Kori bustard (Ardeotis kori struthiunculus) occurrence in the Serengeti grass plains, northern Tanzania
Abstract
enThe kori bustard (Ardeotis kori struthiunculus) is indigenous to grasslands and lightly wooded savannahs of southern and eastern Africa. The species is categorized as near threatened in its entire range due to anthropogenic factors and low reproductive rates. The aim of this study was to analyse the impact of grass colour, grass height, season and location on the density/occurrence of this bird species in the Serengeti grass plains, Tanzania. Data were collected from January 2014 to June 2015 using transect counts in four seasons: (i) short dry, (ii) long rain, (iii) long dry and (iv) short rain seasons, respectively. The mean density of kori bustard in the grass plains was 0.25 ± 1.01 per 0.2 km2 with near-significant differences among the study sites. The occurrence of kori bustard was high in the medium height (11–30 cm) during the long rain and short dry seasons. The kori bustard density is relatively low, and the distribution varies with grass height and season. We suggest that conservation efforts should be directed at preventing its local extinction by protecting the habitat from excessive human activities, such as livestock grazing and illegal offtake.
Résumé
frL'outarde kori (Ardeotis kori struthiunculus) fréquente les prairies et les savanes peu boisées d'Afrique de l'Est et du Sud. L'espèce est classée « quasi menacée » dans la totalité de son aire de répartition à cause de facteurs anthropiques et d'un faible taux de reproduction. Le but de cette étude était d'analyser l'impact de la couleur et de la hauteur de l'herbe, de la saison et de la localisation sur la densité/la présence de cette espèce dans les plaines herbeuses du Serengeti, en Tanzanie. Nous avons collecté des données entre janvier 2014 et juin 2015 en faisant des comptages sur transects pendant quatre saisons : (i) petite saison sèche, (ii) grande saison des pluies, (iii) grande saison sèche et (iv) petite saison des pluies. La densité moyenne d'outardes kori dans les plaines herbeuses était de 0.25 ± 1.01/0.2 km² avec des différences quasi significatives entre les sites étudiés. La fréquence des outardes kori était élevée dans les herbes de hauteur moyenne (11–30 cm) pendant la longue saison des pluies et la petite saison sèche. La densité d'outardes kori est relative faible, et leur distribution varie avec la hauteur de l'herbe et les saisons. Nous suggérons d'orienter les efforts de recherche vers la prévention de leur extinction locale, en protégeant l'habitat contre l'excès d'activités humaines, tels le pâturage du bétail et les prélèvements illégaux.
Introduction
The future survival of wildlife is threatened by increasing human pressure due to high human population increases in many countries in sub-Saharan Africa (Abbitt, Scott & Wilcove, 2000; Ceballos & Ehrlich, 2002; Balmford, Green & Jenkins, 2003; Gaston, Blackburn & Goldewijk, 2003; Harcourt & Park, 2003). For a significant period of time, habitat fragmentation and loss has caused displacement and extinction of threatened bird species in various ecosystems (Alroy, 2001; McKee et al., 2004; Senyatso, Collar & Dolman, 2013).
The largest kori subspecies (Ardeotis kori) is indigenous to the grasslands and lightly wooded savannahs of southern and eastern Africa. The nominate subspecies Ardeotis kori kori occurs in Botswana, Zimbabwe, Namibia, southern Angola, South Africa and Mozambique (Johnsgard, 1991), while Ardeotis kori struthiunculus occurs in Ethiopia, Kenya and Tanzania. Seemingly, the species has undergone considerable population decline in all its range states, except in Zambia and Angola. In East Africa, the population of Ardeotis kori struthiunculus had declined by 21% by the late 19th century (Senyatso, Collar & Dolman, 2013). The species is listed in Appendix II of CITES and categorized as near threatened by the IUCN (BirdLife-International, 2013).
Kori bustard (Ardeotis kori spp.) populations are dramatically declining in both open and protected areas that are close to human settlements due to illegal hunting, habitat loss, low reproductive rates, the disappearance of large mammals and collision with power lines (Dale, 1990; Collar, 1996; Lawes, Fly & Piper, 2006; Thiollay, 2006; Manu, Peach & Cresswell, 2007; Magige et al., 2009). Fifteen years ago, the kori bustard population in South Africa was estimated to number 2000–5000 birds (Anderson, 2000), although the numbers are thought to be decreasing throughout southern Africa (Senyatso, Collar & Dolman, 2013). Shaw (2013) found that the annual mortality of kori bustard in the Nama Karoo district alone, due to collision with power lines, was approximately one-seventh of the South African population, that is 720 individuals. In the early 1970s, the kori bustards in Africa were one of the most characteristic elements of the northern Sahelian avifauna, except in the heart of Tenere; however, by 2004, no single bustard was observed (Thiollay, 2006). Increased human population adjacent to protected areas has shown to cause pressure on wildlife through illegal hunting (Herremanns, 1998; Loibooki et al., 2002; Kideghesho, 2006). In Burkina Faso, studies have revealed that the main factors correlated with large bird extinction or decline in protected areas were hunting, habitat degradation and fragmentation due to intensive cattle grazing, wood cutting and farming (Thiollay, 2006). A similar situation was observed in the Serengeti ecosystem where the kori bustard population is declining due to habitat degradation and hunting (Mmassy & Røskaft, 2014).
