DIURNAL MOVEMENT PATTERNS, HABITAT USE AND ENERGY COST OF LOCOMOTION BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RESERVE, KENYA By Timothy Mwanzia Ikime A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Zoology Master of Science 2015 ABSTRACT DIURNAL MOVEMENT PATTERNS, HABITAT USE AND ENERGY COST OF LOCOMOTION BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RESERVE, KENYA By Timothy Mwanzia Ikime Large carnivore population sizes are declining worldwide and this trend has been associated with habitat loss, diminishing prey availability, poaching, human - wildlife conflict and trade in their products. Large carnivores may feed on prey that are larger than themselves and are faced with challenges of meeting their daily food requirements. Here, I examined the diurnal (10am - 4pm) movement patter ns, habitat choice, and energy cost of locomotion by female spotted hyenas ( Crocuta crocuta ) in the Masai Mara National Reserve (MMNR) in Kenya. These patterns were investigated for a period of one year (2013). Two management regimes were employed in the M MNR throughout my study period, with intensive anthropogenic activity occurring on one side of the MMNR (disturbed) but not the other (undisturbed). I tested the hypothesis that the anthropogenic activity would be associated with differences in movement pa tterns, habitat choice and energy cost of locomotion between the two sides among groups of spotted hyenas fitted with GPS collars. I expected to see hyenas in the disturbed side of the MMNR travelling longer distances, using habitats char a cterized by dense r vegetati ve cover, and using more energy in locomotion. As expected, my results showed hyenas i n the disturbed side of MMNR travelled significantly longer distances t han they did in the undisturbed side , and used habitats covered with more dense vegetatio n than expected . However , we saw hyenas in the undisturbed side of the Reserve using significantly more energy in locomotion than they did in the disturbed side . iii This work is dedicated to my wife, daughter, parents, and siblings for being close to me despite the distance that kept us apart during my graduate studies at MSU. iv ACKNOWLEDGEMENTS This work would not have been a success if it were not for the help I received from several people. I thank the office of the Narok County G overnment and Brian Heath, Director of the Mara Conservancy for allowing us to conduct this research in the Masai Mara National Reserve. I am grateful to the Mara Hyena Project (MHP) personnel for their help during data collection for this work and making my stay at both Serena and Talek camps enjoyable. I thank the Michigan State University (MSU) and the Zoology Department for admitting me into their graduate programs. The department staff have been amazing to me. All members of the Holekamp lab have been more than helpful during my time at MSU. I thank my guidance committee members namely Dr. Kay Holekamp, Dr. Scott Winterstein and Dr. Catherine Lindell for their efforts and encouragement throughout my graduate life. I am deeply grateful to my advisor Dr . Kay Holekamp for agreeing to work with me both at MSU and in the field (the Masai Mara National Reserve). She gave me all the scholarly support I required and was more than an academic advisor to me. I am also deeply grateful to all the other people who provided support throughout this work. In particular , I am grateful to Zach Laubach and David Green , who spent enormous amount of time with me, and were always there for me throughout the drafting of this work. I thank Dr. Richard Hill for his scholarly a dvice in the energy calculations in this thesis. My roommate David Makacha was very instrumental. I thank all the MSU MasterCard scholars, the Foundation staff and the MasterCard Foundation Scholars Program for funding my graduate studies at MSU. v TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ............................. viii LIST OF FIGURES ................................ ................................ ................................ .............................. x KEY TO ABBRE VIATIONS ................................ ................................ ................................ ................ xii CHAPTER 1 ................................ ................................ ................................ ................................ ...... 1 GENERAL INTRODUCTION ................................ ................................ ................................ .............. 1 REFERENCES ................................ ................................ ................................ ................................ .. 15 CHAPTER 2 ................................ ................................ ................................ ................................ .... 20 DIURNAL MOVEMENT PATTERNS BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RESERVE, KENYA ................................ ................................ ................................ ........ 20 INTRODUCTION ................................ ................................ ................................ ......................... 20 Methods ................................ ................................ ................................ ................................ .... 