Exploring Linkages Between Landscape Patterns and Freshwater and Estuarine Bivalves in the Coast Range of Oregon

Spatial configurations of landscape variables (biotic, abiotic, and socio-ecological) affect and are affected by ecological processes and species in watersheds. This dissertation explores relationships among landscape patterns, ecosystem processes and bivalve species dynamics in coastal watersheds in Oregon, USA. I approached this broad topic through two primary avenues of research: investigating cross-ecosystem threats from pesticide use in forestland management to downstream aquatic environments, and the landscape ecology of an at-risk freshwater mussel species.

Terrestrial land use activities present cross-ecosystem threats to riverine and marine species and processes. Specifically, pesticide runoff can disrupt hormonal, reproductive, and developmental processes in aquatic organisms, yet non-point source pollution is difficult to trace and quantify. In Oregon, state and federal forestry pesticide regulations, designed to meet regulatory water quality requirements, differ in buffer size and pesticides applied. To identify exposure and uptake of contaminants in coastal watersheds, I collected freshwater and estuarine bivalves Margaritifera falcata, Mya arenaria, and Crassostrea gigas from eight Oregon Coast watersheds to examine forestry-specific pesticide contamination. Additionally, during a 45 day period in the spring of 2019, I sampled sixteen coastal watersheds for current-use water-borne herbicides commonly used in forestland vegetation management. In 38% of bivalve samples, one or more of twelve unique pesticides were detected (two herbicides; three fungicides; and seven insecticides). Frequency and maximum concentrations varied by season, species, and watershed, with indaziflam (herbicide) the only current-use forestry pesticide detected. At 80% of sampling locations integrative passive water samplers detected at least one of four commonly used herbicides, with hexazinone and atrazine most commonly detected. An additive effects model using slope, herbicide activity notified during the sampling window, and recent clearcut harvest notifications predicted variation in total herbicide accumulation (R2=0.8914). The model was then applied to predict concentrations in un-sampled watersheds throughout Oregon's coastal region at three watershed scales using Hydrologic Unit Codes (HUCs) 8, 10, and 12. Details about types and levels of exposure provide insight into effectiveness of current forest management practices in controlling transport of forest-use pesticides at multiple scales.

Freshwater mussels have declined across the region following widespread degradation of freshwater habitat and other aquatic species, including parallel declines in salmonid species, which serve as host fish for larval western pearlshell mussels (Margaritifera falcata). M. falcata are native to Pacific coastal watersheds in Oregon and beyond, but their comparative distribution, habitat, host species interactions, and health have not been investigated in detail. To understand population dynamics of extant M. falcata in Oregon's small coastal watersheds, I analyzed a dataset of stream survey observations collected over a recent ten year period for presence/absence of mussels, explored reach-scale habitat characteristics in relation to persistence of populations, and summarized the current distribution of surveyed mussels and their co-occurrence with host fish species in coastal drainages. I also collected M. falcata at eight locations within Oregon's Coast Range and compared condition indices among sites. Overall naïve occupancy in surveyed areas was 12.3%, close to half of predicted occupancy (ψ = 0.24, CI= 0.19-0.31) based on modeling repeated visits over a ten year assumed closure period. Mussel occupancy was positively correlated with habitat variables (% of pool, count of boulders and stream temperature), providing new information about reach-scale habitat associations in Oregon's coastal watersheds. Using a host fish co-occurrence analysis, I found that probability of mussel observations was positively correlated with presence of coho (Oncorhynchus kisutch) and chinook (O. tshawytscha) salmon, and negatively associated with steelhead (O. kisutch) based on logistic regression. Condition varied significantly among mussel collection locations (n=8), and healthiest animals were found in areas draining small catchments. Spatial relationships between existing distribution, host species and habitat variables outlined in this study answer questions about coastal freshwater mussel populations in Oregon and identify "priority areas" for further research, conservation, and population assessment within this region.

Scully-Engelmeyer, Kaegan Michael
Product Number: 
Year of Publication: 
Portland State University. Earth, Environment, & Society Ph. D. Program