A standardised protocol for roadkill detection and the determinants of roadkill in the greater Mapungubwe Transfrontier Conservation Area, Limpopo Province, South Africa
- Authors: Collinson, Wendy Jane
- Date: 2013
- Subjects: Roadkill -- South Africa -- Mapungubwe Site , Traffic safety and wildlife -- South Africa -- Mapungubwe Site
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:5606 , http://hdl.handle.net/10962/d1002056 , Roadkill -- South Africa -- Mapungubwe Site , Traffic safety and wildlife -- South Africa -- Mapungubwe Site
- Description: Despite evidence suggesting that road traffic is a major threat to biodiversity loss, very little is known about its actual impact on wildlife populations in South Africa. Globally, road density and traffic volumes are increasing, and although huge budgets are devoted to the construction and upgrading of roads, there is little or no allocation to mitigation measures for protecting fauna in most countries, particularly Africa. Further, no global standardised protocol exists for the rapid assessment of roadkill or the most economical and efficient approach for assessing roadkill rates. Using vehicle field trials, the reliability of detecting artificially deployed roadkill was assessed. Roadkill detection rates decreased significantly at speeds >50 km/h and were also significantly influenced by light conditions (i.e. detection success was greater when the sun was high) and the position of the roadkill on the road (i.e. smaller roadkill on verges were often missed). These results suggest that roadkill sampling was most effective between 1.5 hours after dawn and 1.5 hours before dusk and that driving at slower speeds (<50 km.h⁻ₑ) was required to detect roadkill. This protocol was implemented across three ecological seasons on a 100 km paved road and a 20 km unpaved road in the Greater Mapungubwe Transfrontier Conservation Area, Limpopo Province, South Africa. Driven daily over a 120-day period (three periods consisting each of 40 days), a total of 1,027 roadkill were recorded. These comprised 162 species from all terrestrial vertebrate groups with birds being the most commonly encountered roadkill (50% of all incidents). The high numbers of vertebrates identified as roadkill suggests that road traffic could have potentially unsustainable impacts on wildlife populations and hence the biodiversity of the area. Seventeen variables were identified as possible determinants of roadkill occurrence with season, rainfall, minimum and maximum temperature, habitat type, grass height, grass density, fence type and vehicle type significantly influencing roadkill numbers. Significantly more roadkill were detected on the paved road (9.91/100km) than on the unpaved road (1.8/100km) probably because of greater traffic volumes and the increased speed that vehicles travelled on the paved road. Warmer temperatures and increased rainfall in the preceding 24 hours also increased road mortality numbers as animals tended to become more active during these times. Interestingly, more roadkill was detected in open roadside habitats compared to dense roadside habitats on both the paved and unpaved roads and when grass on the roadside verge was of intermediate height. Open habitat possibly may provide a natural corridor for wildlife which ultimately end up on the road. Roadkill numbers increased when certain other physical barriers, such as cattle fences, were present, probably because these barriers were more penetrable than electric fencing. A series of mitigation measures are proposed to reduce the impacts of roads on wildlife in South Africa. These mitigation measures highlight the need to address the balance between the development of a country’s transport infrastructure and the conservation of its fauna. It is important that research on the impacts of roads becomes standardised to enable robust statistical comparisons which will provide a greater understanding of the potential threats to vertebrate biodiversity
- Full Text:
- Date Issued: 2013
- Authors: Collinson, Wendy Jane
- Date: 2013
- Subjects: Roadkill -- South Africa -- Mapungubwe Site , Traffic safety and wildlife -- South Africa -- Mapungubwe Site
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:5606 , http://hdl.handle.net/10962/d1002056 , Roadkill -- South Africa -- Mapungubwe Site , Traffic safety and wildlife -- South Africa -- Mapungubwe Site
- Description: Despite evidence suggesting that road traffic is a major threat to biodiversity loss, very little is known about its actual impact on wildlife populations in South Africa. Globally, road density and traffic volumes are increasing, and although huge budgets are devoted to the construction and upgrading of roads, there is little or no allocation to mitigation measures for protecting fauna in most countries, particularly Africa. Further, no global standardised protocol exists for the rapid assessment of roadkill or the most economical and efficient approach for assessing roadkill rates. Using vehicle field trials, the reliability of detecting artificially deployed roadkill was assessed. Roadkill detection rates decreased significantly at speeds >50 km/h and were also significantly influenced by light conditions (i.e. detection success was greater when the sun was high) and the position of the roadkill on the road (i.e. smaller roadkill on verges were often missed). These results suggest that roadkill sampling was most effective between 1.5 hours after dawn and 1.5 hours before dusk and that driving at slower speeds (<50 km.h⁻ₑ) was required to detect roadkill. This protocol was implemented across three ecological seasons on a 100 km paved road and a 20 km unpaved road in the Greater Mapungubwe Transfrontier Conservation Area, Limpopo Province, South Africa. Driven daily over a 120-day period (three periods consisting each of 40 days), a total of 1,027 roadkill were recorded. These comprised 162 species from all terrestrial vertebrate groups with birds being the most commonly encountered roadkill (50% of all incidents). The high numbers of vertebrates identified as roadkill suggests that road traffic could have potentially unsustainable impacts on wildlife populations and hence the biodiversity of the area. Seventeen variables were identified as possible determinants of roadkill occurrence with season, rainfall, minimum and maximum temperature, habitat type, grass height, grass density, fence type and vehicle type