23 Study animals ................................ ................................ ................................ ........................ 23 Collection of spatial data ................................ ................................ ................................ ...... 26 Minimum and maximum temperatures ................................ ................................ .............. 27 Prey counts ................................ ................................ ................................ ............................ 28 Seasons ................................ ................................ ................................ ................................ .. 28 Reserve management regime ................................ ................................ ............................... 29 Statistical analyses ................................ ................................ ................................ .................... 30 Reserve management regime and distance travelled ................................ ......................... 33 Time of the day and distance travelled ................................ ................................ ................ 34 Social rank and distance traveled ................................ ................................ ........................ 35 Prey and distance travelled ................................ ................................ ................................ .. 36 Seasonality and distance travelled ................................ ................................ ....................... 37 Frequency of movement ................................ ................................ ................................ ....... 41 Discussion ................................ ................................ ................................ ................................ . 42 Management regime and movement patterns ................................ ................................ ... 42 Social rank and distance moved ................................ ................................ ........................... 43 Time of day and movement patterns ................................ ................................ ................... 44 Prey availability and movement patterns ................................ ................................ ............ 45 Seasonal variation in movement ................................ ................................ .......................... 46 Conclusion ................................ ................................ ................................ ................................ . 47 REFERENCES ................................ ................................ ................................ ................................ .. 49 CHAPTER 3 ................................ ................................ ................................ ................................ .... 55 HABITAT USE PATTERNS BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RE SERVE, KENYA ................................ ................................ ................................ ........ 55 vi INTRODUCTION ................................ ................................ ................................ ......................... 55 Methods ................................ ................................ ................................ ................................ .... 57 Study site and study animals ................................ ................................ ................................ 57 Generating a classified vegetation map of the Masai Mara National Reserve (MMNR), Kenya ................................ ................................ ................................ ................................ ..... 57 Statistical analyses ................................ ................................ ................................ .................... 59 Results ................................ ................................ ................................ ................................ ....... 60 Habitat types availability in TMC and NCG ................................ ................................ .......... 60 Habitat use relative to availability in TMC and NCG ................................ ........................... 60 Grassland habitat use ................................ ................................ ................................ ........... 60 Shrub land habitat use ................................ ................................ ................................ .......... 61 Bare ground habitat use ................................ ................................ ................................ ....... 61 Ripa rian habitat use ................................ ................................ ................................ .............. 65 Discussion ................................ ................................ ................................ ................................ . 71 Grassland habitat use ................................ ................................ ................................ ........... 71 Shrub land habitat use ................................ ................................ ................................ .......... 72 Bare g round habitat use ................................ ................................ ................................ ....... 72 Riparian habitat use ................................ ................................ ................................ .............. 73 Habitat use outside defended territories ................................ ................................ ............ 73 Conclusion ................................ ................................ ................................ ................................ . 73 REFERENCES ................................ ................................ ................................ ................................ .. 75 CHAPTER 4 ................................ ................................ ................................ ................................ .... 79 ENERGY COST OF LOCOMOTION BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RESERVE, KENYA ................................ ................................ ................................ ........ 79 INTRODUCTION ................................ ................................ ................................ ......................... 79 Methods ................................ ................................ ................................ ................................ .... 80 Study area, and study animals ................................ ................................ ............................. 80 Estimating the energy cost of locomotion ................................ ................................ ........... 80 Statistical analyses ................................ ................................ ................................ .................... 85 Results ................................ ................................ ................................ ................................ ....... 87 Reserve management regime and energy use ................................ ................................ .... 87 Time of day and energy use ................................ ................................ ................................ .. 87 Social rank and energy use ................................ ................................ ................................ ... 87 Prey availability and energy use ................................ ................................ ........................... 88 Seasonality and energy use ................................ ................................ ................................ .. 88 Discussion ................................ ................................ ................................ ................................ . 89 Management regime and energy use ................................ ................................ .................. 89 Social rank and energy use ................................ ................................ ................................ ... 89 Prey and energy use ................................ ................................ ................................ .............. 90 Season and energy use ................................ ................................ ................................ ......... 90 Standardized temperature and energy use ................................ ................................ ......... 91 Conclusion ................................ ................................ ................................ ................................ . 91 GENERAL CONCLUSION S FROM THIS THESIS ................................ ................................ .............. 98 vii REFERENCES ................................ ................................ ................................ ................................ 100 viii LIST OF TABLES Table 2.1: Hyenas studied in TMC ................................ ................................ ................................ 25 Table 2.2: Hyenas studied in NCG ................................ ................................ ................................ . 26 Table 2.3: GLM table of results for distance (Ldist) moved . Distances have been transformed to log base 10 . Bolded cells indicate significant effects ................................ ................................ .... 33 Table 2.4: Table of average diurnal monthly distances travel l ed by hyenas in 2013 in the MMNR 2 Table 3.1: Habitat classification in TMC hyena territories ................................ ............................ 66 Table 3.2: Habitat classification in NCG hyena territor y 66 Table 3.3: Chi - square test results for differences in available habitats in TMC and NCG . Bolded cells represent significant effects . ................................ ................................ ................................ 67 Table 3.4: Chi - square test results for habitat use in TMC. Expected values are presented in parenthesis . ................................ ................................ ................................ ................................ ... 67 Table 3.5: Chi - square test results for habitat use in NCG. Expected values are represented in parenthesis . Bolded cells represent significant effects . ................................ .............................. 6 8 Table 3. 6 : Chi - square test results fo r differences in habitat use compared to available habitats in TMC and NCG. Expected values are presented in parenthesis . Bolded cells represent significant effects 6 8 Table 3. 