| Summary: | Forested catchments support an important ecological relationship between forests and streams. For example, forests filter nutrients and sediments from runoff and contribute terrestrial energy inputs (e.g. leaf litter) that support aquatic food webs. Consequently, disturbance from activities like forestry can affect the transport of matter into streams. A growing concern is the role of forestry in mobilizing mercury (Hg) from terrestrial sources into streams, which is poorly understood at catchment-sized scales. The objectives of my thesis were to evaluate the effect of forest management on spatial distributions of Hg in slimy sculpin (Cottus cognatus) within a catchment and between three catchments managed at different intensities in New Brunswick: minimally, extensively, and intensively harvested (NBR, NBE, and NBI, respectively). Alongside Hg analysis, stable isotopes of carbon, nitrogen, and hydrogen were measured to distinguish the relative contributions of terrestrial and aquatic energy to sculpin diet and trace the movement of Hg across trophic levels. During the fall of 2017, sculpin were sampled from five or six upstream to downstream reaches in each catchment. Where possible, 10 males and 10 females were selected for analyses (n = 287). Total Hg (THg) was measured in dried, homogenized whole body tissue (n = 258) using a Direct Mercury Analyzer, and δ13C, δ15N, and δ2H were measured in dried, homogenized muscle fillets (n = 245). Stable isotope ratios were used in Bayesian mixing models to predict the contribution of terrestrial and aquatic food sources to sculpin diet. Results showed that female sculpin in the minimally disturbed catchment had significantly lower THg levels compared to sculpin from NBE (p = 0.0141). The effect of stream order on THg concentration was significant in all three catchments (ANOVA, p < 0.01), but there were no clear spatial trends. Stable isotope results showed that variations in THg levels could be related to differences in diet, and results in one of the mixing models ...
|
|---|