Climate Change and Variability: Map of Protected and Agricultural Regions in South Central Tanzania



  • Britta Schumacher | Department of Geography, MA/PhD Student
  • Vania Wang, MPH | Department of Geography, PhD Student

Description of the study map and research project

The following map depicts the study area for a research project titled, Farmer perceptions of climate change and variability in villages adjacent to the Udzungwa Mountains National Park, Tanzania. This project takes place in three villages—Magombera, Mang’ula A and Msosa—in south central Tanzania. Magombera and Mang’ula A Villages lie to the east, and Msosa Village to the north west of the Udzungwa Mountains National Park. The park spans over 1,900 square kilometers and is one of only a few areas with protected status within the Eastern Arc Mountains of Tanzania, conserving and supporting biodiversity and endemism within the Eastern Afromontane Biodiversity Hotspot. These mountains are globally renowned for their high concentrations of endemic species and biodiversity (Bunting et al., 2011; Burgess et al., 2007) and stretch from the Taita Hills of Kenya, southward to the Mahenges of Tanzania (Dinesen, Lehmberg, Rahner, & Fjeldså, 2001). These forests are unparalleled in importance for the preservation of biodiversity and endemicity on the African continent (Dinesen et al., 2001; Topp-Jørgensen, Reinhardt Nielsen, Marshall, & Pedersen, 2009).

According to recent studies, these mountains are under increasing threat, resulting from the proximity of tens of thousands of smallholder farmers to the mountains (Burgess et al., 2007; Cordeiro et al., 2007; Harrison, 2006; Rovero, Mtui, Kitegile, & Nielsen, 2012). Should climate change shift the viability of local farming livelihoods, which support 96 percent of individuals in the area, it is likely that farmers will turn to the park for resources (Harrison, 2006).

This project aims to understand the environmental, agronomic, and climatic perceptions of smallholder farmers who live adjacent to the Udzungwa Mountains National Park. Considering the conservation value of this region, it is important to demonstrate how farmers here perceive environmental, climatic, and livelihood changes, to begin understanding potential future impacts and conservation-livelihood strategies. Increasing system resilience here will not only improve smallholder livelihoods, but also ensure future forest health. The following sections show a map of this area and provide brief descriptions of each of the study villages.

Study map

Figure 1: Study Sites Map

Village Descriptions

Magombera Village


Figure 2

Magombera Village (Figure 2) lies at 36°, 56” east; 7°, 49” south on the Kilombero Floodplain in the Kilombero District of the Morogoro Region, just north of the Magombera Forest. This small forest is one of the last remaining tropical lowland forest fragments between the Udzungwas and the Selous Game Reserve, a protected area of exceptional conservation value in south-eastern Tanzania (Gillingham & Lee, 2018). Magombera Village consists of three small and isolated settled areas, interspersed with sparse vegetation, household bustanis (gardens) where vegetables are grown, and small houses constructed primarily of wattle and daub (stick frames plastered with mud) with thatched or corrugated rooves. Most villagers rely primarily on subsistence, rainfed agriculture as their primary livelihood strategy, though many partake in alternative strategies (e.g., livestock rearing, weaving, brewing) to supplement farming.

The village lies in the lowlands, allowing many farmers to grow rice on an annual basis. Elevation, human-made wetland rice paddies, and continuously saturated soils in some parts of Magombera Village provide an opportunity to grow rice throughout seasons when rice production is impossible in other nearby villages, attracting additional in-migrants to Magombera. Agroecologically, the village lies at the transition zone between rice and sugar cane. These crops are grown as monocultural stands in the large (> 1 acre) shambas (plots) that farmers keep on the outskirts of the settled areas. Many farmers are out-growers for the commercial Illovo Sugar Cane operation.

Mang’ula A Village


Figure 3

Mang’ula A (Figure 3) lies at 36o, 54” east; 7o, 50” south on the Kilombero Floodplain, directly to the east of the Udzungwa Mountains National Park entrance. The village is large, with one main settlement. The settlement consists of mostly homes, kitchens and latrines, many of which are made of brick or wattle and daub, occasionally covered stucco, with corrugated or thatched rooves. Many villagers rely on a mixed wage-subsistence livelihood, where farming is supplemented by other livelihood strategies that produce liquid assets (e.g., owning a small business, participating in microfinance schemes).

The village lies at the transition zone between rice, sugarcane and maize. Many fields are inundated and appropriate for rice production seasonally, during the long, masika, rains which occur from March to May, but some farmers own or rent land in the lowlands, where rice farming is appropriate all year.

Msosa Village


Figure 4

Msosa Village (Figure 4) lies at 36° 31” east; 7°, 30” south on the northwestern edge of the UMNP. The village has two settlements. The first settlement lies to the west of the Great Ruaha River, and the second along the smaller Msosa River, both set back from the rivers’ edge by agricultural fields. The settlements are sparse, with very little vegetation, consisting mostly of homes made from fired or mud brick or plaster with corrugated or thatched rooves, latrines, kitchens, and communal, raised stalls for storing onions. Most villagers rely on farming as their main livelihood strategy, though many partake in alternative strategies (e.g., retail shops, livestock keeping) to supplement their livelihoods.

Msosa village experiences unimodal rains but relies on gravity and pump irrigation for farming. Unlike Mang’ula A and Magombera, however, villagers primarily grow crops to sell in major markets, not for subsistence use. Primary commodities include onions, beans, and ground nuts. Farmers often hold and store cash crops in communal storage facilities for sale during the thin months of April and May, when prices are higher and food supplies are low.