significantly influencing roadkill numbers. Significantly more roadkill were detected on the paved road (9.91/100km) than on the unpaved road (1.8/100km) probably because of greater traffic volumes and the increased speed that vehicles travelled on the paved road. Warmer temperatures and increased rainfall in the preceding 24 hours also increased road mortality numbers as animals tended to become more active during these times. Interestingly, more roadkill was detected in open roadside habitats compared to dense roadside habitats on both the paved and unpaved roads and when grass on the roadside verge was of intermediate height. Open habitat possibly may provide a natural corridor for wildlife which ultimately end up on the road. Roadkill numbers increased when certain other physical barriers, such as cattle fences, were present, probably because these barriers were more penetrable than electric fencing. A series of mitigation measures are proposed to reduce the impacts of roads on wildlife in South Africa. These mitigation measures highlight the need to address the balance between the development of a country’s transport infrastructure and the conservation of its fauna. It is important that research on the impacts of roads becomes standardised to enable robust statistical comparisons which will provide a greater understanding of the potential threats to vertebrate biodiversity
- Full Text:
- Date Issued: 2013
The dispersal of African wild dogs (Lycaon Pictus) from protected areas in the Northern KwaZulu‐Natal Province, South Africa.
- Whittington‐Jones, Brendan Mark
- Authors: Whittington‐Jones, Brendan Mark
- Date: 2012
- Subjects: African wild dog , African wild dog -- Conservation -- South Africa -- KwazuluNatal , African wild dog -- Dispersal -- South Africa -- KwaZulu-Natal , African wild dog -- Effect of human beings on -- South Africa -- KwaZulu Natal
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:5893 , http://hdl.handle.net/10962/d1013414
- Description: The number of African wild dogs Lycaon pictus in Northern KwaZulu‐Natal, South Africa has increased substantially over the last six years. This is largely due to a managed metapopulation approach of introductions onto protected areas and private game reserves since 1998. Because of the increasing wild dog population, the likelihood of rural communities surrounding protected areas encountering dispersing wild dogs in northern KwaZulu‐Natal has also increased. Resident wild dog populations currently occur in Hluhluwe‐iMfolozi Park (HiP), Mkhuze Game Reserve (MGR) of Isimangaliso Wetland Park and Thanda Private Game Reserve (TPGR) all of which are bordered by a matrix of natural habitats and human settlements. Thus, land outside these protected areas could be utilized to expand wild dog distribution and provide connectivity between existing resident populations of wild dogs within KwaZulu‐Natal. To investigate the viability of such an approach, Maximum Entropy Modelling (Maxent) was used to characterize habitat niche selection of dispersing wild dogs, and to identify potential dispersal linkages between current wild dog metapopulation reserves. The model was calibrated using 132 location points collected from 2006 until 2009. From 2008 to 2009 I also conducted a survey of 247 community members in tribal authorities bordering HiP and MGR, to understand factors influencing attitudes towards wild dogs. A habitat suitability model with seven predictor variables had an AUC of 0.96 (SD = 0.02) and indicated four variables which best predicted probability of presence for dispersing wild dogs: elevation, road density, land cover and human density. The results suggest that elevation and land cover may be of greater influence for dispersing wild dogs than human density or activity. Elevation was the single most effective variable indicating a higher probability of presence for dispersing wild dogs in lower lying locations; peaking at approximately 200 – 300m a.s.l. Wild dogs also showed a preference for Woodland and Bushland habitats which in KwaZulu‐Natal tend to be found on lower lying topography. A preference for areas of lower human density and a highest probability of presence at road densities of approximately 0.7km/km2 or less would suggest that while wild dogs may show a tendency to avoid areas of high human activity, they can coexist in close proximity to humans. Respondent’s attitudes, knowledge of wild dogs and livestock husbandry were interpreted by the development of a set of indices. Attitudes were positively related to formal education levels and wild dog‐specific education, but were not influenced by demographic factors such as gender, age and employment status. Eighty three percent of respondents believed efforts to protect wild dogs should continue. Respondents with higher numbers of livestock tended to have more positive attitudes towards wild dogs despite generally incurring higher losses to carnivore depredation than those with less livestock. This appears to be because the financial burden of livestock losses to those with fewer livestock is perceived to be a loss of a greater proportion of total financial wealth. The study highlighted the substantial scope for improvement in livestock management. Theft (34 percent), drought (30 percent) and disease (14 percent) were ranked as the greatest problems facing livestock owners while predators were ranked as the greatest problem by only 4 percent of respondents. My findings suggest that wild dogs are generally viewed positively or with ambivalence. Concerns over the potentially increasing threat to livestock, as natural prey numbers outside protected areas decline and wild dogs disperse from natal packs, are likely to be manageable. However, wild dog population expansion within KwaZulu‐Natal will continue to rely on managed core populations on perimeter‐fenced metapopulation reserves with tolerant communities and landowners contributing to the connectivity of isolated reserves. Generation of tolerance can be instilled through continued advocacy and education, supported by conflict mitigation initiatives and rapid response to conflict reports. The implementation of incentive schemes for adjoining private landowners to co‐manage wild dog populations will need to be addressed in conjunction with managed metapopulation practices and law enforcement, to promote range expansion, and reduce potentially lethal edge effects and wild dog‐human conflict.
