13Environmental change is expected to increase the frequency and severity of problems 14 caused by harmful algal blooms. We investigated the ecology of phytoplankton blooms in UK 15 canals to determine the environmental predictors and spatial structure of bloom 16 communities. The results revealed a significant increase in bloom presence with increasing 17 elevation. As predicted, higher temperatures were associated with a greater probability of 18 blooms, but the relationship between temperature and bloom occurrence changed across 19 landscapes. At the minimum level of agricultural land, the probability of bloom presence 20 increased with increasing temperature. Conversely, at the maximum level, the probability 21 decreased with increasing temperature. This pattern could be due to higher temperatures 22 increasing phytoplankton growth rates despite lower nutrient concentrations at low levels of 23 agricultural land, and nutrient depletion by rapidly growing blooms at high levels of 24 agricultural land and temperatures. Community composition exhibited spatial autocorrelation: 25 nearby blooms were more similar than distant blooms. Hydrological distances through the 26 canal network showed a stronger association with community dissimilarity than Euclidean 27 distances, suggesting a role for hydrological connectivity in driving bloom formation and 28 composition. This new knowledge regarding canal phytoplankton bloom origin and ecology 29 could help inform measures to inhibit bloom formation. 30 31 Keywords: algal bloom, cyanobacteria, climate, land use, health, connectivity, canal 32 hypotheses using a novel data source which arises from a bloom reporting system in 131 operation in England. 132 133 METHODS 134 Land use data, including patterns of natural, agricultural, and urban land, were obtained from 135 the Land Cover Map (LCM) 2007