater diversifica- (1,275 km2). From the headwaters through the sam- tion of functionally significant morphological traits pling area, the Bladen River lies within a nature reserve than comparable assemblages of African cichlids. A protecting the biodiversity of a landscape covered by possible reason for more extensive adaptive radiation dense and essentially undisturbed rainforest. among Central American cichlids may be an evolu- tionary history with lower competition or predation Fish sampling from species from other freshwater fish orders that are rich in species, such as the Characiformes and Fishes were collected from a range of mesohabitats Siluriformes in South America and Africa (Winemiller including riffles, runs, deep pools, vegetated areas and 1991a, 1995). connecting tributaries. Fish capture methods included In addition to having species with specialized cast net, seine, hook-and-line and dip net. Sampling morphologies, particularly with regard to feeding was performed in each habitat at intervals throughout (Eaton 1943; Barel 1983; Hulsey and García de León the day and evening. Because changes in fish diet and 2005), species assemblages of Central American morphology depend on body size and stage of cichlids reveal evidence of dietary and habitat development, we summarized and analyzed data for partitioning (Winemiller and Pianka 1990; Winemiller individuals of each species that corresponded to the et al. 1995; Soria-Barreto and Rodiles-Hernández adult size interval. Fish specimens were euthanized in 2008). Nonetheless, cichlid species with extreme MS-222 then preserved in 10% formalin and later morphologies sometimes behave as ecological gen- transferred to 75% ethanol for storage. Specimens eralists. Liem (1980, 1991) proposed that morpholog- measured for the morphological analysis were also ical specialization permits more efficient foraging for examined for dietary analysis. Astatheros robertsoni a subset of food resources during periods of intensi- was not collected during our 2006–2007 survey, but fied competition. Others have proposed that morpho- specimens were observed and a few were collected logical specializations are shaped by optimal foraging within the study reach by the second author during in response to functional tradeoffs associated with December 2005–January 2006. These specimens were food acquisition and processing, irrespective of used for comparative morphological and dietary competition (Robinson and Wilson 1998). The present analyses. study examines associations among morphology, diet and habitat use within a natural assemblage of cichlid Habitat sampling fishes in a Central American stream. The field study was performed during the annual dry season when Several physicochemical variables were investigated stream discharge is relatively low and stable, popula- at each site where fish were observed or collected. tion densities are greatest, and species interactions are Temperature (°C), dissolved oxygen concentration most likely to influence community patterns. (mg·L−1), and conductivity (µs) were measured with Environ Biol Fish (2010) 88:143–152 145 a YSI model 85 meter, and pH was determined with a Snorkeling at each point was done for approximately handheld electronic pH meter. These variables 5 min. The observer visually identified and counted revealed low variation and were not considered in individuals of all fish species and one person on shore the statistical analyses. Maximum water depth and recorded data and time. After each point was snor- channel width were measured with a weighted keled, 5 min was allowed to pass before snorkeling the measuring tape. Flow velocity (m·s−1) was measured next point in order to minimize the effect of fish by timing a neutrally buoyant object drifting across a disturbance. In addition, during both habitat sampling predetermined distance. and snorkeling surveys, all movement between tran- In addition to physiochemical variables, other sects was done on shore to prevent disturbance as well. habitat characteristics were recorded at each site After all habitat and snorkel data were compiled, a where fish were collected, including presence of vegetation index, substrate index, and structural index aquatic vegetation, types and density of instream were calculated for each transect point and collection structures (e.g., large rocks, coarse woody debris), site. Each index was calculated by dividing the total and substrate composition. Substrate categories were number of categories observed at the location or point bedrock, silt/clay, sand, mud rich in organic matter, by the maximum observed categories. The indices leaf litter, woody debris, gravel (<3 cm diameter), were used to calculate an overall index of habitat cobble (4–10 cm diameter), rocks (11–25 cm diame- complexity. The following equation was used to ter), and bedrock (>26 cm diameter). calculate the index of habitat complexity: Habitat data were collected at each site where fish Vegetation index þ Substrate index þ Structural index were collected, however, these data were also collect- ed along ten randomly selected transects along