Behaviour and density are basic aspects of a species' ecology, but remain poorly known for the free-ranging kori bustard and little data exist on wild birds (Mwangi, 1988; Osborne & Osborne, 1998; Hallager & Boylan, 2004; Hallager & Lichtenberg, 2007; Lichtenberg & Hallager, 2008).
Birds' density is often positively correlated with habitat suitability and a sign of minimum anthropogenic disturbance (Huhta, Mappes & Jokimaki, 1996; Martin, 1998; Clark & Shutler, 1999; Robertson & Hutto, 2006). Serengeti National Park holds the largest natural grasslands in the world and is famous for its large aggregates of mammals and birds. However, the bird community structure of these grasslands has been poorly studied (Gottschalk, Ekschmitt & Bairlein, 2007). Maintaining the Serengeti National Park grasslands will provide critical refuges for grassland birds, including kori bustards, Karamoja apalis (Apalis karamojae), Francolins, secretary bird (Sagittarius serpentarius) and all vulture species, all of which have shown disturbing declines in recent decades (Herremanns, 1998; Sinclair, Mduma & Arcese, 2002).
Therefore, for effective management and conservation action of free-ranging kori bustard in the Serengeti plains, a reliable population estimate is required (Braun, 2005; Marques et al., 2013). The aim of this study was to identify vital environmental factors that influence the occurrence and density of kori bustard within four study sites in the Serengeti grass plains.
In this study, we predicted that (i) because we performed our fieldwork in similar habitats, the density of kori bustard will be similar within the four study areas with no seasonal differences, and (ii) kori bustard will be equally distributed, regardless of differences in grass height and colour.
Materials and methods
Study area
This study was conducted in the Serengeti grass plains in an area located on the eastern Serengeti National Park and western Ngorongoro Conservation Area grass plains, Tanzania, between 1o and 3o30' S and 34o and 36o E (Figure 1). Serengeti National Park (14,763 km2) is characterized by highland savannahs dominated by thorny woodland trees (Acacia, Commiphora, Ficus, Combretum and Podocarpus) and extensive grass plains (Herlocker, 1976). However, the study sites in the eastern Serengeti and western Ngorongoro plains are dominated by grasslands. The climate is warm and dry with mean annual temperatures between 15°C and 27°C (Sinclair, 1995).
Rainfall on the Serengeti plains is bimodal, and the long rain season occurs between March and May and the short rain season between November and January (Norton-Griffiths, Herlocker & Pennycuick, 1975). Rainfall is generally lower in the south and east (500 mm year−1) of the ecosystem than in the north and west (950–1150 mm year−1). Vegetation cover in Serengeti National Park is influenced mainly by soil type and rainfall and can be broadly classed into eastern grass plains, central acacia woodlands and northern broadleaf forests (Sinclair, 1995).
The Ngorongoro Conservation Area (8094 km2) spans vast expanses of highland plains, savannah, savannah woodlands and forests, from the plains of the Serengeti National Park in the north-west, to the eastern arm of the Great Rift Valley. The southern and eastern slopes of the highlands receive high rainfall, but the western plains lie in the mountains' rains shadow and receive less precipitation.
In the Serengeti National Park, human activities are prohibited and contrast that of Ngorongoro Conservation Area in which human activities and settlements are allowed. The Ikoma Open Area allows human settlement and some human activities, including livestock keeping, hunting by residents, farming and firewood collection. The neighbouring Ikorongo and Grumeti Game Reserves are partially protected and only allow trophy hunting.
Data collection
The data were collected at four sites including Naabi–Ngorongoro–Ndutu South (south-west, SW), Naabi–Ngorongoro–Ndutu North (north-west, NW), Maasai–Barafu–Golkopjes North (north-east, NE) and Maasai–Barafu–Golkopjes South (south-east, SE) (Figure 1). The study areas are predominantly treeless and dominated by short grass with a maximum height of 30 cm, along with some small trees and shrubs in a few kopje areas. Kori bustard observations were recorded on a monthly basis from January 2014 to June 2015. The study was divided into four seasons: (i) short dry season (January – March), (ii) long rain season (April – June), (iii) long dry period (July – September) and 4) short rain season (October – December). The study sites were selected because of previous anecdotal observations of large numbers of kori bustards during specific periods.