- Full Text:
- Date Issued: 2012
- Authors: Whittington‐Jones, Brendan Mark
- Date: 2012
- Subjects: African wild dog , African wild dog -- Conservation -- South Africa -- KwazuluNatal , African wild dog -- Dispersal -- South Africa -- KwaZulu-Natal , African wild dog -- Effect of human beings on -- South Africa -- KwaZulu Natal
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:5893 , http://hdl.handle.net/10962/d1013414
- Description: The number of African wild dogs Lycaon pictus in Northern KwaZulu‐Natal, South Africa has increased substantially over the last six years. This is largely due to a managed metapopulation approach of introductions onto protected areas and private game reserves since 1998. Because of the increasing wild dog population, the likelihood of rural communities surrounding protected areas encountering dispersing wild dogs in northern KwaZulu‐Natal has also increased. Resident wild dog populations currently occur in Hluhluwe‐iMfolozi Park (HiP), Mkhuze Game Reserve (MGR) of Isimangaliso Wetland Park and Thanda Private Game Reserve (TPGR) all of which are bordered by a matrix of natural habitats and human settlements. Thus, land outside these protected areas could be utilized to expand wild dog distribution and provide connectivity between existing resident populations of wild dogs within KwaZulu‐Natal. To investigate the viability of such an approach, Maximum Entropy Modelling (Maxent) was used to characterize habitat niche selection of dispersing wild dogs, and to identify potential dispersal linkages between current wild dog metapopulation reserves. The model was calibrated using 132 location points collected from 2006 until 2009. From 2008 to 2009 I also conducted a survey of 247 community members in tribal authorities bordering HiP and MGR, to understand factors influencing attitudes towards wild dogs. A habitat suitability model with seven predictor variables had an AUC of 0.96 (SD = 0.02) and indicated four variables which best predicted probability of presence for dispersing wild dogs: elevation, road density, land cover and human density. The results suggest that elevation and land cover may be of greater influence for dispersing wild dogs than human density or activity. Elevation was the single most effective variable indicating a higher probability of presence for dispersing wild dogs in lower lying locations; peaking at approximately 200 – 300m a.s.l. Wild dogs also showed a preference for Woodland and Bushland habitats which in KwaZulu‐Natal tend to be found on lower lying topography. A preference for areas of lower human density and a highest probability of presence at road densities of approximately 0.7km/km2 or less would suggest that while wild dogs may show a tendency to avoid areas of high human activity, they can coexist in close proximity to humans. Respondent’s attitudes, knowledge of wild dogs and livestock husbandry were interpreted by the development of a set of indices. Attitudes were positively related to formal education levels and wild dog‐specific education, but were not influenced by demographic factors such as gender, age and employment status. Eighty three percent of respondents believed efforts to protect wild dogs should continue. Respondents with higher numbers of livestock tended to have more positive attitudes towards wild dogs despite generally incurring higher losses to carnivore depredation than those with less livestock. This appears to be because the financial burden of livestock losses to those with fewer livestock is perceived to be a loss of a greater proportion of total financial wealth. The study highlighted the substantial scope for improvement in livestock management. Theft (34 percent), drought (30 percent) and disease (14 percent) were ranked as the greatest problems facing livestock owners while predators were ranked as the greatest problem by only 4 percent of respondents. My findings suggest that wild dogs are generally viewed positively or with ambivalence. Concerns over the potentially increasing threat to livestock, as natural prey numbers outside protected areas decline and wild dogs disperse from natal packs, are likely to be manageable. However, wild dog population expansion within KwaZulu‐Natal will continue to rely on managed core populations on perimeter‐fenced metapopulation reserves with tolerant communities and landowners contributing to the connectivity of isolated reserves. Generation of tolerance can be instilled through continued advocacy and education, supported by conflict mitigation initiatives and rapid response to conflict reports. The implementation of incentive schemes for adjoining private landowners to co‐manage wild dog populations will need to be addressed in conjunction with managed metapopulation practices and law enforcement, to promote range expansion, and reduce potentially lethal edge effects and wild dog‐human conflict.
- Full Text:
- Date Issued: 2012
- «
- ‹
- 1
- ›
- »