Amon Murwira

Monitoring change in the spatial heterogeneity of vegetation cover in an African savanna

The extent to which a new intensity‐dominant scale approach to characterizing spatial heterogeneity from remote sensing imagery can be used to monitor two‐dimensional changes (i.e. variability and patch size) in the spatial heterogeneity of vegetation cover (estimated from a Landsat Thematic Mapper (TM)‐derived Normalized Difference Vegetation Index (NDVI)) was tested in the Sebungwe region in north‐western Zimbabwe between 1984 and 1992. Intensity of spatial heterogeneity (i.e. the maximum variance obtained when a spatially distributed landscape property is measured with a successively increasing window size) was used to measure variability in vegetation cover. Dominant scale of spatial heterogeneity (i.e. the window size at which the maximum variance in the landscape property is measured) was used to measure the dominant patch dimension of vegetation cover. This approach was validated by testing whether the observed change in the dominant scale and intensity of spatial heterogeneity of vegetation cover between 1984 and 1992 was related to changes in the proportion of arable fields. The results also indicated that there was a significant relationship (p<0.05) between changes in the proportion of agricultural fields and changes in the intensity and the product of intensity and dominant scale of spatial heterogeneity (intensity×dominant scale), suggesting that the new approach captures observable changes in the landscape, and is not an artefact of the data. The results imply that the intensity‐dominant scale approach to quantifying spatial heterogeneity in remote sensing imagery can be used for a comprehensive characterization and monitoring of changes in landscape condition.

The use of multi-temporal MODIS images with ground data to distinguish cotton from maize and sorghum fields in smallholder agricultural landscapes of Southern Africa

In this study, we test whether we can significantly (p < 0.05) distinguish cotton (Gossypium hirsutum L.) fields from maize (Zea mays L.) and sorghum (Sorghum bicolor) fields in smallholder agricultural landscapes of the Mid-Zambezi Valley, Zimbabwe, using a temporal series of 16-day Moderate Resolution Imaging Spectroradiometer – normalized difference vegetation index (MODIS NDVI) data. We test this hypothesis at different phenological stages over the growing season, that is, early green-up onset, late green-up onset, green-peak, early senescence and late senescence. We also statistically compare the rate of change in the greenness of the three crops at the three phenological stages. Results show that we can significantly (p < 0.05) distinguish cotton fields from maize and sorghum fields using 16-day MODIS NDVI data during the late green-up onset as well as during the green-peak stage of the three crops. Our results indicate that cotton can successfully be distinguished from maize and sorghum in spatially heterogeneous smallholder agricultural landscapes using temporal MODIS NDVI.

Predicting maize yield in Zimbabwe using dry dekads derived from remotely sensed Vegetation Condition Index

Abstract Maize is a key crop contributing to food security in Southern Africa yet accurate estimates of maize yield prior to harvesting are scarce. Timely and accurate estimates of maize production are essential for ensuring food security by enabling actionable mitigation strategies and policies for prevention of food shortages. In this study, we regressed the number of dry dekads derived from VCI against official ground-based maize yield estimates to generate simple linear regression models for predicting maize yield throughout Zimbabwe over four seasons (2009–10, 2010–11, 2011–12, and 2012–13). The VCI was computed using Normalized Difference Vegetation Index (NDVI) time series dataset from the SPOT VEGETATION sensor for the period 1998–2013. A significant negative linear relationship between number of dry dekads and maize yield was observed in each season. The variation in yield explained by the models ranged from 75% to 90%. The models were evaluated with official ground-based yield data that was not used to generate the models. There is a close match between the predicted yield and the official yield statistics with an error of 33%. The observed consistency in the negative relationship between number of dry dekads and ground-based estimates of maize yield as well as the high explanatory power of the regression models suggest that VCI-derived dry dekads could be used to predict maize yield before the end of the season thereby making it possible to plan strategies for dealing with food deficits or surpluses on time.