7 : Hyena locations inside and outside the ir defended territories in TMC. ..................... 69 Table 3.8 : Hyena locations inside and outside the ir defended territor y in NCG. ........................ 70 Table 3.9 : Chi - square test results for habi tat use within and outside defended territories in TMC and NCG. Expected values are presented in parenthesis . Bolded cells represent significant effects . ................................ ................................ ................................ ................................ .......... 70 Table 4.1: Table of mean monthly temperatures in TMC nad NCG. Temperatures were recorded daily by sensors fitted on GPS collars and thereafter averaged into monthly records .............. 8 3 Table 4.2: Table of average diurnal monthly energy cost of locomo tion b y hyenas in TMC and NCG ................................ ................................ ................................ ................................ ............... 8 4 ix Table 4.3: . ................................ ....................... 8 5 Table 4.4: GLM Table of results for energy cost of locomotion . Energy values have been transformed to log base 10. Bolded cells represent significant ... 9 2 x LIST OF FIGURES Figure 1.1: Map of Africa and Kenya showing location of the MMNR ................................ ........... 8 Figure 1.2: Map of the MMNR showing the territories of the three study clans ......................... 1 0 Figure 1.3: GPS collar like those applied to adult 1 1 Figure 2.1: Hyena spatial movement patterns in the MMNR ................................ ....................... 24 Figure 2.2: Mean monthly minimum and max imum temperatures in TMC and NCG ................. 31 Figure 2.3: Prey density (animals/km 2 ) in TMC and NCG ................................ ............................. 32 Figure 2.4: Total monthly rainfall recorded in the MMNR ................................ ........................... 32 Figure 2.5: Management regime and average daily distance moved in TMC and NCG. Asterisk represents significant difference and error bars represent standard errors .............................. 34 Figure 2.6: Time of day and average daily distance moved in the MMNR . Asterisk represents significant difference and error bars represent standard errors ................................ ................. 3 5 Figure 2.7: Social r ank and average daily distance moved in the MMNR . Error bars represent standard errors ................................ ................................ ................................ ............................ 3 8 Figure 2.8: Prey availability and average daily distance moved in the MMNR . Error bars represent standard errors ................................ ................................ ................................ ............ 38 Figure 2.9: Seasonal ity and average daily distance moved in the MMNR . Error bars represent standard errors ................................ ................................ ................................ ............................. 3 9 Figure 2.10: Rank differences and everage daily distance moved in TMC and NCG. Asterisk represents significant difference and error bars represent standard errors .. 39 Figure 2.11: Prey availability and average daily distance moved in TMC and NCG. Error bars represent standard errors . 40 Figure 2.12: Seanal variation and average daily distance moved in TMC and NCG. Error bars represent standard errors ..41 Figure 3.1: Classified habitat map of the MMNR 6 2 Figure 3.2: Classified map of the MMNR showing the combined habitats 6 3 xi Figure 3.3 : Classified habitat map of the MMNR showing hyena locations ................................ . 6 4 Figure 3. 4 : Classified habitat map of the MMNR showing habitat use outside defended territories ................................ ................................ ................................ ................................ ...... 6 5 Fig. 4.1: Management regime and average daily energy cost of locomotion in TMC and NCG . Asterisk represents significant difference, and e rror bars represent standard errors ................ 9 3 Figure 4.2: Time of day and average daily energy cost of locomotion in the MMNR . Asterisk represents significant difference, and error bars represent standard errors .............................. 9 4 Figure 4.3: Social rank and average daily energy cost of locomotion in the MMNR . E rror bars represent standard errors ................................ ................................ ................................ ............ 9 4 Figure 4.4: Prey avilability and average daily energy cost of locomotion in the MMNR . Asterisk represents significant difference, and error bars represent standard errors 9 5 Figure 4.5: Seasonality and average daily energy cost of locomotion in the MMNR . Error bars represent s tandard errors 9 5 Figure 4.6: Social r ank and average daily energy cost of locomotion in TMC and NCG . Asterisk 6 Figure 4. 7 : Prey availability and average daily energy cost of locomotion in TMC and NCG. Error bars represent standard errors ................................ ................................ ................................ .... 9 6 Figure 4.8 : Seasonal ity and average energy cost of locomotion in TMC and NCG. Error bars represent standard errors ................................ ................................ ................................ ............. 9 7 Figure 4.9: The rel a tionship between hyena body mass (kg) and energy cost of locomotion .9 7 xii KEY TO ABBREVIATIONS MSU - Michigan State University NMNR - Masai Mara National Reserve MHP - Mara Hyena Project NCG Narok County Government TMC The Mara Conservancy MJ Mega Joules J Joules O 2 Oxygen Mm Millimeters Ml Milliliters Minimum m Temperature 0 C Degree Celsius SS Serena South SN Serena North TW Talek West No. Number Id. Identity NGO Non - Governmental Organization GPS Global Positioning System GCS Geographic Coordinate System UTM Universal Transve rse Mercator xiii WGS World Geodetic System GLM Generalized Linear Model UD Utilization Distribution Kg Kilogram M Meters Log Logarithm Fig. Figure Ldist Log base 10 transformed distance CO 2 Carbon dioxide H High L Low Yr Year IUCN International Union for Conservation of Nature Eqn Equation % - Percentage USGS United States Geological Society v o Speed (m/s) Mb Body mass V O Volume of Oxygen used Hrs Hours Km Kilometer ESRI Environmental Systems Research I nstitute VHF Very High Frequency Std Standardized 1 CHAPTER 1 GENERAL INTRODUCTION Biodiversity is facing widespread competition with humanity for space and resources (Balmford et al., 2001). M any species, including large carnivores, are increasingly coming into conflict with people (Woodroffe and Ginsberg, 1998). Members of the mammalia n order Carnivora , most of which are predators , number approximately 226 species , ( Treves and Karanth, 2003) and feed on animals at lower trophic positions. Carnivores often regulate or limit the numbers of their prey, thereby influencing the structure and function of entire ecosystems (Estes et al. 1998; Berger et al. 2001; Terborgh et al., 2001). As a result, carnivore management remains a key concern to conservation biologists all over the world (Treves and Karanth, 2003). The 31 largest carnivores (wit belong to five families: Canidae, Felidae, Mustelidae, Ursidae, and Hyaenidae (Ripple at al., 2014) vertebrates feared to face a mass extinction of 30 - 96% ( Rosenzweig et al., 2012), large carnivores are no exception. Large carnivores generally have small population sizes. The majority of these species (61%) are listed by the International Union for the Conservation of Nature (IUCN) as threatened (vulnerable, endangered , or critically endangered), and are at risk of local or total extinction (Ripple et al., 2014). In this thesis, I will be focusing on spotted hyenas ( Crocuta crocuta ), which belong to the Carnivore family Hyaenidae. Populations of carnivores are declini ng around the globe, often with dramatic effects on lower trophic levels (Estes et al., 2011). These declines are most severe in large species, which 2 require large areas with intact prey communities, and which are prone to killing livestock (Woodroffe, 200 0). Large carnivores typically range over such wide areas that it can be difficult to maintain viable populations without some individuals coming into close proximity to humans, posing serious threats to human safety and domestic livestock (Packer et al., 2013). As a result, large carnivores are usually among the first species to disappear from landscapes, often with strong cascading effects on ecosystem structure and function (Estes et al., 2011). In the past century, carnivore populations have experienced drastic, global reductions due to increasing human population densities, habitat loss and fragmentation, reduced prey availability, and elevated rates of conflict (Gittleman et al., 2001). Conservationists have therefore sought methods to promote human ca rnivore coexistence outside the confines of national parks and wilderness areas (Dickman et al., 2011). Human - - rich diet and large home ranges that draw them into recurrent competiti on with humans, who have very similar needs to those of the carnivores themselves. Indeed, many large carnivore species are specialized for ungulate predation; therefore some individuals readily kill domesticated ungulates when opportunities to do so arise (Karanth et al., 1999). This is a worldwide problem, exemplified by wolves ( Canis lupus ) and coyotes ( Canis latrans ) killing sheep in North America , pumas ( Puma concolor) and jaguars ( Panthera onca ) taking cattle in South America , numerous carnivore gener a preying on cattle and goats in Africa, and tigers ( Panthera tigris ) and leopards ( Panthera pardus ) killing livestock in Asia (Jackson & Nowell, 1996; Kaczensky, 1999). Between 1992 and 2001, black bears ( Ursus americanus) killed 429 livestock in the stat e of Wisconsin (U.S.A.), whereas wolves ( Canis lupus ) killed 164 livestock (Treves et al., 2002). Under some 3 conditions, individual carnivores attack humans, with tragic consequences for all (Treves & Naughton - Treves, 1999; Rajpurohit & Krausman, 2000). Be cause active persecution and accidental killing by local people remain the most important causes of mortality for many predators (Woodroffe & Ginsberg, 1998), carnivore declines are likely to roughly track the expansion of human populations in the future ( Woodroffe, 2000). Regional and international trade in carnivore skins, bones and other body parts may also encourage local people to kill predators (Woodroffe, 2000). In Africa, which is home to the last intact guild of large carnivores (Cozzi et al., 2012 ), large carnivore numbers have declined considerably over the last 30 years (Western et al., 2009 ) . large carnivores are important generators of income through tourism and hunting (e.g., Western & Henry, 1979). Conservation of large carnivores may therefore make economic sense, particularly in dry rangelands of limited value for agriculture (Ogada et al., 2003). After lions ( Panthera leo ), spotted hyenas are the largest carnivores in Africa (Tilson and Henschel, 1986), and they are also by far the most abundant large carnivores on the continent. Spotted hyenas are opportunistic hunters, targeting whichever prey specie s are locally most abundant (Kruuk, 1972; Coop er, 1990; Holekamp et al., 1997b). These hyenas can survive in environments from which other large predators such as cheetahs ( Acinonyx jubatus ), lions, and African wild dogs ( Lycaon pictus ) have been extirpated; if hyenas also vanish, a particular habitat has most likely become very severely degraded (Mills and Hofer, 1998). It has become a 4 - first century will depend on the goodw ill of local communities (Lamprey and Reid, 2004) in order to avoid such extirpations. Schuette et al., (2013) found that distance to active human settlements has the strongest influence on habitat occupancy by carnivore species in the southern Rift valley of Kenya indicating the likelihood of encounters between carnivores and people. This suggests that, at least in the southern rift valley of Kenya, carnivores adjust patterns of occupancy in reaction to human space use in quite a fine - scaled manner, rather than simply avoiding areas associated with high levels of human activity over the long term. Occupancies by lions, spotted hyenas, and baboons ( Papio spp.), all potential high - conflict species, were high near active human settlements, whereas occupancies by striped hyenas ( Hyaena hyaena ) and black - backed jackals ( Canis mesomeles ), which are low - conflict species, were low near active human settlements. In general, Schuette et al., (2013) found that seasonal changes in land use by humans and livestock trigge red seasonal changes in carnivore occupancy patterns. In th e study by Schuette et al. (2013) , the probability of detecting lions was highest between 10:00 pm and 3:59 am, peaked between 12:00 am and 1:59 am, and was followed by a secondary peak at dawn fr om 6:00 to 7:59 am. This pattern is consistent with frequent nocturnal hunting, followed by movement into hiding cover around dawn. By contrast, spotted hyena detection probability was highest between 12:00 am and 5:59 am, with a peak between 12:00 am and 1:59 am and a secondary peak from 8:00 pm to 9:59 pm. This partial partitioning of time apparently allows for coexistence of large carnivores living in same ecosystem. L arge carnivores including lions, spotted hyenas, and medium carnivores, including strip ed hyenas, and black - backed jackals, occurred at high rates in the community - run conservation area in 5 which no people resided, but declined with increasing human utilization of the landscape (Schuette et al. 2013) . Dense vegetative cover appears to provide refuge to lions and spotted hyenas exposed to people and livestock (Boydston et al., 2003b; Kolowski and Holekamp, 2009). Selective harassment or killing of lions may happen if local pastoralists perceive lions as the most destructive, daring and aggressive large predators (Omondi, 1994) and thus respond to lion attacks by directing greater retaliative aggression toward them. This may also happen if lions remain closer to villages after attacks, are more reluctant to run away from people, or escape over s horter distances than other predators when detected by humans. Due to habitat loss, poaching and other disturbances, resident wildlife populations have declined by more than 70% over the last 20 years in the Masai Mara National Reserve, Kenya (Ottic hilo, 2000; Serneels & Lambin, 2001). Henceforth, I shall refer to this spectacular wildlife among the highest recorded in African savannas (Ogutu & Dublin, 2002) but is unusually low at the edge of the Reserve in areas adjoining pastoral ranches (Ogutu & Dublin, 2002, 2004), implicating possible negative impacts of pastoralism on lion density and distribution. Intensive and systematic searches for lions (Ogutu & Dublin, 2004; Reid et al., 2003) support the notion that lion numbers are lower in the pastoral areas than elsewhere, portending a severe threat to the long - term viability of the lion populations inhabiting such areas. Like lions and African wild dogs, spotted hy enas are strongly dependent on protected areas or zones of low human density that contain sufficient numbers of suitable prey to support them (Mills and Hofer, 1998). Thus, the future of these species lies inside rather than outside 6 large conservation area s. However, leopards and spotted hyenas appear better able to adapt to habitats modified by people (both may be sighted in some towns in East Africa) than do lions, wild dogs or cheetahs. This adaptability might reflect their avoidance of people by shiftin g from crepuscular to strictly nocturnal activity, and/or their ability to survive by scavenging when natural prey are depleted (Woodroffe, 2000). Spotted