Impact of Land Subdivision and Sedentarization on Wildlife in Kenya’s Southern Rangelands

Subdivision and sedentarization of pastoral communities is accelerating rapidly across the African rangelands, posing a severe threat to wildlife populations, but few studies have looked quantitatively at the ecological impact of sedentarization. Here we look at the impact of sedentarization on wildlife by comparing ecologically matched subdivided and unsubdivided Maasai pastoral lands (ranches) in semiarid southern Kenya. We found no significant difference in livestock densities on the two ranches but there was a significantly higher wildlife density on the unsubdivided ranch, in both dry and wet seasons. Nonetheless, the unsubdivided ranch still had a higher percentage of grass biomass and ground cover and lower grazing pressure than the subdivided ranch. Distribution of homesteads (bomas) was mostly random on the subdivided ranch, with little area unaffected by human settlement. On the contrary, the unsubdivided ranch had a highly clumped boma distribution pattern, resulting in much of the land being relatively far from permanent human settlement. We show that the regular distribution and permanence of settlements following subdivision and sedentarization greatly reduces wildlife populations both through direct displacement and a reduction of forage. Relative to mobile pastoralism on open rangelands, sedentarization leads to reduced seasonal movements of livestock, lowered grass biomass, and slower grass recovery after very dry periods. This study points to the need to maintain mobile, large-scale herd movements to avoid the heavy impact on grasslands associated with sedentarization of pastoral settlement and herds.     

Perspective article: The evolving use of stocking rate indices currently based on animal number and type in semi‐arid heterogeneous landscapes and complex land‐use systems

Abstract

Methods of determining stocking rate based on metabolic mass, animal type and biomass are reviewed in the context of equilibrial/disequilibrial paradigms. In South Africa, the calculation of ‘carrying capacity’ is based on conversion of animal species to metabolic mass equivalents. This assumes homogenous systems that tend to some point of equilibrium. It is applied widely in commercial livestock systems involving one or two species. Examination of a case study in the Lowveld of the Northern Province, South Africa, showed that the determination of ‘stocking rate’ in this multi‐herbivore and heterogenous system, overestimated the ‘carrying capacity’ of the reserve over 20 years. The actual animal numbers in the system dropped by approximately 4 000 kg km^?2 after a drought in the early period of the study into the bounds as determined by a model incorporating rainfall and animal biomass. An approach to determine stocking density using animal type, biomass, rainfall and vegetation parameters is suggested. The development of this as a coarse‐scale (regional) and ranch‐specific model to cover a range of scales and heterogeneity in key resources is advocated.     

Cattle herd vulnerability to rainfall variability: responses to two management scenarios in southern Ethiopia

We examine how the system of grazing management of cattle in savanna rangelands affects the herd response to drought. We have used long-term time series data to evaluate the effects of management on drought-induced cattle mortality using traditional livestock management practices. There was no control of stocking densities, as compared to a government ranch where stocking densities would be adjusted in accordance with available pasture. We tested the responses under two scenarios. Scenario 1: Response of cattle herds to inter-annual rainfall variability (IRV) under a regulated grazing management system; this provides more reliable predictions of cattle population and performance in terms of herd mortality and calving rates than does the communal land use system. Scenario 2: Regardless of the management system, similar trends in cattle populations will be observed in response to IRV. The results of the study showed that fluctuations in cattle numbers, herd mortality and calving rates were highly correlated with IRV, with stronger linear impacts in accordance with scenario 2. In both management systems, cattle herd sizes and calving rates declined during periods of drought, followed by slow recovery. Cattle populations in Borana rangelands in southern Ethiopia did not recover for a period of two decades. We conclude that a management system based on control of stocking densities did not improve herd survival, as compared with traditional drought management strategies. This contradicts common expectations. Increased drought frequencies aggravated cattle mortality and lowered calving rates. The implication of the findings is that regardless of adjusted stocking density, livestock populations in the arid savanna ecosystems of southern Ethiopia remain at risk from climate change.     

