Intra-annual Variation in Fish Communities and Habitat Associations in a Chihuahua Desert Reach of the Rio Grande / Rio Bravo Del Norte

—Anthropogenic alterations to lotic systems are often implicated in global declines among stream-dwelling fishes. A primary step toward mediating fish declines is improving our understanding of species–environment relationships; unfortunately, such information is limited. The goal of this study was to assess relationships between fish communities and environmental variables in a relatively intact portion of the Rio Grande / Rio Bravo del Norte in the southwestern United States and northern Mexico. We quantified the relationship between riverine habitats and local fish communities at 7 sites in the Big Bend reach of the Rio Grande / Rio Bravo del Norte during 2006. Monthly collections yielded 10,565 individuals representing 20 species, including 6 species listed as threatened or otherwise at risk. Species richness varied among sites (range 13–19) as did diversity (1 – D: 0.49–0.72) and density (16–45 fish ⋅ 100 m–2). Fish community composition indicated spatial and temporal variations, but habitat characteristics indicated more variation among sampling sites than among months. Spatial variation in community structure correlated with site-specific habitat characteristics, and most threatened or at risk species were associated with run or riffle geomorphic units containing higher current velocities and gravel to cobble substrates. Temporal variation in community structure correlated with fall monsoonal flooding, during which composition of dominant species shifted from the habitat generalist red shiner (Cyprinella lutrensis) to the imperiled, regionally endemic Tamaulipas shiner (Notropis braytoni). Results from this study suggest that large flood pulses and maintenance of habitat heterogeneity are necessary for the persistence of both declining and intact local fish communities in the Rio Grande / Rio Bravo del Norte. RESUMEN.—Las alteraciones antropogénicas a los sistemas lóticos están con frecuencia implicadas en las declinaciones globales entre peces de arroyos. Un paso primordial para mediar la disminución de los peces, es mejorar nuestro entendimiento de las relaciones entre las especies y su ambiente; desafortunadamente, dicha información es limitada. El objetivo de este estudio fue evaluar las relaciones entre las comunidades de peces y las variables ambientales en una parte relativamente intacta del Río Grande / Río Bravo del Norte al suroeste de los Estados Unidos y al norte de México. Cuantificamos la relación entre los hábitats ribereños y las comunidades locales de peces en 7 sitios del tramo Big Bend del Río Grande / Río Bravo del Norte durante el 2006. Las colecciones mensuales aportaron 10,565 individuos representando a 20 especies, incluyendo a 6 especies listadas como amenazadas en riesgo. La riqueza de especies varió entre sitios (con un rango de 13–19) al igual que la diversidad (1 – D: 0.49–0.72) y la densidad (16–45 peces ⋅ 100 m–2). La composición de la comunidad de peces indicó variaciones espaciales y temporales, pero las características del hábitat indicaron más variaciones entre los sitios de muestreo que entre los meses. La variación espacial en la estructura de la comunidad estuvo correlacionada con las características del hábitat específicas al sitio, y la mayoría de las especies amenazadas o en riesgo estuvieron asociadas con unidades geomórficas de correderas o rápidos con corrientes de velocidades mayores y sustratos de grava o guijón. La variación temporal en la estructura de la comunidad estuvo correlacionada con las inundaciones otoñales del monzón, durante las cuales la especie dominante cambió de la generalista carpita rojiazul (Cyprinella lutrensis), a la especie en peligro y regionalmente endémica sardinita pechona de Tamaulipas (Notropis braytoni). Los resultados de este estudio sugieren que los pulsos de inundación grandes y el mantenimiento de la heterogeneidad del hábitat son necesarios para la sobrevivencia de las comunidades locales de peces, tanto en declinación como las intactas, en el río Bravo del Norte. fishes are considered extirpated (Hughes et al. 2005); and in the Rio Grande / Rio Bravo del Norte Basin (hereafter, Rio Grande), 43% of native species are missing from all or portions of their historical range (Calamusso et al. 2005). Declines in abun dance and distribution of native fishes are often associated with