How a monitoring dataset, an adaptive management framework, and ecological comparisons of selected fish groups can guide conservation

Abstract

Freshwater habitats are amongst the most threatened systems globally, and stream and river biodiversity is extremely vulnerable to human and climate impacts. A challenge for research and management professionals seeking to conserve native, freshwater biodiversity is how to develop a process that links past and present data to guide future data collection, restoration actions, and management decisions. The purpose of this research is to illustrate how an adaptive management framework, applied to monitoring data for carefully chosen groups of fish, can guide conservation planning. Chapter 1 develops my team’s adaptive management framework and illustrates the use of the framework for one stable, native Kansas fish species. Chapter 2 demonstrates how the framework can be used to analyze monitoring data for three native Kansas fish species (two common and one uncommon). Our framework is comprised of an iterative 10-step cycle within which we embedded a 6-step, statistical subloop. Each iteration of this framework prioritizes a tractable question, identifies focused taxa and scales, uses a directed literature review to provide context, wrangles appropriate fish-relevant habitat variables, and applies data cleaning procedures. Then weight of evidence is accumulated by combining many visualization tools (i.e., fish maps, proportional resource maps, prediction maps, ridgeline plots, box plots, histograms, pie plots), multiple logistic regression, and probability plots. The final step of the 1st iteration identifies data gaps and testable predictions for future field sampling that will be analyzed in the 2nd iteration. In a proof of concept, my team compared data analysis of two common fish [Emerald Shiner (Notropis atherinoides), Central Stoneroller (Campostoma anomalum)] and one uncommon fish [Plains Minnow (Hybognathus placitus), Kansas threatened]. Data analysis of combined fish taxa, chosen based on a thoughtful, multi-criteria decision tree, enhanced conservation insights. Our multiple logistic regression models consistently identified priority regressors. Our weight of evidence approach clarified ambiguous regression trends. Prediction maps, paired with visualization tools, identified promising sites for future Plains Minnow restoration. Our approach proposes a continually evolving series of structured interactions among researcher-manager teams to accumulate actionable knowledge through a process of shared question identification, data analysis, and discussion of next steps. Monitoring data, research data, and data tests of management outcomes all have value for applied problem solving. However, if different types of data and different datasets are not connected and coordinated, opportunities for conservation success are lost. Our framework and proof of concept show a way to make these connections. This framework is an example of an implementable, adaptive management approach that can compare distributional patterns of thoughtfully chosen fish taxa to aid restoration efforts of threatened, freshwater systems.