Disease incidence in ranch and pastoral livestock herds around Lake Mburo National Park,in South Western Uganda

A longitudinal study was done to monitor incidence of livestock diseases in ranch and pastoral herds around Lake Mburo National Park, in South-western Uganda with a high level of wildlife (impala and zebra) and livestock (cattle and goats) interactions. East Coast Fever (ECF), abortions, helminthosis and starvation due to drought were major livestock disease problems encountered. ECF was a major disease affecting mainly calves, causing mortality both in ranch (8.5%) and pastoral (8.2%) herds. Meanwhile, drought was a more serious in pastoral than ranch herds (p < 0.05), killing 28.5 % and 5.8% of adult female cattle in respective production systems. Other endemic diseases found and controlled by vaccination were FMD, CBPP and blackquarter. In both cattle and goats helminthosis was causing mortalities, 3.3% and 0.8% among calves in pastoral and ranches respectively; and 4.8% and 6.9% among kids and sub-adult goats respectively. Many cases of abortions occurred among pastoral herds. In conclusion, the disease incidence among livestock was the same or relatively low as compared to other areas without wildlife.     

Cattle Drive Salmonella Infection in the Wildlife–Livestock Interface

The genus Salmonella is found throughout the world and is a potential pathogen for most vertebrates. It is also the most common cause of food‐borne illness in humans, and wildlife is an emerging source of food‐borne disease in humans due to the consumption of game meat. Wild boar is one of the most abundant European game species and these wild swine are known to be carriers of zoonotic and food‐borne pathogens such as Salmonella. Isolation of the pathogen, serotyping and molecular biology are necessary for elucidating epidemiological connections in multi‐host populations. Although disease management at population level can be addressed using a number of different strategies, such management is difficult in free‐living wildlife populations due to the lack of experience with the wildlife–livestock interface. Herein, we provide the results of a 4‐year Salmonella survey in sympatric populations of wild boar and cattle in the Ports de Tortosa i Beseit National Game Reserve (NE Spain). We also evaluated the effects of two management strategies, cattle removal and increased wild boar harvesting (i.e. by hunting and trapping), on the prevalence of the Salmonella serovar community. The serovars Meleagridis and Anatum were found to be shared by cattle and wild boar, a finding that was confirmed by 100% DNA similarity patterns using pulse field gel electrophoresis. Cattle removal was more efficient than the culling of wild boar as a means of reducing the prevalence of shared serotypes, which underlines the role of cattle as a reservoir of Salmonella for wild boar. To our knowledge, this is the first attempt to manage Salmonella in the wild, and the results have implications for management.     

Interest in Participating in Integrated Wildlife/Livestock Land Uses in a Rural Area in South Africa

Wildlife ranching is the fastest growing industry in South Africa, with most ranches having been converted from large-scale livestock farms. This article investigated perceptions and attitudes of rural communities located in wildlife-rich rural areas toward integrated wildlife/livestock land uses, and analyzed socio-demographic factors influencing household interest in these land uses. Data were collected from 130 households selected through stratified random sampling techniques from nine villages adjacent to the Kruger National Park. Results showed a general interest in pursuing wildlife land uses, but differences in attitudes toward wildlife between households with cattle and without. Multinomial regression results indicated that age, education level, monthly income, and cattle ownership significantly impacted the odds of a household being interested in participation or not. Addressing concerns about human–wildlife conflict and clarifying potential benefits to households and the community is necessary for increased interest in integrated land uses.     

