Fall 2019

Document Type

Master's Thesis (Open Access)

Degree Name

Master of Science (M.S.)


Natural Sciences


As perennial river systems become increasingly intermittent worldwide, there is a growing need to focus attention on developing dry-phase bioassessment tools for non-perennial rivers and streams (NPRS). NPRS play key ecological roles in watershed functioning but can be impacted by many of the same disturbances affecting perennial rivers. However, we lack tools to assess the ecological health of NPRS like the well-developed bioassessment tools used for perennial systems. For this reason, there has been an increasing need to develop biological indicators to assess the health of rivers when they are dry.

We sampled 39 streams in Southern California, USA during the dry phase to assess the responses of terrestrial arthropod and bryophyte assemblages to human activities. We developed 230 metrics which characterized aspects of terrestrial arthropod assemblages (ground-dwelling and vegetation-dwelling) and bryophyte assemblages, and described the richness, abundance, taxonomic composition, diversity, and feeding groups (for arthropods) or growth forms (for bryophytes) of each assemblage. We accounted for metric bias associated with natural variation by adjusting metrics that were influenced by naturally occurring environmental gradients. We evaluated the ability of the 230 metrics to distinguish reference sites from non-reference sites impacted by human activity in the watershed (e.g., urbanization, agriculture, cattle grazing). Eight metrics had large responses (i.e., absolute t-statistic > 1.80) to human activity, and these metrics could potentially be used as biological indicators of dry stream ecological health. These eight metrics included representatives from each of the three indicator assemblages and included measures of taxonomic composition, richness, abundance and arthropod feeding groups. We then assessed the relationship between the responses of these eight metrics and several measures of human activity using quantile regression and found three of the eight metrics were significantly (p < 0.05) limited by human activity.

These results indicate that terrestrial assemblages may be used to distinguish reference from non-reference NPRS during the dry phase. These findings support the eventual integration of dry phase bioassement into river monitoring programs. However, conceptual models describing the causal mechanisms that drive the terrestrial responses to human activity are still needed to understand how these activities should be managed to protect dry river health. We also need to include the linkages between the indicator assemblages and ecosystem functions of NPRS in these conceptual models.