In text

  • Bunting, G., Burgess, N., Carret, P., Silva, N. De, Gordon, I., Jbour, S., … Woldemariam, T. (2011). Ecosystem Profile: Eastern Afromontane Biodiversity Hotspot.
  • Burgess, N. D., Butynski, T. M., Cordeiro, N. J., Doggart, N. H., Fjeldså, J., Howell, K. M., … Stuart, S. N. (2007). The biological importance of the Eastern Arc Mountains of Tanzania and Kenya. Biological Conservation, 134(2), 209–231.
  • Cordeiro, N. J., Burgess, N. D., Dovie, D. B. K., Kaplin, B. A., Plumptre, A. J., & Marrs, R. (2007). Conservation in areas of high population density in sub-Saharan Africa. Biological Conservation, 134(2), 0–8.
  • Dinesen, L., Lehmberg, T., Rahner, M. C., & Fjeldså, J. (2001). Conservation priorities for the forests of the Udzungwa Mountains, Tanzania, based on primates, duikers and birds. Biological Conservation, 99(2), 223–236.
  • Gillingham, S., & Lee, P. C. (2018). People and protected areas : a study of local perceptions of wildlife crop-damage conflict in an area bordering the Selous Game Reserve , Tanzania. 37(3), 316–325.
  • Harrison, P. (2006). Socio-Economic Baseline Survey of Villages Adjacent to the Vidunda Catchment Area , Bordering Udzungwa Mountains National Park National Park. WWF.
  • Rovero, F., Mtui, A. S., Kitegile, A. S., & Nielsen, M. R. (2012). Hunting or habitat degradation? Decline of primate populations in Udzungwa Mountains, Tanzania: An analysis of threats. Biological Conservation, 146(1), 89–96.
  • Topp-Jørgensen, E., Reinhardt Nielsen, M., Marshall, A. R., & Pedersen, U. (2009). Relative densities of mammals in response to different levels of bushmeat hunting in the Udzungwa Mountains, Tanzania. Mongabay.Com Open Access Journal -Tropical Conservation Science Tanzania. Tropical Conservation Sciencecom Open Access Journal -Tropical Conservation Science, 22(211), 70–8770.


  • OpenStreetMap contributors. (2018). Roads, rivers, streets, cities, villages retrieved from OSM via
  • Platts, P.J., Burgess, N.D., Gereau, R.E., Lovett, J.C., Marshall, A.R., McClean, C.J., Pellikka, P.K.E., Swetnam, R.D., Marchant, R. (2011). Delimiting tropical mountain ecoregions for conservation. Environmental Conservation 38(3): 312-324.
  • World Database for Protected Areas, (2015). Protected areas Tanzania polygons.


Reconceptualizing geography

The following is an excerpt from a critique I wrote for our first-year philosophy of science course. It gives my perspective on the field of geography, as someone who entered a highly-regarded doctoral program in geography having never taken a geography course.

Like many of my colleagues within the cohort of UCSB geography doctoral students, I’m not a geographer in a traditional sense. My background spans multiple disciplines, including—but not restricted to–microbiology, public health, computer science, and more recently, network science. Naturally then, my perception of space is colored by these experiences; and by virtue of having the considerable fortune to dedicate a life to scholarship, I’m afforded the time and intellectual resources to ponder about ‘space’, ‘region’, and ‘place’. This piece is mainly a meditation on ‘region’ and how it should be conceptualized given the context of our times. Here, I will humbly suggest some alternative definitions for ‘region’, while reflecting on how my prior learning, life-experiences, and personal philosophies give rise to these insights. I propose an alternative frame onto which space–and its constituent components–can be modeled.

Unlike biological classification–which rests on a system of taxonomy that funnels living beings into categories of genera and species–geographical classification is noticeably more nebulous. Is a taxonomy of ‘place’ a conceiveable possibility? If geographers attempt to emuluate biological taxonomy as a classification model, would it be possible to define spatial entities that are divorced from the influence of politics and society?

Geographical researchers strive hard to claim legitimacy within the scientific community as a ‘hard’ science. Spatial science is the answer to geography’s perpetual goal to establish itself as legitimate, objective, and free from the shackles of culture and other unquantifiables that have colored the field of geography. However, the fundamental units of geography–place, region, and space–are deeply intertwined with political, cultural, and historical contexts. Therefore, I argue that region should not be removed from its social context. The quantitive methods of spatial science can be used to analyze phenomena that are subjective, social, political, and cultural.

Given today’s dizzying slew of technological and computational tools, mathematical and statistical models have been constructed to simulate and predict human behavior, social dynamics, and even cultural dynamics. For instance, ‘culturomics’ is a burgeoning branch of sociology that uses Big Data techniques to analyze tomes of cultural records. Some notable studies describe examining large corpora of digitized literature to track the development of the English language through history; or using a computational technique called topic modeling to extract contextual meanings from text. The common thread of these studies lie in their use of Big Data, graph theory, and stochastic modeling to find patterns within collections of digitized human artifacts.

My own work rests within the realm of models, and other mathematical and statistical descriptors that aim to distill biological and social behavior into their most salient components. As a trained microbiologist and public health worker, I’m conditioned to view phenomena as tiny interacting entities that progressively affect larger systems–starting from the organism itself, to the organism it ails (usually the human host), to the affected community. When I conceptualize ‘space’, I think similarly. I’m advocating conceptualizing space and its constituent components from a relational perspective. Spatial entities can be modeled as nodes in a network. Relationships, symbolized through edges, can mean that two entities are close in physical proximity or related given a comparison of nodal attributes. Networks can multi-layered, dynamic, and temporal, with entities of many related networks interacting. By design, this network model can be as simple or as complex as the researcher desires.

The concept of space is changing. Vidal’s region and Hawthorne’s chorology are no longer sufficient descriptors of space and place. In this age of computation, it would be a folly to ignore the power of graph models and data science in our conceptualization of space.