The sero‐epidemiology of Coxiella burnetii (Q fever) across livestock species and herding contexts in Laikipia County,Kenya

Coxiella burnetii, the causative agent of Query fever (Q fever), is among the most highly infectious zoonotic pathogens transmitted among livestock, with chronic effects challenging to veterinary and medical detection and care systems. Transmission among domestic livestock species can vary regionally due to herd management practices that determine which livestock species are raised, whether or not livestock are in contact with wildlife, and the susceptibility of these livestock to infection. To explore how different livestock management practices are associated with the risk of infection in multispecies environments, we carried out a comparative study of three types of herd management systems in the central Kenyan county of Laikipia: agro‐commercial, mixed conservancy/commercial, and smallholder ranches. We tested C. burnetii antibody seroprevalence in four common livestock species. Across all management types, the highest seroprevalence was in camels (20%), followed by goats (18%), sheep (13%), and cattle (6%). We observed a lower odds of testing seropositive for young compared to adult animals (adjusted OR = 0.44 [95% CI 0.24, 0.76]), and for males compared to females (adjusted OR = 0.52 [95% CI 0.33, 0.80]). Animals from mixed conservancy/commercial and smallholder operations had a higher odds of testing seropositive compared to animals from agro‐commercial ranches (adjusted OR = 5.17 [95% CI 2.71, 10.44] and adjusted OR = 2.21 [95% CI 1.17, 4.43] respectively). These data suggest that herd management practices might affect the transmission dynamics of C. burnetiiin arid African ecosystems like those seen in Kenya where several transmission modes are possible, risk of drought has promoted new livestock species such as camels, and multiple wildlife species may co‐occur with livestock on the landscape. Further longitudinal studies are needed to disentangle the mechanisms underlying these patterns, and further explore transmission patterns between wildlife, domestic animal, and human populations.     

Effectiveness of electro-fencing for restricting the ranging behaviour of wildlife: a case study in the degazetted parts of Akagera National Park

Several national parks on the African continent have been fenced to mitigate human–wildlife conflict, but the effectiveness of different fencing strategies seems to vary between regions. Here, we assessed the effectiveness of electro-fencing for restricting movement patterns of several wild ungulates in Akagera National Park (NP), Rwanda. Following the turmoil of the Rwandan genocide and civil war (1991–1995), large numbers of returning war refugees brought about an increasing pressure on land use, after which considerable parts of the western Akagera NP and adjacent Mutara Game Reserve were degazetted, followed by the construction of an electric fence in 2013 and a helicopter-based game-drive of ungulates into the modern park. We conducted distance sampling of wildlife and domestic livestock in the Uruhita plain – now no longer part of Akagera NP but bordering the modern Akagera NP to the west – for extended periods of time (2010–2015) before and after fencing. For abundant species population densities could be calculated, whereas for other species encounter frequencies are reported. We report on continuously high cattle densities, and increasing encounter frequencies of sheep/goats in the Uruhita plain. We found decreasing densities of zebra (Equus quagga) and decreasing encounter frequencies of waterbuck (Kobus ellipsiprymnus), stagnant encounter frequencies in the case of topi (Damaliscus lunatus jimela) and stagnant densities of impala (Aepyceros melampus). The latter species can jump fences, and we argue that competitive release and reduced predation may be the prime factors leading impala to (temporarily) exploit grasslands outside the modern national park.     

Using State and Transition Models to Determine the Opportunity Cost of Providing Ecosystem Services

A ranch-level model using state-and-transitions models for three ecological sites is used to determine the trade-offs of providing various ecosystem services. The hypothetical ranch is located in northern Colorado and is based upon area average ecological site characteristics and livestock production practices. Management decisions include stocking rate and brush control. The model includes exogenous factors such as precipitation and fire. The model solves for optimal decisions over an infinite planning horizon using stochastic dynamic programming. Results show that a ranch cannot provide all ecosystem services in tandem at their highest level, implying that land managers must decide which ecosystems service they want to provide. Also, it is much cheaper in terms of foregone profitability for a ranch to continue to provide a specific ecosystem service that is already provided by the ranch rather than try to transition the ranch to a new ecological state in order to provide a service currently not provided.     

