Date of Thesis

Spring 2023


Land transformation from natural to agriculturally dominated landscapes has many negative impacts on surrounding terrestrial and aquatic habitats. Specifically, conversion of land from forest to agricultural use, combined with farming and livestock practices, creates complex physical stressors on stream communities. The stream experiences geomorphic changes following riparian zone removal, such as increased surface runoff, narrowed stream channels, sediment transport, and erosion. Without canopy coverage, these streams are exposed to increased light exposure and solar radiation, leading to overall higher water temperatures.

Without a riparian zone, streams contain very little, large woody debris (LWD) and often lack recruitment of new wood to the channel due to lack of woody riparian vegetation, which often alters the biological community found in the stream, such as shredders, collectors, and subsequently, fish. Much of what is understood about lotic systems comes from studies that use small-scale, easy to access locations in just a portion of a whole watershed. Habitats and biological communities studied at local scales are influenced by processes operating at both the reach-scale and watershed-scale, but studies of responses at fine scales do not distinguish between these groups of processes, therefore providing at best an incomplete understanding of the whole watershed. To understand how to mitigate impacts of agriculture on streams and aquatic biota, a reach-scale and watershed-scale approach is needed.

This study aims to evaluate reach-scale and catchment-scale effects of four land cover types (agriculture, forest, forest buffer, and planted buffer) within predominantly agricultural watersheds in Centre and Lycoming counties. Criteria for selecting research sites included having restoration practices (planted riparian buffers) and active agriculture. Final watershed selection was made after ground truthing. Each stream was then segmented into reaches categorized by land cover type (agriculture, forest, planted riparian buffer, forest buffer) and a HOBO temperature logger was deployed at the top and bottom of each reach. HOBOs were deployed from May 2022 to October 2022, and temperature was recorded once every hour for the duration of deployment. Temperature data were used to calculate average temperature per week, average daily maximum temperature, and average daily minimum temperature at each site. Upper and lower sites were established in each stream to assess monthly water quality and fish community structure. 100-meter single pass electrofishing was conducted at each site, accompanied by habitat surveys, and discharge calculations. Fish communities were analyzed using the Thermal Fish Index of Pennsylvania (TFI). Observed maximum weekly temperature data from 17 July 2022 and National Land Cover Data attributes (NLCD) were used to train a Spatial_Stream_Network (SSN) model. This model was used to eliminate spatial dependence of upstream HOBOs and create predicted temperature values for unsampled reaches.

Agricultural reaches consistently and considerably increased stream temperature by 2.17 ± 3.25 °C/km. Forested reaches decreased temperature by 0.18 ± 0.65 °C/km, forest buffer increased reach temperature by 0.18 ± 0.40 °C/km, and planted buffers increased temperature by 0.13 ± 1.1 °C/km. Fish assemblages reflected changes that were expected based on the maximum reach temperature. Buffalo Run had the lowest TFI score, indicative of cold-water temperatures, and Halfmoon Creek had the highest, indicative of warm-water temperatures. The SSN temperature model identified several land cover types that have significant impact on maximum stream temperature, including 42 (Evergreen Forest), 43 (Mixed Forested), 71 (Grassland), and 82 (Cultivated Crops).

This study was conducted to quantify the influence of riparian zone vegetation (reach-scale) and land use variables (watershed-scale) on stream temperature in multiple stream networks with agricultural influence and to determine how stream thermal regimes impact fish communities. Overall, streams with more agricultural reaches had higher temperatures and fish communities containing higher proportions of warm-water species. Maintaining and restoring forested riparian buffers could prove to be an effective way to reduce stream warming in agricultural streams with corresponding benefits to stream biota.


Fish, Streams, Agriculture, Riparian Buffer, Temperature, Central Pennsylvania

Access Type

Masters Thesis (Bucknell Access Only)

Degree Type

Master of Science



First Advisor

Matthew McTammany

Available for download on Tuesday, May 06, 2025