BACKGROUND: Traditional fishing communities are strongholds of ethnobiological knowledge but establishing to what degree they harbor cultural consensus about different aspects of this knowledge has been a challenge in many ethnobiological studies.
METHODS: We conducted an ethnobiological study in an artisanal fishing community in northeast Brazil, where we interviewed 91 community members (49 men and 42 women) with different type of activities (fishers and non-fishers), in order to obtain free lists and salience indices of the fish they know. To establish whether there is cultural consensus in their traditional knowledge on fish, we engaged a smaller subset of 45 participants in triad tasks where they chose the most different fish out of 30 triads. We used the similarity matrices generated from the task results to detect if there is cultural consensus in the way fish were classified by them.
RESULTS: The findings show how large is the community\'s knowledge of fish, with 197 ethnospecies registered, of which 33 species were detected as salient or important to the community. In general, men cited more fish than women. We also found that there was no cultural consensus in the ways fish were classified.
CONCLUSIONS: Both free-listing and triad task methods revealed little cultural consensus in the way knowledge is structured and how fish were classified by community members. Our results suggest that it is prudent not to make assumptions that a given local community has a single cultural consensus model in classifying the organisms in their environment.

Recording local ecological knowledge (LEK) is a useful approach to understanding interactions of the complex social-ecological systems. In spite of the recent growing interest in LEK studies on the effects of climate and land use changes, livestock mobility decisions and other aspects of agro-pastoral systems, LEK on forage plants has still been vastly under-documented in the West African savannas. Using a study area ranging from northern Ghana to central Burkina Faso, we thus aimed at exploring how aridity and socio-demographic factors drive the distributional patterns of forage-related LEK among its holders. With stratified random sampling, we elicited LEK among 450 informants in 15 villages (seven in Ghana and eight in Burkina Faso) via free list tasks coupled with ethnobotanical walks and direct field observations. We performed generalized linear mixed-effects models (aridity- and ethnicity-based models) and robust model selection procedures. Our findings revealed that LEK for woody and herbaceous forage plants was strongly influenced by the ethnicity-based model, while aridity-based model performed better for LEK on overall forage resources and crop-related forage plants. We also found that climatic aridity had negative effect on the forage-related LEK across gender and age groups, while agro- and floristic diversity had positive effect on the body of LEK. About 135 species belonging to 95 genera and 52 families were cited. Our findings shed more light on how ethnicity and environmental harshness can markedly shape the body of LEK in the face of global climate change. Better understanding of such a place-based knowledge system is relevant for sustainable forage plants utilization and livestock production.