Sustaining indigenous Maasai Alalili silvo-pastoral conservation systems for improved community livelihood and biodiversity conservation in East African rangelands

2.4 Research design and sampling techniques Prior to main data collection, a reconnaissance survey was conducted to confirm the availabil- ity of Maasai Alalili systems within districts in the proposed study regions. The sampling in this study involved a cross-sectional research design whereby both purposive and stratified random sampling techniques were applied. Administrative boundaries of the Arusha and Manyara regions were marked through Quantum Geographical Information System (Q-GIS) software version 3.12.3 [67] to enhance the sampling process of the Maasai land with the assis- tance of shape files collected from Tanzania Wildlife Research Institute (TAWIRI). Five dis- tricts from the two regions were sampled purposively, targeting pastoral and agro-pastoral Maasai communities (Fig 1). A purposive sampling of the five districts dominated by the Maa- sai communities was adopted to enhance the identification process of the Alalili systems and livestock herd size grazed in the sampled Alalili systems through key informant interviews and focused group discussion respectively [68, 69]. Key informant interviews were conducted in each district for the purpose of identifying villages that have Alalili systems as well as preparing the list of all Alalili systems from identified villages. Key informants included district game officers (DGOs), rangeland management officers (ROs), ward/village executive officers (VEO/ WEO), and members of the village rangeland management committee. A total of 61 interviews

were done in the study area whereby five were conducted at district levels and 56 at ward lev- els. DGOs and ROs from all five districts were consulted through direct physical meetings con- ducted at each district council office headquarters during a reconnaissance survey. WEO/VEO from a total of 88 wards of the sampled districts were contacted with assistance from DGOs and ROs via phone/mobile calls considering the limited time, available resources, and broader range of the study area. The interviews between DGOs, ROs, and WEO/VEO enhanced to reach out the leaders of the respective village rangeland management committees who gave the list of all Alalili systems present in their villages. A total of 298 Alalili systems were identified and numbered continuously, from which 40% (119 Alalili systems) were sampled to cover a wider range of the study area through randomization technique with the adoption of random number tables. Focused group discussion comprised of one member of the village rangeland management committee and two or three members of the pastoral households who graze live- stock in the sampled Alalili systems for estimation of livestock herd size. Stratification of the study area was conducted with regard to the existing land use practices in the Arusha and Manyara regions. Six types of land use strata were plotted through Q-GIS and distinguished using the topographic map of sampled regions. The land use strata were open areas, WMA, GCA, NCA, national parks, and game reserves. National parks and game

reserves were not accounted for mapping of Alalili as Maasai Alalili silvo-pastoral conserva- tion systems are not allowed in the national parks and game reserves. Stratification of the sam- pled villages that comprised the sampled Alalili systems was done based on different land-use systems with respect to their distribution. An empirical evidence-based method (Ground- truthing technique) was applied through field visitation and direct observation. DGOs and ROs assisted in verifying the exact land use types in the study area whereby GCA covered the largest proportion (45%) followed by NCA (20%), open areas (19%), and WMA (16%) [70]. 2.5 Data collection All Alalili samples were visited in each sample village and their respective locations were recorded at the center of each surveyed Alalili system with a GPS to enhance mapping through the Q-GIS platform. Since the specific boundaries and actual sizes (ha) of each Alalili were not known, four radii were recorded from the center of each visited Alalili system to the north, south, east, and west directions assuming that Alalili has an irregular polygon. The Alalili PLOS ONE Sustaining Alalili systems in rangelands of East Africa PLOS ONE | https://doi.org/10.1371/journal.pone.0303649 May 16, 2024 6 / 24