Plants · 8 authors, 5 centres
This summary was generated by AI from a single paper. It has not been reviewed by a clinician and is not clinical advice. Verify against the source before acting on it.
This study used regression, binomial, and Bayesian models to identify plant orders and families in Kenya that are significantly overused for medicine and food. Key medicinal outliers included Fabales, Sapindales, and Rutaceae, while food outliers included Gentianales and Anacardiaceae. These findings help prioritize taxa for bioprospecting and conservation.
**Background:** Globally, medicinal and food plants are not randomly selected from the flora; certain taxa are preferentially used. Understanding these patterns can guide drug discovery and conservation. Kenya, with over 5000 indigenous plant species and diverse ethnic groups, has rich ethnobotanical knowledge but lacks quantitative analyses of plant selection patterns. This study aimed to identify orders and families significantly overrepresented for medicinal and food uses in Kenya using three complementary statistical approaches.
**Methods:** An extensive literature search (74 articles, 4 reports, 3 books, 11 dissertations for medicinal plants; 19 articles and 2 books for food plants) was conducted to compile a database of 1298 medicinal and 664 food plant species from Kenya's flora (5883 species, 188 families, 50 orders). Three quantitative models were applied: (1) Linear regression (LINEST function) to calculate residuals and relative regression residuals (R-values), with positive outliers defined as R > 0.5; (2) Binomial analysis (BINOMDIST function) to test significance of departure from expected proportions (p < 0.05); (3) Bayesian analysis (BETA.INV function) to compute 95% credible intervals, identifying taxa where the inferior interval exceeded the superior interval for the whole flora.
**Key Results:** For medicinal taxa, 14 positive outlier orders were identified by all three models, all significant (p < 0.05). Fabales had the highest regression residual (66.16), while Sapindales had the highest R-value (1.1605). Thirty-eight positive outlier medicinal families were identified; 34 were significant (p < 0.05). Rutaceae had the highest R-value (1.6808), and Fabaceae had the highest residual (63.2). For food taxa, 16 positive outlier orders were recovered; 13 were significant (p < 0.05). Gentianales had the highest residual (45.27), and Sapindales had the highest R-value (2.3654). Forty-two positive outlier food families were identified; 30 were significant (p < 0.05). Anacardiaceae had the highest R-value (5.163), and Fabaceae had the highest residual (28.72). Seven orders (Celastrales, Cucurbitales, Ericales, Magnoliales, Malvales, Rosales, Sapindales) and 14 families (e.g., Amaranthaceae, Anacardiaceae, Rutaceae) were significant outliers for both medicine and food.
**Clinical Implications:** The identification of significantly overused medicinal families such as Rutaceae (containing antimalarial and anticancer compounds like nitidine) and Euphorbiaceae (rich in terpenoids) supports phylogenetic-based bioprospecting. The overlap between medicinal and food taxa suggests some plants serve dual purposes, but differences indicate independent selection criteria. These results prioritize taxa for pharmacological screening and conservation, especially in Kenya's arid regions where plant-based medicine is prevalent. The study also highlights that large families like Asteraceae are not necessarily overused, emphasizing the need for quantitative approaches over simple species counts.