Trophic interactions (i.e., predator-prey relationships) among target reef fish comprising the majority of catch by all fishing gears employed in five study sites along the Kenyan coast.
To determine which species were caught by each gear type being used by fishers within these sites, we used a long-term fish catch dataset collected by the Wildlife Conservation Society. The dataset included surveys from 25 landing sites along the Kenyan coast conducted continuously between 2010 and 2016. For each observation, onsite observers identified landed catch at the species level in addition to the gear used. Observers were present at landing stations every sampling day before the arrival of boats and stayed until the entire landing process was concluded. Although all patrols were conducted during daylight hours, the sampling method does not exclude catches attributed to nighttime fishing activities, as observers also intercepted fishers returning from their overnight fishing, ensuring that each gear used at each site was sampled and that each species landed was recorded. The number of patrols conducted per month were not stratified, but similar intervals of sampling were maintained within this randomized block design to detect long-term catch trends. Data was collected at least 8 days per month, translating to a total of 599 sampling days over the survey period and an observed total of 19,467 individual fish caught across all gear types. Most gears used in multispecies coral reef fisheries incidentally catch a number of species infrequently. We therefore focused on species comprising the majority of the total catch for each gear type, excluding all species that comprised less than 1% of the total catch. This resulted in 36 key target reef fish species. Trophic interactions capturing predator-prey relationships among the 36 key target reef fish species were estimated based on a combination of diet, relative body size, and habitat use (likelihood of encounter). The corresponding ecological network is thus undirected, with edges representing trophic interactions between fish species. Diet and body size (maximum length) data were taken from FishBase, and broad patterns of habitat use (i.e., pelagic, demersal, coral-dominated, macroalgal bed) were estimated from published records and expert first-hand knowledge of coral reef ecologists at the ARC Centre of Excellence for Coral Reef Studies at James Cook University, Australia. Detail on the specific fish species consumed from analyses of gut contents is largely unavailable, yet piscivorous coral reef fish are known to be generalists in terms of the species they consume. We thus took a conservative approach, considering one species to prey on another if its diet was predominantly piscivorous, its body length was large (ca. ≥ 2 times) compared to that of the prey species, and the two species occupied a similar habitat. We did not identify potential predatory links for any species whose diet was not primarily piscivorous as any likely predation on other targeted species was likely to be too infrequent to have a meaningful effect on prey populations.