Winds, waves, warm waters, weekdays, and which ways boats are counted influence predicted visitor use at an offshore fishing destination
Quantifying the number of recreational fishers is important for many aspects of managing coastal resources. Unfortunately, quantifying recreational boaters in offshore settings has proven difficult due to their distance from shore and a lack of cost-effective methods to monitor small boats (<10 m length). We investigated visitor-use at an offshore marine protected area (MPA) in the southeastern USA. We used multiple methods of counting boats (satellites, buoy camera, passive acoustics, and boat-based observations) and a generalized linear modeling approach to identify environmental and calendar-based predictor variables that influenced visitation. Based on the model, predicted visitor-encounter rates were estimated for various weather and calendar-based scenarios, and the probability of detecting a hypothetical change in visitation with each counting method was examined through a power analysis. The most important predictors were day of the week, special day (e.g., tournament), water temperature, and wave height. Boat counts were 2–5 times higher on weekend days than on weekdays. More boats were predicted on weekdays with good weather (defined as water temperature 24 °C, wave height 0.5 m), than weekends with decent weather (17 °C and 1 m). Considering weekends alone, those with good weather were predicted to have 5 times higher visitation than weekends with decent weather. Predicted visitation was highest on calm days, dropped by ∼75 % when wave height reached 1 m, and was essentially zero when wave height exceeded 1.5 m. Highest counts were predicted when water temperature was warmest and gradually declined as temperatures cooled. For the buoy camera and passive acoustic boat-count methods, power analysis suggested that 3–6 years of typical samples before and after a hypothetical 25 % increase in visitation would be needed to have an 80 % chance of detecting the change. Other techniques would take 14 or more years of typical samples. The process used here for investigating visitation can be adapted to other offshore or remote locations.