Content area
Full text
INTRODUCTION
An emerging view in community ecology is that local assemblage organization is constrained not only by local processes, but also by larger scale environmental factors and available species pool (Ricklefs 1987, Roughgarden 1989, Menge and Olson 1990, Ricklefs and Schluter 1993). This implies that when ecological generalizations are sought at regional scales, information on abiotic constraining factors and historical processes (including biogeography) must be explicitly included to understand any observed patterns, though local processes or habitat constraints may also need to be invoked to explain residual variation in the regional pattern (Duarte 1991).
Hydrological regime is a significant constraint on lotic assemblage structure. Extremes of flow and patterns of flow variability can directly influence local community structure, as has been demonstrated by a number of studies for both fish (Horwitz 1978, Meffe 1984, Coon 1987, Bain et al. 1988, Jowett and Duncan 1990, Fausch and Bramblett 1991) and invertebrates (reviewed in Poff and Ward 1989, Fisher and Grimm 1991). For individual fish species, spates cause direct mortality for both juveniles (e.g., Seegrist and Gard 1972, Schlosser 1985, Harvey 1987) and adults (e.g., Harrell 1978, Toth et al. 1982), and the timing of high flows may serve as an environmental cue for spawning (John 1963, Nesler et al. 1988). As the importance of physical disturbance to streams becomes increasingly recognized (Resh et al. 1988), interest has grown in testing the hypothesis that significant variation in assemblage structure among streams is explicable in terms of hydrological patterns, which can vary substantially over even short geographic distances (Poff and Ward 1989, Biggs et al. 1990, Poff and Allan 1993).
In attempting to identify environmental determinants of assemblage structure across geographic domains, assemblages and environmental factors should be viewed at comparable scales (see O'Neill et al. 1986, Wiens 1989, Levin 1992). Margalef (1968) pointed out that large-scale patterns are best detected with coarse-grain data. For ecological communities relative abundance data provide fine-grain information because they emphasize local peaks in species performance, while species presence/absence data emphasize a coarser grain of environmental tolerance (Allen and Starr 1982, cf. Rahel 1990). In streams, availability of long-term discharge data allows great flexibility in describing environmental variability in ecologically relevant terms (Poff and Ward 1989). Both fine-grained and coarse-grained hydrological descriptors can...