Effects of Red Shiner (cyprinella Lutrensis) on Red River Pupfish (cyprinodon Rubrofluviatilis )

A proposed desalinization project that would affect the entire upper Red River basin, Oklahoma-Texas, would, if completed, raise the possibility of invasion by red shiner, Cyprinella lutrensis, into stream reaches from which it has been restricted by naturally high salinity. Because introduced red shiners are known to impact native fish communities, we asked if red shiner could have an adverse effect on Red River pupfish, Cyprinodon rubrofluviatilis, which typically occurs in reaches of the Red River with high salinity. In an array of artificial streams we examined effects of red shiner on survival, condition, and reproduction of Red River pupfish. In the presence of red shiner, pupfish successfully produced larvae, but fewer juvenile pupfish survived to potentially recruit. RESUMEN-Un proyecto propuesto de desalinizaci6n que afectaria a toda la cuenca del Red River superior, Oklahoma-Texas, aumentaria, si es completado, la posibilidad de una invasi6n de la sardinita roja, Cyprinella lutrensis, en riachuelos en donde no ha estado por una salinidad naturalmente alta. Preguntamos si la sardinita roja podria occasionar un efecto negativo sobre el pez cachorito del Red River, Cyprinodon rubrofluviatilis, ya que la sardinita roja introducida suele impactar las communidades naturales de peces. Cyprinodon rubrofluviatilis tipicamente se encuentra en las cuencas del Red River con alta salinidad. Examinamos los efectos de la sardinita roja en la supervivencia, condici6n y reproduci6n del pez cachorito del Red River en una serie de riachuelos artificiales. En presencia de la sardinita roja, el pez cachorito produjo larvas con 6xito pero menos juveniles del pez cachorito sobrevivieron para su posible reproducci6n. The parapatric distribution of species along environmental gradients can reflect a trade-off between the competitive ability of a species and its ability to withstand environmental conditions (e.g., Connell, 1961). Abundance patterns of stream fishes are known to vary along environmental gradients, and patterns of little or no overlap between species have been interpreted as the outcome of competitive interactions (e.g., Taylor and Lienesch, 1996). However, it is often difficult to determine causal factors and to separate relative effects of various biotic and abiotic factors without experimental studies. Human activities can inadvertently affect fish distributions by changing environmental conditions. For example, impoundment of rivers has dramatically affected natural temperature and chemical gradients (Baxter, 1977). In addition, changes in regional temperature regimes in response to global warming may disrupt thermal gradients and thus the distribution of fish species in streams (Matthews and Zimmerman, 1990; Rahel et al., 1996). The impetus for the present study is the proposed desalinization of a naturally saline river to improve water quality for municipal and agricultural purposes, a process that could alter distribution of fishes along a salinity gradient. The Red River (Oklahoma-Texas) is characterized by high salt concentrations (>10 ppt) in the headwaters due to percolation of ground water through marine salt beds (Echelle et al., 1972). Salinity decreases downstream where freshwater tributaries enter the river. Fish assemblages in the headwaters typiTHE SOUTHWESTERN NATURALIST 44(3):287-295 SEPTEMBER 1999 The Southwestern Naturalist cally are depauperate (<10 species/collection), with Red River pupfish, Cyprinodon rubrofluviatilis, and plains killifish, Fundulus zebrinus, being the dominant species (Echelle et al., 1972). In lower reaches, red shiner, Cyprinella lutrensis, western mosquitofish, Gambusia affinis, and numerous other native minnows are abundant (Echelle et al., 1972; Taylor et al., 1993). Although pupfish can tolerate low salinity (Echelle et al., 1972), they are thought to be poor competitors and to exist in high abundance only in areas where high salt concentrations limit entry of other native species (>10 ppt; Echelle et al., 1972). Increasing demand for water in Oklahoma and Texas has led to projects that reduce salinity of the Red River. Two desalinization structures already exist in the upper Red River basin and seven others have been proposed by the United States Army Corps of Engineers as part of the Red River Chloride Control Project. Currently, the Red River is still highly saline; however, completion of additional desalinization plants could cause a basin-wide reduction in salinity. Under this scenario, it is likely that other species that are abundant downstream under lower salt concentrations could invade the upper reaches of the basin. In particular, red shiner has been shown to rapidly invade disturbed areas and has had negative effects on native fish species (Karp and Tyus, 1990; Rinne, 1991; Douglas et al., 1995). Given the potential for increased interaction between red shiner and Red River pupfish if desalinization were to occur, we examined interactions between these species to assess direct effects of the former upon the latter. METHODS-Experimental Streams-Experimental streams were located at the University of Oklahoma Biological Station and consisted of eight independent riffle-pool units (Fig. 1). Each unit was filled with natural sand, gravel, and cobble. Substratum was sculptured to form concave pools and riffles with sinuous flow to simulate a natural stream. In each unit, water was pumped from a downstream collection box using a submersible 0.25 hp (horse power) pump into an elevated holding tank where water overflowed into an upstream collection box. Upstream and downstream boxes were closed with 0.5 cm plastic mesh to block movement of adult fish into collection boxes. Current velocity within each unit was heterogeneous, ranging from 0 to 25 cm/ sec in riffles and 0 to 10 cm/sec in pools (Gelwick, Elevated Holding Tank Pool Depth 45cm