Abstract

Wild yam species (Dioscorea spp.) provide important supplementary food and thus contribute directly to the livelihood and well-being of local people in SW Madagascar. Given the ongoing exploitation of this resource, there is a necessity to identify and predict yam species distribution along environmental gradients to improve our understanding of sustainable management of this resource. Therefore, the aims of the current study were to identify predictors for the distribution of important wild yam species and to spatially predict their availability. To this end species abundance and environmental variables were collected in the field using a systematic sampling approach within a yams collection area of four villages (58 plots). A redundancy analysis (RDA) was conducted to investigate the relationship between wild yam species and 12 environmental variables. Species distribution models and species response curves were established for the most abundant wild yam species using nonparametric multiplicative regression (NPMR). These models were subsequently used in conjunction with geospatial data for predictive mapping. RDA depicted a clear pattern of species habitats with D. alatipes occurring in dry spiny forests on calcareous soils at remote places, while D. bemandry and D. fandra were found in forest habitats on sandy soils with high harvest intensities. The NPMR models explained 88% (D. alatipes), 82% (D. bemandry) and 37% (D. fandra) of the variation in species abundance. Sensitivity analysis indicated the importance of vegetation structure, human interventions, and soil characteristics in determining wild yam distribution. Predicted distribution maps showed that the population of wild yam is scarce and mostly located in restricted areas of open spiny forests and dry spiny forest thickets, where harvest intensity is high. This study highlights the need for long-term assessment and public awareness actions on yam harvest practices as well as the importance of anthropogenic factors for the distribution of yam as a key forest resource in SW Madagascar.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call