Estimating Fish Species Richness across Multiple Watersheds

Abstract

Assessing fish species richness at the scale of an entire watershed or multiple watersheds is important when designing conservation areas and maintaining aquatic biodiversity. Estimating biodiversity at this scale requires considering the effects of habitat heterogeneity within and across drainages on the species-area relationship (SAR). I examined the SAR using unusually complete data to assess fish species richness in minimally disturbed watersheds on large public lands in the Sand Hills ecoregion, southeastern United States of America (USA). My objectives were to compare (1) true richness with estimates produced by different species richness estimators and sampling designs and (2) species richness among reservations. Accurate estimates were obtained for five contiguous watersheds (780 km2 total) by using Chao 2 or first-order jackknife estimators, coupled with (1) a stratified design that apportioned sampling effort over 25 sample sites based on major spatial correlates of assemblage composition, including stream size and drainage basin identity and (2) sufficient sampling effort to collect enough individuals to include rare species. The greatest species richness was in streams within a large land holding characterized by greater instream habitat diversity, less disturbed land coverage, more forested land, and closer proximity to source pools than other reservations. Species richness in these streams was within the range observed in high diversity Neotropical and Indomalayan realms.

Keywords

• stream fish
• species richness
• drainage basins
• species-area relationship
• species richness estimators
• sampling effort
• sampling design
• Sand Hills ecoregion
• USA