Seasonal Community Occupancy of Fishes in a Lowland (Arkansas, <scp>USA</scp> ) Headwater Stream

ABSTRACT Among lowland fish communities, there is limited knowledge of the importance of headwaters. This information is needed to understand their role in fish communities. We addressed this gap by assessing seasonal changes in fish communities in lowland headwater streams (Arkansas, USA). 16 sites were sampled seasonally (winter, spring, summer and fall) in four events from Jan. to Oct. 2023. Multiple sampling gears (backpack electrofishing, kick nets and seines) were used to increase species detection. Fish presence/absence data, site and survey covariates were analysed using multiseason, community occupancy models to estimate μ , a value between 0 and 1, which represents the likelihood that an individual species is a member of the community and essentially represents that species occupancy (species probability), Ω , a value between 0 and 1, which represents the average occupancy probability across all species in the community across sites (community occupancy mean), and p , a value between 0 and 1, which is the probability of observing a species at a site given that the species is present (detection probability). A total of 37 species from 10 families were observed. Of the many models, we explored the best three: μ (canopy cover), Ω (canopy cover), p (.); μ (catchment area), Ω (catchment area), p (stream flow); μ (catchment area), Ω (catchment area) and p (.). The model incorporating canopy cover suggests that μ varied by species, Ω (± SE) varied by season and that p was constant. Overall, seasonal variation exists at the species and community levels; canopy cover and catchment area are the main drivers of μ and Ω and stream flow is the main driver of p . This research aids in the understanding of fish patterns and species richness on a temporal scale, improving knowledge of lowland headwater stream functionality in South Central Plains watersheds.