Spatial modelling of Bacillus anthracis ecological niche in Zimbabwe

Anthrax continues to cause significant mortalities in livestock, wildlife and humans worldwide. In Zimbabwe, anthrax outbreaks have been reported almost annually over the past four decades. In this study we tested whether anthrax outbreak data and a set of environmental variables can be used to predict the ecological niche for Bacillus anthracis using maximum entropy modelling for species geographical distribution (Maxent). Confirmed geo-referenced anthrax outbreaks data for the period 1995-2010 were used as presence locations and a set of environmental parameters; precipitation, temperature, vegetation biomass, soil type and terrain as predictor variables. Results showed that the environmental variables can adequately predict the ecological niche of B. anthracis (AUC for test data=0.717, p<0.001), with soil type as the most important predictor followed by variance of vegetation biomass and maximum temperature. These results imply that the model we tested may be used by animal health authorities in devising better control strategies for anthrax.

Remote sensing of the link between arable field and elephant (Loxodonta africana) distribution change along a tsetse eradication gradient in the Zambezi valley, Zimbabwe.

We investigated whether the proportion of remotely sensed arable fields increased along a tsetse eradication gradient in the Sebungwe region. We also investigated whether and to what extent this increase in arable fields affected the distribution of the African elephant (Loxodonta africana) between the 1980s and 1990s. Results showed a relatively higher increase in the proportion of arable fields in the zone cleared of tsetse by 1986 compared to the zone that was still tsetse infested by the same date. Results also showed contrasting patterns in the relationship between the proportion of the habitat under arable fields and elephant distribution between the two periods. Specifically, in the 1980s, when arable field cover was between 0% and 11%, there was a weak (p > 0.05) positive relationship between elephant presence and the proportion of the habitat under arable fields. In contrast, a significant (p < 0.05) negative relationship emerged in the 1990s, when arable field cover ranged between 0% and 88%. Furthermore, the results demonstrated that the change in the probability of elephant presence between the early 1980s and the early 1990s was significantly (p < 0.05) related to the change in the proportion arable fields. In conclusion, this study demonstrated that the expansion of arable fields in the Sebungwe was greater in areas where tsetse had been eradicated compared with areas that were still tsetse infested. Overall, the results suggest that using remotely sensed data, we can conclude that tsetse eradication led to the redistribution of elephants in response to arable field expansion.

Estimating wood volume from canopy area in deciduous woodlands of Zimbabwe

In this study we tested the predictive ability of canopy area in estimating wood volume in deciduous woodlands of Zimbabwe. The study was carried out in four sites of different climatic conditions. We used regression analysis to statistically quantify the prediction of wood volume from canopy area at species and stand level using field data. Our results revealed that canopy area significantly (P < 0.05) predicted wood volume at both levels. Furthermore, the results show that at the species-specific level, canopy area explained 54–81% of the variance in wood volume with standard error of estimate (SEE) ranging from 0.056 to 0.71 m3. At stand level, canopy area significantly (P < 0.05) explained 58–84% of the variance in total wood volume with SEE ranging from 0.15 to 3.99 m3 ha−1. Across all study sites, the relationship between canopy area and wood volume at stand level was best described by a logistic regression function, with a R2 value of 0.65 and SEE of 0.7 m3. We concluded that canopy area significantly (P < 0.05) predicted wood volume of dominant tree species in Zimbabwean deciduous woodlands. The relationship between wood volume and canopy area provides an opportunity of estimating wood volume using remote sensing as canopy area can be viewed and measured from aerial, as well as satellite-borne sensors.

An IKONOS-based comparison of methods to estimate cattle home ranges in a semi-arid landscape of southern Africa

We estimated home range (HR) and core areas of cattle herds in a semi-arid rangeland in southern Africa using the fixed kernel and the local convex hull (LoCoH) methods. We also compared the HR values of the two methods with the aid of high spatial resolution IKONOS imagery. We also compared area estimates determined by the two methods at different probability contours of the utilization distribution (UD). Results showed that the LoCoH performed better than the kernel method in estimating UD. We found significant (p < 0.05) differences concerning the area estimated at each probability contour of the UD between the fixed kernel and LoCoH methods. However, both methods produced similar land cover preferences within the HR and core areas. Based on IKONOS-imagery-aided evaluation, our results imply that LoCoH determines core areas in HR analysis better than the kernel method, while both methods can be used in preference analysis.

