Theses and Dissertations (Mammal Research Institute)

Permanent URI for this collectionhttp://hdl.handle.net/2263/39987

Browse

Recent Submissions

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Factors that affect owl ecology in an agricultural matrix in east Gauteng
    (University of Pretoria, 2019-11) Keith, Mark; cghannweg@gmail.com; Hannweg, Caroline Grace
    Anthropogenic impacts are extensive and affect species at individual and population levels as well as communities. To understand these anthropogenic impacts we have to understand various factors, such as agriculture, and the expansion of roads and how these affect populations, species and communities. The Boesmanspruit Highveld Grassland ecosystem, between Springs and Devon, Gauteng, is a critically endangered ecosystem, dominated by agriculture in the area, urbanisation and extensive road networks. The agricultural crops include maize, soya beans, and cultivated fields for animal feed. The region also contains the vulnerable African grass-owl (Tyto capensis), and three other owl species namely, the western barn owl (Tyto alba), the spotted eagle-owl (Bubo africanus) and the marsh owl (Asio capensis). The extensive road network, and particularly the N17 highway in this area have resulted in continued owl roadkills in the area. During 2002-2003 an average of 9.2 dead owls per kilometre per annum was reported. Various factors contribute to these high mortality numbers, one being that maize trucks carrying grain along the N17 may spill grain onto the road, which lures rodents to feed on the waste and in turn attracts owls to the road where they can easily hunt, putting the owls in danger of being struck by vehicles at night. The aim of this research was to assess the species of owls in the area and their abundance, estimate their occupancy and unpack the ecological and anthropogenic factors that may be driving owl occupancy in the area. Road surveys were done for 5 nights, in October 2018, January 2019 and April 2019 allowing for a 15 night occupancy estimate. Only three of the reported four owl species were found, namely the western barn owl, the spotted eagle-owl and the marsh owl. The data collected during these surveys were used in an occupancy model to determine the occupancy of owls in the area, providing an estimate of 0.817 (SE = 0.102). The abundance was determined to be approximately 2 owls per 4 km2 (SE = 1.21). Covariates were also used to assess the factors that may be affecting the high occupancy of owls in the area. The covariate that had the greatest effect on the occupancy of owls in the area was water sources. With water sources incorporated into the models, occupancy increased to 1 (SE = 3.427 x 10-16) from the initial 0.817. This increased the abundance estimates to approximately 6 owls per 4 km2 (SE = 3.091). Comparing occupancy across the different months that were sampled, it was found that owl occupancy was highest in October 2018 at 0.996 (SE = 0.155), and lowest in April 2019 at 0.526 (SE = 0.173). Food resources are a key driver for owl occupancy. Rodent communities and density was assessed in the area. The rodents caught were mostly Mastomys species (92%) and the remaining rodents were identified as Rhabdomys dilectus (8%). Up to 55 rodents were caught in a hectare, with the highest number of rodents caught in the April 2019 sampling and the lowest number of rodents caught in the October 2018 sampling. The abundance of rodents increased as the height of the crops increased and fields where soya beans were being grown had the largest number of rodents. Comparing the species caught in the live traps of the rodent survey with an unpublished small pellet analysis done in the area, the barn owls regurgitated pellets contained four species of small mammals (Otomys spp., Gerbilliscus spp., Mastomys spp. and a shrew species Myosorex spp.), while only two rodent species were trapped in the fields. The asynchronous owl-rodent abundance reported in the study indicate that the three owl species are likely more driven by habitat availability and nesting opportunities, rather than purely by food availability. It is therefore important that the owl breeding seasons, and in particular grass-owl breeding seasons are taken into consideration when managing agriculture in the area. It is also important that when making decisions on where to build new infrastructure, areas where owls may be more likely to nest should be avoided. It should also be noted that where rodents are being controlled by ecologically based methods, such as by owls, it is likely that these methods alone are not effective enough to control the levels of rodents in the area. Therefore, it is important that other ecologically responsible methods of rodent control are used in the area.
  • Thumbnail Image
    Item
