Multiscale drivers of phytoplankton communities in north‐temperate lakes

Abstract

Multiple factors operating across different spatial and temporal scales affect β‐diversity, the variation in community composition among sites. Disentangling the relative influence of co‐occurring ecological drivers over broad biogeographic gradients and time is critical to developing mechanistic understanding of community responses to natural environmental heterogeneity as well as predicting the effects of anthropogenic change. We partitioned taxonomic β‐diversity in phytoplankton communities across 75 north‐temperate lakes and reservoirs in Alberta, Canada, using data‐driven, spatially constrained null models to differentiate between spatially structured, spatially independent, and spuriously correlated associations with a suite of biologically relevant environmental variables. Phytoplankton β‐diversity was largely independent of space, indicating spatial processes (e.g., dispersal limitation) likely play a minor role in structuring communities at the regional scale. Our analysis also identified seasonal differences in the importance of environmental factors, suggesting a general shift toward greater relevance of local, in‐lake (e.g., nutrients and Secchi depth) over regional, atmospheric and catchment‐level (e.g., monthly solar radiation and grassland coverage) drivers as the open‐water growing season progressed. Several local and regional variables explained taxonomic variation jointly, reflecting climatic and land‐use linkages (e.g., air temperature and water column stability or pastureland and nutrient enrichment) that underscore the importance of understanding how phytoplankton communities integrate, and may serve as sentinels of, broader anthropogenic changes. We also discovered similar community composition in natural and constructed water bodies, demonstrating rapid filtering of regional species to match local environmental conditions in reservoirs comparable to those in natural habitats. Finally, certain factors related to human footprint (e.g., cropland development) explained the composition of bloom‐forming and/or toxic cyanobacteria more than the overall phytoplankton community, suggesting their heightened importance to integrated watershed management.

Quantifying seasonal succession of phytoplankton trait‐environment associations in human‐altered landscapes

“…Total phosphorous and nitrogen concentrations ranged from 0.0015 to 0.7500 and 0.0672 to 14.0210 mg/L, respectively, indicating oligotrophic to hypereutrophic status. Water ion chemistry and nutrient levels reflect both natural and land‐use factors (Loewen et al 2020), including extensive development for cropland, pastures, and canals/drainage (0–89, 0–50, 0–56% of catchments, respectively) among prairie grasslands in the southern portion of the Province. Logging (0–28% of catchments) is a more common disturbance among forest (0–81%) and wetland (0–32%) dominated catchments in northern and western areas of the study region.…”

“…Additional environmental factors were derived from remote sensing data to capture regional variation in land‐cover, climate, and basin and watershed morphometry. Detailed protocols for phytoplankton sampling and enumeration, water quality sampling, and geospatial processing are described in Supplementary Material (Appendix 1) of Loewen et al (2020). To limit the number of comparisons in our analysis, environmental variables were restricted to 31 candidates identified as being significant contributors to phytoplankton community composition from their prior analysis of the data (Table 1).…”

“…Waterbodies were sampled as part of the Alberta Government's lake monitoring program, in collaboration with the Alberta Lake Management Society's LakeWatch Program, and spanned large biogeographic and anthropogenic gradients covering prairie grassland, aspen parkland, boreal, foothills, and Rocky Mountain ecoregions. Descriptive statistics for environmental variables were previously presented by Loewen et al (2020) and are briefly summarized here. Sampling locations had maximum depths ranging from 0.5 to 50 m along a 1372 m elevation gradient, with mean catchment slopes ranging from 0.52 to 23.22 .…”

“…Phytoplankton occurrence was determined from composite water samples combining euphotic depth-integrated subsamples collected at 10 locations in each lake. Taxonomic identifications, based on diagnostic morphological characteristics, and cell enumerations were conducted by a professional taxonomist at Bio-Limno Research & Consulting, Inc. (Halifax, Nova Scotia) following the Utermohl's sedimentation method (Lund et al 1958 Loewen et al (2020). To limit the number of comparisons in our analysis, environmental variables were restricted to 31 candidates identified as being significant contributors to phytoplankton community composition from their prior analysis of the data (Table 1).…”

The Role of Environmental Processes and Geographic Distance in Regulating Local and Regionally Abundant and Rare Bacterioplankton in Lakes

“…The logistical constraints of sampling across the length and breadth of New Zealand meant that lakes were sampled in different seasons and years. Seasonal differences have previously been shown in planktonic communities (Loewen et al, 2020;Schallenberg et al, 2021) with the prevalence of local and regional level drivers changing (Loewen et al, 2020). However, we acknowledge that the sampling design meant that groups of lakes close together were sampled at the same period of time which could impact the assessment of the comparison between environmental processes and geographic distance.…”