Relationship between remotely sensed variables and tree species diversity in savanna woodlands of Southern Africa

In this study, we test whether and in what way tree species diversity in three savanna woodland sites is related to Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) satellite imagery-derived indices. We test the use of standard deviation of near-infrared (stdev NIR) radiance and tree canopy cover estimated through the soil-adjusted vegetation index (SAVI) in estimating diversity. We use linear regression analysis to test the existence as well as determine the nature of the relationship between tree species diversity estimated from fieldwork data and stdev NIR radiance and SAVI. Our results show that tree species diversity has a significant (p < 0.05) hump-shaped response to variations in stdev NIR radiance and SAVI. Furthermore, results show that the combination of stdev NIR and SAVI explains between 60% and 64% of the variations in tree species diversity, an improvement of between 30% and 54% explained by the indices individually. We conclude that ASTER remotely sensed data can successfully be used to estimate tree species diversity in savanna woodlands.

Cotton fields drive elephant habitat fragmentation in the Mid Zambezi Valley, Zimbabwe

Abstract In this study we tested whether cotton fields contribute more than cereal fields to African elephant ( Loxodonta africana ) habitat loss through its effects on woodland fragmentation in the Mid-Zambezi Valley, Zimbabwe. In order to test this hypothesis, we first mapped cotton and cereal fields using MODIS remotely sensed data. Secondly, we analysed the effect of the area of cotton and cereal fields on woodland fragmentation using regression analysis. We then related the fragmentation indices, particularly edge density with elephant distribution data to test whether elephant distribution was significantly related with woodland fragmentation resulting from cotton fields. Our results showed that cotton fields contributed more to woodland fragmentation than cereal fields. In addition, results showed that the frequency of the African elephant increased where cotton fields were many and small relative to cereal fields. We concluded that cotton fields are the main driver of woodland fragmentation and therefore elephant habitat in the Mid-Zambezi Valley compared with cereal fields.

Predicting forest carbon stocks from high resolution satellite data in dry forests of Zimbabwe: exploring the effect of the red-edge band in forest carbon stocks estimation

In this study, we tested whether the inclusion of the red-edge band as a covariate to vegetation indices improves the predictive accuracy in forest carbon estimation and mapping in savanna dry forests of Zimbabwe. Initially, we tested whether and to what extent vegetation indices (simple ratio SR, soil-adjusted vegetation index and normalized difference vegetation index) derived from high spatial resolution satellite imagery (WorldView-2) predict forest carbon stocks. Next, we tested whether inclusion of reflectance in the red-edge band as a covariate to vegetation indices improve the models accuracy in forest carbon prediction. We used simple regression analysis to determine the nature and the strength of the relationship between forest carbon stocks and remotely sensed vegetation indices. We then used multiple regression analysis to determine whether integrating vegetation indices and reflection in the red-edge band improve forest carbon prediction. Next, we mapped the spatial variation in forest carbon stocks using the best regression model relating forest carbon stocks to remotely sensed vegetation indices and reflection in the red-edge band. Our results showed that vegetation indices alone as an explanatory variable significantly (p < 0.05) predicted forest carbon stocks with R2 ranging between 45 and 63% and RMSE ranging from 10.3 to 12.9%. However, when the reflectance in the red-edge band was included in the regression models the explained variance increased to between 68 and 70% with the RMSE ranging between 9.56 and 10.1%. A combination of SR and reflectance in the red edge produced the best predictor of forest carbon stocks. We concluded that integrating vegetation indices and reflectance in the red-edge band derived from high spatial resolution can be successfully used to estimate forest carbon in dry forests with minimal error.