Fragmentation and Drought Legacy Correlate with Distribution of Burrhead Chub in Subtropical Streams of North America

Abstract

Burrhead Chub Macrhybopsis marconis is a species of special concern endemic to subtropical streams in south Texas. We documented life history attributes and historical patterns in abundance and distribution to aid in understanding range‐wide declines of the species among the Colorado and Guadalupe‐San Antonio River systems. Life history results suggest Burrhead Chub lives for 2 years, reaches sexual maturity at age‐1, and spawns multiple clutches during March–September. Review of museum vouchers revealed Burrhead Chub is now missing from 26% of its historical range, including most impounded stream segments. There was a significant difference in the size of stream fragments for which Burrhead Chub is missing or still persists. Timing of many extirpations occurred after fragmentation and coincided with a record drought (1949–1959) when range‐wide stream flows were exceedingly low during the Burrhead Chub reproductive season. Furthermore, extensive historical collections taken during 1950–1961 in the Guadalupe River revealed a decline in relative abundance in upstream reaches as the drought continued. Based on these data, we hypothesize that Burrhead Chub experienced range‐wide declines associated with historical drought conditions in desiccated stream fragments. Following the drought, stream flows recovered but recolonization by Burrhead Chub was blocked by impoundments in many fragments. These findings support previous linkages between stream connectivity, flow magnitude, and the persistence of fish populations while providing insight into the mechanisms driving freshwater fish declines in the plains of North America.

Fragmentation and dewatering transform Great Plains stream fish communities

“…Sizing nodes using longitudinal lengths of fragments as well as codes denoting patterns in stream desiccations represented the application of graphics characterized by heterogeneous node resolutions (ErTs et al 2012). Similar approaches have been used to illustrate the network-scale spatial distribution and connectivity of essential habitats to enhance conservation of fishes in and outside of the Great Plains (ErTs et al 2011, Perkin et al 2013b. In this study, graphics illustrated the positive relationship between fragment length and water availability in a spatially explicit context.…”

“…In cases where there were multiple large tributaries, we included the length associated with the longest tributary when calculating fragment length. Although fragment network length (including uninhabited small tributaries) might also be used as a measure of habitat connectivity (e.g., Perkin et al 2013b), we relied on maximum longitudinal length to facilitate comparisons with recent studies reporting that fragmentation disrupts longitudinal connectivity and is critical for persistence of some fishes Urbanczyk 2013, Worthington et al 2014a). In each fragment, we selected USGS stream flow gages located nearest to the center of the fragment (measured using hydrologic distance; rkm) and with daily stream flow data encompassing the entire period between 1970 and 2013.…”

“…We selected 39 stream fragments distributed among the five basins and including all large-order (e.g., greater than fourth order; Strahler 1957, Perkin et al 2013b) stream segments historically inhabited by pelagic-spawning fishes. Fragments were defined by barriers included in the NABD as well as natural upstream limits of fish distributions as defined in previous work (Perkin and Gido 2011) and were limited to the area downstream of the western edge of the High Plains Aquifer (where groundwater depletion begins).…”

“…To begin, we developed patch-based graphs for each basin characterized by heterogeneous node (habitat patches) and link (connectivity among patches) resolutions (ErTs et al 2012). We defined nodes as fragments of stream between barriers and excluding impoundment water (n ¼ 39 fragments; Cote et al 2009), and sized nodes in proportion to the longitudinal length of each fragment (Perkin et al 2013b). We then constructed stream networks by inserting links between immediately adjacent nodes so that links constituted some form of barrier fragmenting the abutting nodes.…”

“…In rivers most affected by these alterations, homogenization is driven by the loss of a reproductive guild of fishes characterized by production of eggs and larvae that develop as they passively drift down the river continuum Bestgen 1997, Hoagstrom and. Members of this ''pelagic-broadcast spawning'' reproductive guild decline in association with fragmentation of riverscapes (Dudley and Platania 2007, Perkin and Gido 2011, Wilde and Urbanczyk 2013, reduced discharge magnitude (Cross et al 1985, Wilde and Durham 2008, Durham and Wilde 2009, and especially when both factors are combined (Perkin et al 2013b, Worthington et al 2014a. Though changes in distributions of benthic and pelagic fishes occur together in time, it is unclear if declining native fishes and expanding nonnative and native-invasive fishes are responding similarly to underlying environmental alterations (Douglas et al 1994.…”