Rethinking livestock encroachment at a protected area boundary

Abstract

The presence of livestock inside protected areas, or "livestock encroachment," is a global conservation concern because livestock is broadly thought to negatively affect wildlife. The Maasai Mara National Reserve (MMNR), Kenya, exemplifies this tension as livestock is believed to have resulted in the declining wildlife populations, contributing to the strict and sometimes violent exclusion measures targeting Maasai pastoralists. However, research embedded in the real-world setting that draws insights from the social-ecological contexts is lacking. In this study, we conducted 19 mo of ecological monitoring covering 60 sites in MMNR and found that cattle presence inside the reserve did not significantly impact most co-occurring wild herbivores at the current intensity. Using the Hierarchical Modeling of Species Communities and Gaussian copula graphic models, we showed that cattle had no direct associations-neither negative nor positive-with nearly all wild herbivores despite frequently sharing the same space. Moreover, we did not detect resource degradation correlated with cattle presence near the MMNR boundary. Given the colonial legacy and land use history of Mara, entering MMNR becomes the only viable option for many herders. These results corroborate the emerging perspective that the ecological impacts of extensively herded livestock on wildlife might be more nuanced than previously thought. To effectively balance the needs of people, livestock, and wildlife, the current rigid livestock exclusion measures need to be reassessed to holistically consider herbivore ecology, local land use history, and modern politics of protected area management.

Keywords: East Africa; herbivore assemblage; pastoralist; wildlife conservation; wildlife–livestock conflict.

Fig. 1.

Study area and sampling design. The left panel shows the timeline of key events about the establishment and management of Maasai Mara National Reserve (MMNR, green bounded area in the map) and neighboring private conservancies (orange bounded areas) in Kenya. For more detailed information on the land history, see SI Appendix, Supporting Text. The right panel demonstrates the 60 sampling sites (purple dots) along five transects extend from the MMNR boundary 12 km into the reserve.

Fig. 2.

Occurrence of cattle and eleven wild herbivores using dung counts as the proxy in the Maasai Mara Nature Reserve from May 2018 to November 2019. (A) Wildlife and livestock occurrence summarized over the 60 sample sites overlapped with monthly precipitation (mm, gray dash line). Lower panels show the predicted cattle (B), zebra (C), and wildebeest (D) occurrence within the 95% credible interval inside MMNR as a function of distance to the reserve boundary. The prediction was conducted for the driest month when cattle use of the reserve is typically at its peak and when wildlife–livestock conflicts were believed to be the greatest (July, average precipitation 13.4 mm, calculated from the 20-y TerraClimate precipitation history). Each line represents a single draw from the posterior distribution of the Hierarchical Modeling of Species Communities model, and a total of 4,000 predictions were made for each species.

Fig. 3.

Species association networks derived from three different methods showing positive (blue), negative (red), or no association (white) among the 12 herbivore species in MMNR. (A) Raw species occurrence correlations based on Pearson’s correlation; (B) Species correlation estimated using the residual association structure among species from the Hierarchical Modeling of Species Communities, which controls for effects of environmental variables; (C) Direct species association estimated as the conditional dependence among species from Gaussian copula graphic models, which controlled for effects of environmental variables as well as mediator species. The width and shade of links represent the magnitude of correlation (SI Appendix, Fig. S3).