We collected data by driving a vehicle slowly (20 km per h) in randomly selected 1-km-long transects (n = 1897). All birds within 100 m of each side of the observation vehicle were counted; thus, each transect covered 0.2 km². When a 1-km transect was completed, another 1 km was driven without making any recordings. All kori bustards observed within each transect line were recorded, including those that flew away but were within transects at the start of the observation. Kori observations within transects were performed by visual scanning of the immediate surroundings, and Zeiss binoculars (10x40, field 7.10) when required, while a Leica CRF 900 (200 m) rangefinder was used to determine the distance from the vehicle. The censuses were conducted in all four study sections (NW, SW, NE and SE). A minimum of 25 transects were conducted daily for four days, starting at the beginning of every month and including a minimum of 100 transects per month. Kori bustards were also recorded outside the transects, but these observations were not included in the analysis. Grass height was initially measured by tape measure (100 cm) followed by visual estimates of grass height over the entire area of observation.
Data analyses
Because kori bustard density data were not normally distributed, we log-transformed the data to make them normally distributed. Before log transformation, we added the integer 1 to all observations to avoid transformation of zeros. Therefore, all statistical tests are conducted with log-transformed data.
We determined the density of kori bustards in relation to the following variables: grass colour (green = 100% green, greenish ˃50% green, brown = 0% green); height of grass was described as very short (less than 10 cm), medium height (11–30 cm) and tall (>30 cm); season and study site (SW, SE, NW and NE). These four areas were the result of dividing the study area of Figure 1 into four equal sites.
We calculated the total density of kori bustard per 0.2 km² (i.e. all koris observed on both sides of the vehicle during the 1-km transect) and compared the mean density among the study sites. Furthermore, an analysis of variance was used to compare the mean kori bustard density in relation to grass colour, grass height and season in the four study areas. Chi square (χ²) tests were used to compare variations of kori bustard observation with season and habitat, that is grass height and colour. Linear regression analyses were used to relate the density of kori bustard to the four study sites, grass height and colour at different seasons.
Results
The kori bustards were regularly seen in the same area, and most observations were of single birds (70.2%). The mean density of kori bustard in the Serengeti grass plains was 0.25 individuals per 0.2 km² (±1.01, SD, n = 1897, not log-transformed data). The density was 0.12 individual/0.2 km² (±0.48, SD, n = 314) in the NE area, 0.24 individual/0.2 km² (±0.86 SD, n = 264) in the SW area, 0.28 individual/0.2 km² (±1.28 SD, n = 514) in the SE area and 0.28 individual/0.2 km² (±1.01 SD, n = 805) in the NW area. There was no statistically significant difference in kori bustard densities among the four study areas (log-transformed data: ANOVA; F = 2.423, df = 3 and 1893, P = 0.064).
Although the most common colour of the grass was green when it was longer than 10 cm, a significant difference in the frequencies of green colour existed in different grass height groups (χ² = 165.5, df = 4, P < 0.001; Table 1). A significant difference in kori bustard density existed in transects with different grass heights (log-transformed data: ANOVA; F = 7.527, df = 2 and 1894, P < 0.001), with the highest densities in short grass habitats ranging from 11 to 30 cm. Statistically significant variations in densities also occurred with grass colour (log-transformed data: ANOVA; F = 5.319, df = 2 and 1896, P = 0.005).
| Grass colour | Grass height | Number of observations | ||
|---|---|---|---|---|
| 0-10 cm | 11-30 cm | >30 cm | ||
| Green | 238 | 346 | 4 | 588 |
| Greenish | 244 | 158 | 11 | 413 |
| Brown | 649 | 242 | 5 | 896 |
| N total | 1131 | 746 | 20 | 1897 |
Kori bustard densities differed highly significantly among the four different seasons (log-transformed data: ANOVA; F = 12.382, df = 3 and 1893, P < 0.001). The density was highest during the long rain season (April–June: 0.33 individual/0.2 km² ±1.40 SD, n = 610) and the short dry season (January–March: 0.30 individual/0.2 km² ±0.92 SD, n = 677), while the density was the lowest during the long dry season (July–September: 0.01 individual/0.2 km² ±0.12 SD, n = 208) and the short rain season (October–December: 0.14 individual/0.2 km², ±0.58, SD, n = 402).
A linear regression analysis with kori bustard density (log-transformed) as the dependent variable and area, season, grass height and grass colour as independent variables was significant (r2 = 0.017; ANOVA F = 7.989, df = 4 and 1892, P < 0.001). However, only season (B = −0.013, t = −3.645, P < 0.001) and area (B = 0.006, t = 1.961, P = 0.050) were significant contributors in explaining the density variation. Grass length (P = 0.409) and grass colour (P = 0.220) did not contribute significantly in explaining this variation.
