Insights into landforms and soils for explaining plague hosts-vectors interaction in Western Usambara Mountains, Tanzania

PhD. Thesis

Western Usambara Mountains in Lushoto District, Tanzania, are highly populated owing to favourable climatic conditions for production of both cash and food crops such as coffee, tea, maize, beans, potatoes and assorted fruits and vegetables. Some areas in the district are also potential as tourist destinations. However, clusters of villages in the district are reported to be hot-spots of plague recurrence that has caused deaths of thousands of people. Plague outbreaks in some villages of Lushoto District show a high frequency (on average, more than 1 year out of three) while other villages, very nearby (<10 km) have low plague frequency (less than once in 10 years). In these areas, plague has persisted for over 20 years in some villages. It is not known why plague has persisted for years in some villages but is not observed in others in the neighbourhoods. Also the irregularity of the disease outbreaks (both in time and space) pose a serious risk to communities. Earlier studies show that this persistence and the foci of the disease could not be explained by differences in fauna composition or human domestic behaviour. Hence, it was thought that a critical analysis of landscape factors at farm scale could provide explanations for plague occurrence. This study attempted to establish a landscape ecological framework in which the importance of factors of landforms, soils and soil microclimate linked to rodents, fleas and rodent burrows can be analysed as proxy for plague risk prediction. Specifically the study was aimed to i) analyse landforms and associated soil properties in relation to occurrence of small mammals (rodents) and fleas, ii) investigate the status of soil micronutrients with respect to landforms and small mammals; abundance in the plague risk area, iii) investigate the influence of landform characteristics and surface attributes at farm scale on rodent burrows, and iv) examine the influence of soil physical properties and soil microclimatic attributes on rodent burrows’abundance, portals’ orientation and burrows’ use. The study employed standard field survey and mapping of landforms, surface attributes and soils using remote sensing and GIS techniques. Soil samples were analysed in the laboratory for physical and chemical properties following internationally accepted methods. Trapping of small mammals was conducted on purposely selected sites using traps of different types and sizes. The trapped small mammals were counted and recorded, and treated with ethanol for further analysis. Fleas were removed from the animals by brushing the fur using ethanol, counted, recorded and stored. In the field, rodent burrows were surveyed systematically in the mapped landforms at farm scale. Data were explored using correlation and descriptive statistical analysis while regression analysis was carried out by generalised linear models (GLM) and boosted regression trees (BRT) in the framework of R statistics and Geographic Information System (GIS). The results show that phosphorus, base saturation, elevation and slope aspect are soil and landform characteristics which favour diversity of rodents (p<0.001); with Shannon Diversity Index increasing with elevation with values of 1.58, 1.62 and 1.69 for Plain, Escarpment and Plateau respectively. Fleas abundance was discouraged by hill-shade and magnesium while it was favoured by available phosphorus and organic carbon (p<0.001). The studied DTPA extractable Fe, Mn, Cu and Zn were variable. Fe ranged from 2.13 to 399.4 mg/kg soil, with a mean of 65.3 mg Fe/kg soil across the geomorphic units. Mn ranged from 0.59 to 266.28 mg Mn/kg soil while Cu ranged from 0.25 to 8.19 mg/kg soil with a mean of 2.98 mg/kg soil. Results show that Zn ranged from 0.08 to 19.6 mg/kg soil. Higher concentrations of Mn and Zn occurred mostly in the topsoils which also contained relatively high organic matter. Iron and Cu showed more or less equal distribution with soil depth but generally micronutrients declined with soil depth. Iron was found to significantly (p<0.01 and p<0.05) influence abundance of plague hosts and vectors. It is concluded that micronutrient levels vary in function of soil types and geomorphic units. Whereas higher concentrations were found in the plateau and upper part of escarpment, lower levels were in the plain. Fe had positive influence on abundance of plague hosts and vectors. Landforms in the studied plague frequency landscapes vary significantly in terms of elevations and slopes (p<0.05). Abundance of rodent burrows varies significantly (p<0.05) with different landform characteristics in the studied plague frequency landscapes, with high abundance in the high plague frequency. Surface cover favoured rodent burrowing in high plague frequency areas. Elevation and slopes interactively provided a favourable environment for rodent burrowing. In the medium plague frequency landscape, slope length and slope aspect were the dominant landform characteristics favouring rodent burrows. Hence high elevations and vegetation cover favoured rodent burrowing; thus, these attributes should be considered in future plague research. It has been established in this study that topsoil depth, a natural Ap/Ah horizon was an important soil physical property (P<0.05) which influenced the abundance of rodent burrows with a relative influence of 39.5 percent. The results further showed that there was no statically significant influence of soil physical properties and soil microclimate attributes (P>0.05) on burrow port orientation. The results showed that both soil depth and temperature encouraged burrow use in the high plague frequency landscape (p<0.05). Soil depth, dry soil consistence and subsoil relative humidity (RH (%) within 30 cm depth) significantly (p<0.05) could explain the burrow abundance in the medium plague frequency landscape. The use of burrows depended on the resistance of burrow to collapse, a characteristics which is dependent on the compaction of topsoil. However, dry subsoil consistence discouraged burrowing. Subsoil dry consistence had a significant (p<0.05) negative influence on burrow portal orientation in the medium plague frequency area, something attributed to compaction of the soils. It is concluded that soil physical and soil microclimate in the high plague frequency landscape encourage burrowing, a phenomenon correlating with the reported human plague contact risks. The information obtained from this study could be applied for surveillance, monitoring and control of rodent populations.

Landscape-ecological Clarification of Bubonic Plague Distribution and Outbreaks in the Western Usambara Mountains, Tanzania (LEPUS) via VLIR own initiative project