Effects of rainfall,competition and grazing on flowering of Osteospermum sinuatum (Asteraceae) in arid Karoo rangeland

The shrub Osteospermum sinuatum, an important fodder plant in Karoo rangelands, flowered after heavy rain in autumn, winter and spring. The number of flowers produced per bush was positively conelated with basal stem diameter and rainfall in the 12 weeks before anthesis. Flowering increased when neighbouring plants of all species were removed. Grazing during flowering and seed set, by sheep stocked at recommended densities, reduced potential seed set by as much as 90%. Birds, hares and small antelope consumed 10–50% of the flowers. Size‐class distributions indicated that little recent recruitment had taken place on a ranch where palatable plants were scarce and where O. sinuatum flower production was severely depressed by grazing sheep.     

Wildlife or livestock? New directions for developing communal rangelands in South Africa

The paper agrees with three fundamental points raised by Vetter (in this issue), whilst highlighting an emerging trend in wildlife land use, which should be considered in policy making. Firstly, the paper supports the argument that communal rangelands are important as objects of biodiversity conservation and, secondly, the argument that livestock production alone is not sufficient as a mechanism for alleviating poverty, but rather livelihoods diversification and multisectoral strategies are required for this purpose. Lastly, the paper supports the need for mechanisms to ensure tenure security for sustainable use of communal rangelands. I argue that the historical shift to wildlife land use previously observed on private farms in South Africa is now inevitable on communal rangelands and will require explicit consideration in the policy on communal rangelands. Although such development in the communal rangelands has potential to meet the important goals of livelihoods diversification as indicated by Vetter, it also has potential to threaten existing rural livelihoods. The paper suggests several issues to be considered as wildlife becomes part of the rural landscape in some areas, in addition to issues raised by Vetter relevant to policy making.     

Using Range Condition Assessment to Optimize Wildlife Stocking in Tindress Wildlife Sanctuary,Nakuru District,Kenya

Over 70% of Kenya’s wildlife resources occur outside protected areas, in areas where land use practices do not necessarily conform to wildlife conservation standards. Ensuring that land use practices in these areas accommodate wildlife conservation is vital in effectively conserving wildlife in this country. Tindress Farm in Rift Valley offers a good example of a place where economic activities and wildlife conservation can work harmoniously. The farm has set up a 320-ha wildlife sanctuary in the hilly parts of the property to provide a haven for wildlife displaced by human settlements in the surrounding environs. The Tindress Farm management needed to know the diversity and optimum number of wildlife species that the sanctuary could accommodate. This study set out to 1) outline a set of models for objectively calculating wildlife stocking levels and 2) demonstrate the practical use of these models in estimating optimum stocking levels for a specific wildlife sanctuary. After comparing models using forage inventory methods models and utilization-based methods (UM), we opted to use UM models because of their focus on ecological energetics. This study established that the range condition in Tindress Wildlife Sanctuary varied from poor to good (29–69%) and recommended a total stocking density of 158.9 grazer units and 201.4 browser units shared out by the various herbivore species. These estimates remain a best-case scenario. The effects of rainfall, range condition, and condition of the animals should be monitored continuously to allow for adjustments through active adaptive management.     

Presidential Address — 2002: The Grassland Society of Southern Africa should be striving for greater synergy with the farming community

'Let me assure you all that our commitment to the preservation of wildlife and the habitat in the Maasai Mara reserve is total. We do not deny that the reserve is of vital economic importance to us, but we also firmly believe, and proclaim, that wildlife deserves to be protected for its own sake. The animals are God's creatures and we affirm their right to live and survive as undisturbed as possible in this busy and over-populated world. The Mara can provide a relief from the commercial pressures of profit and survival. We must thus emphasize that in the Maasai Mara reserve interests of wildlife are paramount.' Samson Ole Tuya, Chairman Narok County Council — Maasai Mara Tourism Workshop, March 1991

'The chances of wildlife surviving when people are hungry and poor are becoming increasingly slim.' Samson Ole Tuya, Member of Parliament, Narok South. Workshop — Defining the framework for management and conservation planning for the Maasai Mara ecosystem, November 1996