Discussion
The results indicate that kori bustard density in the Serengeti grass plains is approximately 0.25 individual per 0.2 km2, or approximately 1.3 individuals per km2. The mean density did not differ among the four different study sites. According to studies by Frith (1973) and Isakov (1974) in Australia and India, intensive land-use practices have been part of Ardeotis australis' decline in parts of Australia, as well as the decline of Otis tarda in Europe. The few individuals observed in the Serengeti grass plains may be connected to intensive land-use practices in adjacent communities, as well as hunting (Mmassy & Røskaft, 2014). We must emphasize, however, that there is no previous information available about kori bustard densities in the Serengeti grass plains or other areas in Tanzania with which to compare. Thus, repeating this study will be necessary after a period of 5–10 years.
Kori bustard density was highest in green grass habitats with medium height (11–30 cm), probably because of food availability and topographic advantages for vigilance against predators. Habitat structure is considered to be important in determining habitat use by birds (Block & Brennan, 1993). Given that different habitat types vary with respect to composition and spatial distribution of food resources, it could be expected that the habitats used by kori bustards have sufficient species-specific resources. A high kori bustard density in green habitats might be influenced by an abundant food supply, such as insects. The presence of large herds of wild animals in green habitats may be attractive to kori bustards because they may offer protection from predators, ultimately increasing the time available to individual birds for finding and consuming food (Roberts, 1996; Fernández-Juricic, Siller & Kacelnik, 2004). Thus, the combination of food and offer of protection might influence kori bustard to congregate in these green habitats. The presence of fresh animal droppings may also attract beetles and other invertebrates, which are a source of food for the bustard.
Kori bustard density was high during the short dry season (January–March) and the long rain season (April–June). These high densities may be related to breeding season preparations as the male kori bustard courtship display was observed from late October throughout the breeding season during field data collection. The declining number of kori bustards at the study site after the breeding season indicated that the species migrate to more favourable habitats, for example with respect to food availability, during the long dry season (June–September) and migrate back to the plains before the rainy season.
Although not significantly different, the greater number of kori bustard observations in NW and SE, as well as in high grass habitats, that is medium height (11–30 cm), could be related to the kori bustard predator avoidance behaviour from sites with grass that is too long or too short. Given that a different habitat type varies strongly with respect to composition and spatial distribution of food resources, the habitats used by kori bustards are expected to have sufficient resources for the individual's daily demands.
Conclusions and recommendations
In conclusion, kori bustard density varied between the four study sites, but was statistically insignificant; however, seasonal variations in density were significantly different. The density was high during the long rain and short dry seasons. The study also confirmed that the distribution of kori bustard in the study sites was influenced more by grass height (11–30 cm) than by grass colour. Our predictions corresponded better with the season representing the main contributing factor for kori bustard density.
Due to a lack of data and research, kori bustard density recorded in our study in the Serengeti grass plains cannot be categorized as low or high. The species is listed as near threatened by the IUCN, and it is consequently important that the management authorities take appropriate actions to ensure the survival of the species. Protection of suitable habitats of the species within the Serengeti Ecosystem together with public education is imperative. The increasing illegal exploitation of natural resources by local people from communities alongside the Serengeti National Park makes community participation in the conservation and management of the species important. The large seasonal variations in density indicate that the population may utilize large parts of the ecosystem, thereby increasing the likelihood of encountering detrimental anthropogenic activities. Based on an earlier study of illegal offtake of grassland birds and eggs for home consumption (Magige et al., 2009), we advise protected area managers to address such threats posed by illegal offtake.
Acknowledgements
This project was supported by the Intergovernmental Policy Platform on Biodiversity and Ecosystem Services (IPBES) grant from the Ministry of Foreign Affairs through the Norwegian Environment Agency to the Norwegian University of Science and Technology (NTNU) and TAWIRI. This study was further supported by a grant from the kori bustard Species Survival Plan (SSP) and Jacksonville Zoo in the USA for financial support. In particular, we thank Sara Hallager, Smithsonian National Zoological Park and chair of the kori bustard SSP, and Katie Bagley, Zoo Atlanta and vice chair of the SSP, for providing funding for this study through the SSP. We also thank TAWIRI, NCAA and TANAPA for allowing us to conduct kori bustard research in the Ngorongoro Conservation Area and Serengeti National Park. Furthermore, we thank the Serengeti Wildlife Research Centre (SWRC) for logistical support and Juma Mkwizu (driver) and the entire community of SWRC for their hospitality during field work.