human-mediated changes (e.g., dam construction, water diversion) to naturally occurring fluvial conditions (Holden 1979, Wins ton et al. 1991, Hoagstrom et al. 2008), but specific mechanisms and causal factors of spe cies declines remain largely unknown. Conservation approaches in which entire communities rather than specific species are considered hold the greatest potential for identifying and subsequently mediating factors contributing to the decline and extirpation of multiple organisms (Scott and Helfman 2001, Bunn and Arthington 2002). Within the Rio Grande drainage of the southwestern United States and northern Mexico, numerous historically widespread and abun dant native fishes are now less common and occupy a fraction of their historical ranges (Hubbs et al. 1977, Bestgen and Platania 1991, Edwards and Contreras-Balderas 1991, Ed wards et al. 2002, Calamusso et al. 2005, Hoag strom et al. 2008). For example, Rio Grande silvery minnow (Hy bognathus amarus), speckled chub (Macrhybop sis aestivalis), Rio Grande shiner (Notropis jemezanus), and blue sucker (Cycleptus elongatus) are extirpated from numerous reaches (An derson et al. 1995, Platania and Altenbach 1998); and phantom shiner (Notropis orca) and Rio Grande bluntnose shiner (Notropis simus simus) are extinct (Bestgen and Platania 1990, Hubbs et al. 2008). De clines in abundance and distribution of some Rio Grande fishes coincided with increases in habitat generalist and lentic-adapted species such as red shiner (Cyprinella lutrensis), fathead minnow (Pimephales promelas), gizzard shad (Dorosoma cepedianum), and western mos quito fish (Gambusia affinis) after stream alteration throughout the Rio Grande watershed (Calamusso et al. 2005). However, de spite well-documented changes in species abundances and distributions, mechanisms contributing to the decrease of some native fishes and increase of other native or nonnative fishes remain un clear. Additional research is needed to establish the relationship between the ecology of local fish communities and the anthropogenic alterations occurring throughout the basin. Ichthyofaunal changes in the Rio Grande are associated with anthropogenic activities that modify instream habitat and cause concern for the long-term persistence of endemic taxa. Construction of reservoirs, channel straightening, water diversions for agricultural purposes, re duced stream-flow quantity, de graded water quality, and introduction of nonnative species are common in the Rio Grande (Edwards et al. 2002, Calamusso et al. 2005). In some reaches, the mainstem river is completely dewatered because of agricultural and municipal withdrawals (Calamusso et al. 2005); and in other reaches, poor water quality and chemical pollution prompt advisories against water contact and fish consumption (Owens and Niemeyer 2006). Stream fragmentation is thought to be related to the extirpation of fishes belonging to the pelagic-spawning reproductive guild, and only a few remaining populations persist in relatively long fragments (>100 km in length; Dudley and Platania 2007). Such extensive altera tions call into question the feasibility of conserving or restoring intact fish communities within highly modified mainstem reaches (Calamusso et al. 2005). Less impacted river reaches, including the Big Bend National Park reach of the Rio Grande, provide the opportunity to assess the ecology of relatively intact local fish communities (Moring 2002) or, more specifically, to assess habitat associations and suitability for declining or imperiled species (Helfman 2007). Knowledge of habitat associations and suitability is a primary step toward conserving imperiled species (Bond and Lake 2003). This knowledge provides targets for habitat conservation and restoration (Lake et al. 2007) and is necessary for ecoregions such as the Rio Grande where empirical data for instream flow requirements are lacking (Sansom 2008). The goal of this study was to quantify intraannual variation in the habitat ecology of local fish communities in the Big Bend reach of the Rio Grande. We specifically sought to assess occurrence and habitat associations of smallbodied species that are extirpated from Rio Grande reaches separated from the Big Bend region by impoundments and that occur in portions of New Mexico and the Rio Grande terminus near the Gulf of Mexico (e.g., Gido et al. 1997). Historical and contemporary litera ture 2 WESTERN NORTH AMERICAN NATURALIST [Volume 72