Impacts of Small, Surface-Release Dams on Stream Temperature and Dissolved Oxygen in Massachusetts
Abstract
Dams fragment streams and rivers, with >14,000 in New England alone, and have the potential to significantly alter the physical, chemical, and biological characteristics of lotic systems. For example, dams can alter temperature and dissolved oxygen (DO) regimes, which can, in turn, affect species distributions, whole system metabolism, and nutrient processing rates. Moreover, changes in temperature signal life history cues (e.g., emergence, egg-hatching, migration) for many species of aquatic organisms, and present another avenue for dams to alter biotic communities. Despite the prevalence of small dams in the landscape and their potential significant impacts on temperature and DO, dams have not been well-studied and published impacts vary widely across sites. Given the variation in impact, I sought to quantify the impacts of small dams to stream temperature and DO, and to determine the drivers of inter- and intra-site variation in response. To accomplish this, I deployed 160 continuous temperature data loggers at 30 small, surface-release dams in Massachusetts. The majority of sites (61%) had higher temperatures downstream of the dam compared to upstream and most (85%) experienced decreasing temperatures with increasing distance downstream of the dam, such that the warmest temperatures were located closest to the dam. At approximately half of the temperature sites, flow had a homogenizing effect on temperatures throughout the study reach, whereby impacts were more pronounced (e.g., more warming, faster decay rates) under periods of low flow than under high flow conditions. Magnitude of warming varied greatly among sites, and this variation was explained best by landscape position and reservoir volume, with dams in smaller watersheds and with larger reservoir volumes experiencing greater warming magnitudes. Forest cover, dam height, and the presence of an auxiliary spillway best predicted the downstream temperature decay rate, with temperatures cooling fastest downstream of shorter dams in forested basins that did not have an auxiliary spillway. I used continuous DO loggers upstream, within the impoundment, and downstream of 12 dams to identify dam impacts to DO. Most sites experienced lower DO (66%) within the impoundment compared to upstream; however, 58% of the sites showed no difference in diel ranges between these reaches. The effect of dams on downstream DO was mixed, with increases, no change, and decreases relative to upstream condition; however, the majority of sites (58%) experienced a suppressed downstream diel range relative to upstream. The upstream slope, basin size, and dam height drove the impoundment response, such that dams with steeper upstream reach slopes, located in smaller basins, and with shorter dam heights experienced the greatest decreases in impoundment DO relative to upstream. Differences between downstream and upstream DO were best explained by upstream slope and impoundment volume, whereby sites with steeper upstream reaches and larger volumes of water within the impoundment experienced the largest decreases in downstream DO when compared to upstream reaches. These results may help managers prioritize dam removal at sites where a dam is having larger and more negative (e.g., elevated temperatures, decreased DO) impacts, and therefore where the greatest benefits should occur following restoration.