A pseudo time‐series reveals the rapid recovery and high variability of benthic macroinvertebrate populations following catchment wildfire

Abstract

Abstract The increasing risk of wildfire has focused attention on the timescale of the impact and recovery of river ecosystems and methods for their bioassessment. An 18‐year pseudo time‐series was exploited to document patterns in benthic macroinvertebrate impact and recovery and evaluate the efficacy of alternative metrics to assess fire damage. Macroinvertebrates were surveyed by kick‐sampling and data were collected on river habitats. Details of river catchments and wildfire were collated as a GIS database. Macroinvertebrate richness and abundance recovered rapidly, marked by a phase of dynamic increase, followed by relative stability (0–2 years and 3–18 years, respectively). Across sites, richness and abundance were best explained by time since fire. A biotic index of general river quality was ineffective as an indicator of fire damage. While a metric of K‐selected taxa (Odonata richness) was generally indicative of fire‐affected assemblages, a contrasting metric of r‐selected taxa (percentage of chironomids, baetids, and simuliids) was not. Ordination analysis revealed time as a significant determinant of community structure across sites; however, its overall statistical importance was eclipsed by habitat characteristics (water quality, shade, altitude, and latitude) that were associated with ecological variation across both recently affected sites and the putatively recovered communities. These results highlight the stochastic processes – environmental and ecological – that frame the macroinvertebrate response to wildfire. This probabilistic context emphasizes the difficulties of developing indicator taxa for wildfire bioassessment and reinforces the importance of standardized survey protocols and the use of contrasting metrics in the assessment of wildfire impact on the ecological quality of rivers.