Temporal change in urban fish biodiversity-Gains, losses, and drivers of change

Abstract

Our aim was to examine temporal change in alpha and beta diversity of freshwater fish communities in rivers that have urbanized over the same period to understand the influence of changes in land use and river connectivity on community change. We used biological (2001-2018), land use (2000-2015), and connectivity data (1987-2017) from Toronto, Ontario, Canada. We used linear mixed effects models to determine the strength of upstream land use, connectivity, and their changes over time to explain temporal change in alpha and beta diversity indices. We examined beta diversity using the temporal beta diversity index (TBI) to assess site-specific community change. The TBI was partitioned into gains and losses, and species-specific changes in abundance were assessed using paired t-tests. There were more gains than losses across the study sites as measured by TBI. We found little to no significant differences in species-specific abundances at aggregated spatial scales (study region, watershed, stream order). We found different relationships between landscape and connectivity variables with the biodiversity indices tested; however, almost all estimated confidence intervals overlapped with zero and had low goodness-of-fit. More fish biodiversity gains than losses were found across the study region, as measured by TBI. We found TBI to be a useful indicator of change as it identifies key sites to further investigate. We found two high value TBI sites gained non-native species, and one site shifted from a cool-water to warm-water species dominated community, both of which have management implications. Upstream catchment land use and connectivity had poor explanatory power for change in the measured biodiversity indices. Ultimately, such spatial-temporal datasets are invaluable and can reveal trends in biodiversity useful for environmental management when considering competing interests involved with urban sprawl in the ongoing "Decade on Restoration."

Keywords: fish community ecology; freshwater ecology; temporal biodiversity; urban ecology; urban rivers; watersheds.

FIGURE 1

Map of study sites across Etobicoke Creek, Mimico Creek, Highland Creek, the Don River, and the Rouge River. Black points represent study sites. Blue lines represent river or creek watercourse.

FIGURE 2

B‐C plot for the temporal beta diversity index (TBI). Squares represent sites where gains > losses. Circles represent sites where gains < losses. Colors denote watersheds. The location of the centroid line is above the 1–1 line, suggesting the system reports more gains overall than losses. Asterisks represent high TBI sites (prior to holm correction p < .05 sites). Size of points represents relative TBI values.

FIGURE 3

Map of sites with the temporal beta diversity index (TBI) results. Size represents relative TBI value. Plus sign sites have more gains than losses; circle points represent more losses than gains. Red sites represent high TBI sites (pre‐Holm correction significant sites). Gray polygons represent urban intensification. Green polygons represent woodland loss. Purple polygons represent wetland loss. Site identification numbers increase in the upstream direction.

FIGURE 4

Model‐averaged estimates for richness, Shannon diversity, Pielou's evenness, and temporal beta diversity index (TBI) with 95% confidence intervals. The TBI intercept value of 0.62 with 95% CI [0.57, 0.67] not shown due to data visualization limitations. All estimates represent subset estimates.