Temporal variability of stream macroinvertebrate abundance and biomass following pesticide disturbance

We determined the extent of macroinvertebrate recovery in a former pesticide-treated stream (FTS) relative to a reference stream (RS) by examining macroinvertebrates colonizing red maple (Acer rubrum L.) litter bags between 5 to 10 y following pesticide treatment. Mean abundance and biomass, variability in abundance and biomass (using the coefficient of variation [CV]), and assemblage structure were compared both within and among years to assess recovery. The 5 y of study included 3 drought years followed by 2 wet years. Mean total abundance and biomass of macroinvertebrates, and that of most functional feeding groups (FFG) did not significantly differ between streams during this study, nor did within-year variability of these means, indicating macroinvertebrates in FTS had recovered relative to RS. Some exceptions to the above patterns (shredder abundance and mean annual gatherer biomass) resulted from the dynamics of a single taxon in each group (Leuctm and Paraleptophlebia, respectively). Macroinvertebrate assemblage structure in litter bags was similar between streams throughout the study as shown by their similar ordination scores; hence, assemblage structure had also recovered. In each stream, mean annual abundance and biomass of total macroinvertebrates and of each FFG, aside from shredder abundance, differed significantly among years. However, assemblage structure was generally similar among years. Among-year CVs were usually lower than within-year CVs because macroinvertebrate abundance and biomass fluctuated more during a year than it did from year to year, and different processes apparently contributed to the variation observed at these 2 time scales. For example, juvenile development time influenced within-year CVs, indicating that life-history characteristics affected temporal variability of macroinvertebrate abundance and biomass. Examination of both the means and their variances was useful for determining the extent of recovery and how macroinvertebrates responded to natural environmental variability. The detailed analysis of temporal dynamics at different time scales afforded by the CV supported our contention that FTS had recovered from the pesticide application relative to RS. Key uwrds: long-term recovery, within-year variability, among-year variability, coefficient of variation, life history, cohort production interval. The ability of stream macroinvertebrate asNiemi 1990, Mackay 1992, Milner 1994). For exsemblages to recover from disturbance depends ample, species that have high vagility or that on many factors, including 1) life-history charsurvive disturbances can recolonize disturbed acteristics of organisms, 2) timing of the disturareas rapidly, often in a stochastic manner largebance, 3) presence of survivors within disturbed ly dependent on the timing of the disturbance sites, 4) distance of source populations for rerelative to their life-cycle stage (e.g., Gray and colonization, and 5) nature of the disturbance Fisher 1981, Fisher et al. 1982, Wallace et al. (Gore and Milner 1990, Wallace 1990, Yount and i986/ whiles and Wallace 1992, 1995). Subsequent changes in a recovering macroinverte, , . , , , u ,u ^ i A brate assemblage may result from further recol1 E-mail address: [email protected] & J 2 Present address: Department of Agricultural Biolonization and biotic interactions, but also deogy, Kangweon National University, Chuncheon, 200pend on the natural disturbance regime of the 701, South Korea area (Fisher 1983).