    Land-use mosaic effects on insect abundance, bat activity and macadamia production, South Africa
    (University of Pretoria, 2021) Keith, Mark; Weier, Sina; emmswartz@gmail.com; Swartz, Emma
    Rapid increases in the global human population have led to agricultural intensification worldwide, highlighting the need for conserving the remaining fragmented natural areas and incorporating more sustainable farming techniques. South Africa is currently the largest producer of macadamia nuts in the world and macadamia plantings are increasing rapidly, resulting in the loss of natural vegetation. The study area, Hazyview in Mpumalanga, is an agricultural mosaic, with fragments of natural vegetation patches in between. These natural vegetation patches are key to retaining important ecosystem services for the agricultural industry such as pollination and pest-control. This study evaluated how a land-use mosaic, represented by an agricultural landscape with natural vegetation retained around the farms, is related with macadamia nut yield and quality, the activity of insectivorous bat species, and nocturnal insect abundance, with a specific focus on five different macadamia farms. The role and contribution of bats and insects in agro-ecosystems are becoming more apparent, and their importance as indicator species which can reflect environmental changes in plant and insect communities is of key value. With this information, I could gain a better understanding of the relationships between the surrounding biodiversity and the commercial macadamia nut production industry, as well as the potential impact that agriculture-induced landscape fragmentation has on the biodiversity in the region. Firstly, on-farm characteristics (e.g. tree age, density, height) for five farms were used, aswell as their yield and quality measures and land-use types surrounding each farm to determine potential relationships between macadamia nut yield and/or quality and proximity to surrounding natural vegetation. Farm characteristics such as the height and density of the macadamia trees planted had a significant association with yield and quality of the nuts which both decreased with increasing tree height and density. Additionally, a characteristic of land-use type (in this case, extent of waterbodies such as rivers/ dams and wetlands surrounding the farms) was found to have an association with yield. The surrounding natural areas likely act as a reservoir for both insects and their predators, which could be beneficial for the neighbouring agricultural lands by improving pollination, yield and nut quality through biological pest-control. Secondly, the Hazyview landscape composition was assessed, using National land-cover data and buffer zones around each of the five farms. Within a 5 km radius, land-cover composition varied across the five farms, reaching high proportions of natural vegetation (64.83 %), agriculture and forestry plantations (55.78 %), settlements (10.78 %) and water bodies (1.78 %) in some areas. This indicates the extent of the remaining natural vegetation surrounding each farm. Bat echolocation calls (bat activity) and insect collection (insect abundance) were conducted on two study farms. Sample points were located at different edge sites (land-use types), natural vegetation, river and road edges, to determine whether bats (activity) show preferences for foraging over macadamia orchards bordering natural vegetation and if insect abundance follows the same trend. Corresponding with other studies, land-use type, specifically natural vegetation bordering the farms, was significantly associated with increased bat activity and insect abundance within this agricultural landscape, with natural vegetation edges being preferred and road edges being least favoured by both. Bats from all three foraging guilds (clutter, clutter-edge and aerial) were recorded during the study, which indicates that the Hazyview farming region still appears to support a high bat species richness. In conclusion, this study highlights the importance of natural vegetation within an agricultural land-use mosaic and the benefits it can bring, both to the farmer and the surrounding biodiversity. Although ecosystem services were not directly measured, it can be deduced from the association of higher yield and better crop quality with surrounding natural vegetation, where the presence of bat and insects was found to be higher. Given the future increase in agricultural land expansion, information must be made publicly available, to allow farmers to consider all beneficial ecosystem services that the surrounding natural vegetation has to offer, namely pollination and pest-control. Farm management practices, which are aligned with consideration for ecosystem services have the potential to improve ecosystem functioning and service provision across agricultural landscapes.