hyenas from multiple social groups have been killed in villages near the Reserve during livestock dep redation attempts (Kolowski and Holekamp, 2006). The Maasai people, who live near the Reserve, are pastoralists who often respond to livestock depredation by indiscriminately poisoning, snaring, or s pear ing the predators that were putatively responsible fo r livestock depredation (Rudnai, 1979; Omondi, 1994). Altered use of space, social behavior, circadian activity rhythms and demographic structure in spotted hyenas residing at the edge of the Reserve (Boydston et al., 2003b) have also been linked to increa sing interference by livestock grazing within hyena clan territories. Indeed, herders sometimes harass or kill hyenas when they encountered them (K. E. Holekamp, unpubl. Data). Livestock grazing has been shown to be one of the key factors influencing behav ioral plasticity of spotted hyenas in the MMNR (Kolowski et al., 2007; Boydston et al., 2003b; Kolowski and Holekamp, 2009), and hyenas have been reported to run away from pastoralists looking after their livestock on foot (Kolowski and Holekamp, 2009). P atterns of activity are clearly variable in spotted hyenas, yet sources of this variation remain poorly understood. Kolowski et al. (2007) documented the influences of sex and social rank on activity patterns of spotted hyenas. They found that male spotted hyenas tended to be more active than females, particularly during the morning (0700 1100 h), and also tended to 7 exhibit higher movement rates. Neither rates of activity nor movement varied with social rank, but low - ranking females spent more time feeding than did high - ranking females. Finally, female hyenas in territories with daily livestock grazing and high tourist visitation rates showed lower activity and less den use than did hyenas in an undisturbed territory during the times of day when human activi ty coincided with potential hyena activity. The specific times of day when hyena activity was reduced indicated that livestock grazing rather than tourist activity was most likely responsible for observed shifts in activity. However, to date no one has ev er compared hyena activity or movement patterns between neighboring portions of any single protected area in which contrasting management policies were implemented, which is the primary purpose of this thesis research. Boydston et al., (2003a) suggested th at a better understanding of individual variation in space use patterns and the mechanisms by which edge effects can lead to extinction should aid in planning for the protection of wide ranging carnivores around the world. Therefore my own study investigat ed such effects by looking at the movement patterns, habitat usage, and energy costs of locomotion by spotted hyenas in the MMNR (Fig. 1.1), as these might be influenced by anthropogenic activities during the daytime (10am - 4pm ) . The specific purpose of my thesis was to test the hypothesis that anthropogenic activities affect diurnal movement, habitat between 10 am and 4 pm each day; this encompassed the time outside of primary active time (6pm - 9am) (see Kolowski et al., 2007). 8 Figure 1.1: Map of Africa and Kenya showing location of the MMNR. 9 I investigated how social rank (high or low), time of day (am or pm) season (we t or dry), prey density ( a bundant or scarce ), ambient temperature and local management regime within the Reserve (no livestock grazing allowed or livestock grazing allowed) influenced movement, habitat usage and energy expenditure by these hyenas. My work f ocused on 22 adult (> 3 years of age) female spotted hyenas in the Reserve by investigating their behavior patterns in three deployed in radio collars (Fig. 1.3) on the hyenas to track their locations from January 1 st to 31 st December 2013. The Talek West (TW) clan was located near Talek town in an area (see Fig 1.2) where illegal intensive livestock grazing takes place on a daily basis from 9am - 6pm (Kolowski et al., border have been estimated as being over 12,000 for cattle and 16,500 for sheep and goats (Kolowski and Holekamp, 2006). By contrast, both the Serena South (SS) and Seren a North (SN) clans were located in an area where livestock grazing was prohibited (Fig. 1.2). The MMNR (1,500 km 2 ), which lies in southwestern Kenya (1 0 0 ig.1). The Reserve consists primarily of rolling grassland and scattered bushland (predominantly Croton and Euclea species), with riparian forest along the major watercourses. This habitat supports a large diversity of resident herbivores including both grazing and browsing species (Kolowski et al., 2007; Kolowski and Holekamp, 2009). It is bounded by the Serengeti Na tional Park to the south, the Siria escarpment to the west and Maasai pastoral ranches to the north and east (Norton - Griffiths et al., 1975). The rangelands surrounding the MMNR contain year - round 10 communities of resident wildlife, but migratory wildlife al so spill out onto them during the dry season. Figure 1.2: Map of the MMNR showing the territories of the three study clans. 