The Maasai Mara ecosystem has one of the richest assemblages of wildlife in the world and supports about 237 herbivores per km², making it one of the most productive natural terrestrial ecosystems. It maintains the greatest and most spectacular large mammal migration on earth and is the most highly visited tourist area in the East African region. Tourism generates millions of dollars annually while livestock and agriculture contribute significantly to the local and national economy. This paper examines the enormous threats facing the ecosystem and the efforts being made to keep the ecosystem open in order to support a vibrant tourist trade and a healthy livestock economy. These aspects are currently threatened by a burgeoning human population, land subdivision, changing land tenure systems, crop farming, poaching, habitat fragmentation, blockage of wildlife migratory corridors, habitat destruction, sedentarisation, environmental degradation and poverty. Wildlife conservation is in conflict with agriculture, while tourism has taken away the best grazing areas for livestock. The paper also reviews the ecological and socio-economic problems associated with tourism, land use changes and break down of traditional governance structures and how they have weakened the capacity of Maasai Mara to sustain a viable wildlife populations.     

Integrating small mammal community variables into aircraft–wildlife collision management plans at Namibian airports

Understanding ecosystems within and around airports can help to determine the causes and possible mitigation measures for collisions between aircraft and wildlife. Small mammal communities are an important component of the semi‐arid savanna ecosystems of Namibia, its productivity and its ecosystem integrity. They are also a major direct attractant for raptors at airports. The present study compared the abundance and diversity of small mammals between Namibia's 2 main airport properties (Hosea Kutako International Airport and Eros Airport), and among areas of land used for various purposes surrounding the airports. A total of 2150 small mammals (3 orders, 11 species) were captured over 4 trapping seasons. Small mammal abundance was significantly higher at the end of the growing season than during the non‐growing season. The grass mowing regimen in current management plans at the airports resulted in a significant reduction of small mammal abundance at Hosea Kutako during the non‐growing season only, thus indicating that annual mowing is effective but insufficient to reduce the overall abundance of mammal prey species for raptors. Small mammal numbers were significantly higher at Hosea Kutako Airport compared to the cattle and game farming land surrounding the airport, while no differences in small mammal densities or diversity were found for areas with different land uses at and surrounding Eros. The study suggests that the fence around Hosea Kutako provides a refuge for small mammals, resulting in higher densities. It also indicates that different surrounding land use practices result in altered ecosystem function and productivity, an important consideration when identifying wildlife attractants at airports.     

Cluster Analysis of Campylobacter jejuni Genotypes Isolated from Small and Medium‐Sized Mammalian Wildlife and Bovine Livestock from Ontario Farms

Using data collected from a cross‐sectional study of 25 farms (eight beef, eight swine and nine dairy) in 2010, we assessed clustering of molecular subtypes of C. jejuni based on a Campylobacter‐specific 40 gene comparative genomic fingerprinting assay (CGF40) subtypes, using unweighted pair‐group method with arithmetic mean (UPGMA) analysis, and multiple correspondence analysis. Exact logistic regression was used to determine which genes differentiate wildlife and livestock subtypes in our study population. A total of 33 bovine livestock (17 beef and 16 dairy), 26 wildlife (20 raccoon (Procyon lotor), five skunk (Mephitis mephitis) and one mouse (Peromyscus spp.) C. jejuni isolates were subtyped using CGF40. Dendrogram analysis, based on UPGMA, showed distinct branches separating bovine livestock and mammalian wildlife isolates. Furthermore, two‐dimensional multiple correspondence analysis was highly concordant with dendrogram analysis showing clear differentiation between livestock and wildlife CGF40 subtypes. Based on multilevel logistic regression models with a random intercept for farm of origin, we found that isolates in general, and raccoons more specifically, were significantly more likely to be part of the wildlife branch. Exact logistic regression conducted gene by gene revealed 15 genes that were predictive of whether an isolate was of wildlife or bovine livestock isolate origin. Both multiple correspondence analysis and exact logistic regression revealed that in most cases, the presence of a particular gene (13 of 15) was associated with an isolate being of livestock rather than wildlife origin. In conclusion, the evidence gained from dendrogram analysis, multiple correspondence analysis and exact logistic regression indicates that mammalian wildlife carry CGF40 subtypes of C. jejuni distinct from those carried by bovine livestock. Future studies focused on source attribution of C. jejuni in human infections will help determine whether wildlife transmit Campylobacter jejuni directly to humans.     

