Diatom responses to watershed development and potential moderating effects of near-stream forest and wetland cover Nathan J. Smucker 1 , Naomi E. Detenbeck 2 , AND Alisa C. Morrison 3 Office of Research and Development, National Health and Environmental Effects Research Laboratory, Atlantic Ecology Division, US Environmental Protection Agency, Narragansett, Rhode Island 02882 USA Abstract. Watershed development alters hydrology and delivers anthropogenic stressors to streams via pathways affected by impervious cover. We characterized relationships of diatom communities and metrics with upstream watershed % impervious cover (IC) and with riparian % forest and wetland cover in 120-m buffers along each side of upstream networks. Threshold Indicator Taxa ANalysis (TITAN) identified potential threshold responses of diatom communities at 0.6 and 2.9% IC. Boosted regression trees (BRTs) indicated potential thresholds between 0.7 and 4.5% IC at which relative abundances of low- nutrient diatoms decreased and those of high-nutrient, prostrate, and motile diatoms increased. These individual thresholds indicated that multiple stressors or magnitudes of stressors related to increasing watershed % IC differentially affected relative abundances of taxa, and these differential effects probably contributed to a more gradual, but still substantial, change in overall community structure. BRTs showed that near-stream buffers with .65% and ideally .80% forest and wetland cover were associated with a 13 to 34% reduction in the effects of watershed % IC on diatom metrics and community structure and with a 61 to 68% reduction in the effects of watershed % pasture on motile and high-P diatom relative abundances. Watershed % IC and riparian % forest and wetland cover probably affect hydrologic, nutrient, and sediment regimes, which then affect diatom community physiognomy and taxa sensitive to nutrients and conductivity. Our results emphasize the importance of implementing mindful development and protective measures, especially in watersheds near watershed % IC thresholds. Effects of development potentially could be reduced by restoring and conserving near-stream forests and wetlands, but management and restoration strategies that extend beyond near-stream buffers are needed. Key words: streams, urban, impervious cover, riparian buffer, threshold, management practices, nutrients, metrics, algae, periphyton, boosted regression trees, TITAN. More than half of the global population and 80.8% of the USA population resides in cities (UNPD 2006). Future increases in population will further drive development across the landscape and increase demands on and impacts to freshwater ecosystems and areas extending well beyond urban limits (Vo ¨ro ¨smarty et al. 2010, McDonald et al. 2011). Urban land cover makes up a small percentage of the landscape, but it impairs ,100,000 km of monitored streams and is the 2 nd -leading cause of impairment in the USA (USEPA 2002). Watershed development affects the biology, habitat, and ecosystem processes of streams by altering hydrology and delivering anthropogenic stressors via stormwater, failing infrastructure (e.g., leaky sewers and septic tanks), and subsurface flow, all of which are associated with the extent of impervious cover (IC; surfaces imper- meable to water, such as roads and buildings) (Paul and Meyer 2001, Walsh et al. 2005b, Kaushal and Belt 2012). IC is a useful metric because it can be quantified, managed, regulated, and even traded (Schueler et al. 2009). In watershed studies, IC has become an often- used proxy for the complex mixture and temporal variability of stressors attributable to development intensity that are difficult and costly to monitor (Brabec et al. 2002, Moore and Palmer 2005, King et al. 2011). IC reduces evapotranspiration and percolation of rain into soils and ground water, thereby negatively affecting stream condition by altering hydrology and facilitating greater loads of nutrients, sediment, and pollutants (Paul and Meyer 2001, Walsh et al. 2005b). Relationships between IC 1 E-mail addresses: smucker.nathan@epa.gov 2 detenbeck.naomi@epa.gov 3 Student contractor. E-mail: morrison.alisa@epa.gov Freshwater Science, 2013, 32(1):230–249 ’ 2013 by The Society for Freshwater Science DOI: 10.1899/11-171.1 Published online: 15 January 2013 230