Interspecific variation in wildlife responses to cattle,swine and chicken feed in the forests surrounding poultry farms

To strengthen farm biosecurity, wildlife behaviors around livestock environments require significant attention. Livestock feed is considered one of essential factors that attract wildlife to the livestock environment. We experimentally studied wildlife response to cattle, swine, and chicken concentrated feeds in the forests surrounding poultry farms. In 14 feeding sites, four feed conditions were established: without feed (control); cattle feed; chicken feed; and swine feed. Wildlife behaviors at each feed point were monitored using infrared cameras. In 3,175 videos, 10 mammals were photographed on 10 or more occasions. Wildlife more frequently appeared at the points with feed than without feed. In addition, the number of videos that captured foraging or interest behaviors was largest for swine feed, followed by chicken feed, then cattle feed. There was a large difference among wildlife in their response to livestock feeds, although each species did not have a strong preference for a specific feed. Livestock feeds invite frequent visits by high and moderate response groups, especially omnivores and carnivores with omnivorous tendencies. Therefore, to protect against such wildlife intrusion, leftover feed and feed storage must be properly managed. This study also suggests that livestock feeds may not cause intrusions by rare response group species; hence, if their intrusions occur, they may be due to factors other than livestock feed. The study situation can partly reflect actual feed-stealing situations. The results will contribute to consider the properly management to protect livestock environments from wildlife intrusions.     

Molecular and Statistical Analysis of Campylobacter spp. and Antimicrobial‐Resistant Campylobacter Carriage in Wildlife and Livestock from Ontario Farms

The objectives of this study were to (i) compare the carriage of Campylobacter and antimicrobial‐resistant Campylobacter among livestock and mammalian wildlife on Ontario farms, and (ii) investigate the potential sharing of Campylobacter subtypes between livestock and wildlife. Using data collected from a cross‐sectional study of 25 farms in 2010, we assessed associations, using mixed logistic regression models, between Campylobacter and antimicrobial‐resistant Campylobacter carriage and the following explanatory variables: animal species (beef, dairy, swine, raccoon, other), farm type (swine, beef, dairy), type of sample (livestock or wildlife) and Campylobacter species (jejuni, coli, other). Models included a random effect to account for clustering by farm where samples were collected. Samples were subtyped using a Campylobacter‐specific 40 gene comparative fingerprinting assay. A total of 92 livestock and 107 wildlife faecal samples were collected, and 72% and 27% tested positive for Campylobacter, respectively. Pooled faecal samples from livestock were significantly more likely to test positive for Campylobacter than wildlife samples. Relative to dairy cattle, pig samples were at significantly increased odds of testing positive for Campylobacter. The odds of isolating Campylobacter jejuni from beef cattle samples were significantly greater compared to dairy cattle and raccoon samples. Fifty unique subtypes of Campylobacter were identified, and only one subtype was found in both wildlife and livestock samples. Livestock Campylobacter isolates were significantly more likely to exhibit antimicrobial resistance (AMR) compared to wildlife Campylobacter isolates. Campylobacter jejuni was more likely to exhibit AMR when compared to C. coli. However, C. jejuni isolates were only resistant to tetracycline, and C.  coli isolates exhibited multidrug resistance patterns. Based on differences in prevalence of Campylobacter spp. and resistant Campylobacter between livestock and wildlife samples, and the lack of similarity in molecular subtypes and AMR patterns, we concluded that the sharing of Campylobacter species between livestock and mammalian wildlife was uncommon.