11 Figure 1.3: A GPS collar like those applied to adult female hyenas in this study. These dry - season grazing resources in the buffer zone surrounding the Reserve are important to the migrant wildebeest ( Connochaetes taurinus ) and to livestock alike (Singida, 1984). The land uses in the areas surrounding the MMNR include traditional pastoralism , wildlife conservation, tourism, subsistence maize cultivation and commercial wheat cultivation (Serneels et al., 2001). Wildlife conservation and tourism are the only land uses legally permitted within the Reserve. Wildebeest, common zebra ( Equus quagga , formerly Equus burchellii Eudorcas thomsonii ) migrate between MMNR and the Loita p lains within the Masai group ranches to the north - east of MMNR (Norton - Griffiths et al., 1975). The Reserve is also the northernmost destination o f the great migration of zebra and 12 wildebeest from the southern portion of the Serengeti ecosystem. These migratory herbivores are usually present in the reserve from June or July through October each year (Ogutu et al., n popularity poll contest in September 2007 by the international media (Ndegwa and Murayama, 2009). The Sand, Talek and Mara rivers are the major watercourses draining the Reserve. Shrubs and trees fri nge most permanent and seasonal watercourses, and bushes cover many slopes and hilltops (Ogutu et al., 2009). Rainfall in the ecosystem generally increases along a southeast to northwest gradient (Norton - Griffiths et al., 1975), varies strikingly in space and time within the Reserve, and is markedly bimodal during the year. Most rainfall occurs during 2 May. Sunset and sunrise times take place around 18:45 and 06:30 hrs r espectively, with little seasonal variation (Kolowski et al., 2007; Kolowski and Holekamp, 2009). The long dry season spans July October (Norton - Griffiths et al., 1975). The MMNR is jointly managed by the Narok and Transmara County Governments on behalf of the government of Kenya. The Transmara C ounty Government hires a management agency called T he Mara Conservancy, hereafter , r iver (this , hereaft er referred , clan inhabits the area managed by the NCG, whereas the SN and SS clans defend territories in the Mara Triangle. 13 As has been shown in earlier researc h, the TW clan defends a territory along the northeastern Reserve border, in close proximity to a rapidly growing pastoralist population (Kolowski & Holekamp, 2006), and is subjected to intensive daily livestock grazing pressure (Kolowski and Holekamp, 200 9). By contrast, livestock grazing is not allowed in the Mara Triangle. Previous research in the side of the Reserve managed by the NCG showed that 71% of all large carnivore attacks on livestock grazing illegally within the Reserve took place between 11am and 4pm (Kolowski and Holekamp, 2006). All three of our study clans have been continuously monitored by Mara Hyena Project (MHP) personnel for many years, and the database from the entire history of the project was made available to me. GPS collars (Fig . 1.3) have been deployed on several female hyenas in each of these three clans since 2012, and they have been calibrated in such a way that the hyena locations are received at a central position within the MHP camp throughout each 24 hour cycle. Other dat a collected by MHP personnel include biweekly prey counts, daily minimum and maximum temperatures, reproductive states of the hyenas, and livestock & tourist counts done twice every month. All these data were available to me through the database maintained at Michigan State University (MSU), and it is on these data that my thesis is based. I also utilized the seven elements of photo interpretation (size, shape, shadow, pattern, tone, texture and association), G oogle Earth and Landsat images to create rando m forests and thereafter classif ied the different vegetation cover types in the MMNR. As described by Bourgeau - Chavez et al., 2015, random forests were created within the Google Earth and Landsat TM 5 2009 by drawing polygons which corresponded to seven (7 ) different habitat 14 types in the Masai Mara (Fig. 3.1) . The polygons were converted into shape files using Environmental Systems Research Institute (ESRI) ArcMap software (v. 10.2.2). I used R statistical software (R core team, 2015) to convert the shape f iles into a classified habitat type image (raster for the MMNR) . These seven habitat types were buildings, grassland, shrub land, bare ground, forest, river and wetland (Fig 3.1) . Chapter 2 of my thesis describes the diurnal movement patterns of collared female spotted hyenas in terms of linear distances travelled by these females, who rarely disperse from their natal territories ( Chapter 3 of my thesis describes the usage patterns of female spotted hyenas in relation to the different habitat types in which they spend their time. My final chapter (Chapter 4) investigates the energetic costs to the hyenas of their movements and habit at usage, as described in Chapters 2 and 3. Each of these chapters will have Introduction, Methods, Results, and Discussion sections, as well as a brief Conclusion. Then finally, I will end my thesis with General conclusions at the end of Chapter 4. Bec ause this work was done collaboratively with my MHP colleagues, I will use 1 st person plural in Chapters 2, 3, and 4. 15 REFERENCES 16 REFERENCES Balmford, A., Moore, J. L., Brooks, T., Burgess, N., Hansen, L. A., Williams, P., & Rahbek, C. (2001). Conservation conflicts across Africa. Science , 291 (5513), 2616 - 2619. Berger, J., Stacey, P. B., Bellis, L., & Johnson, M. P. (2001). A mammalian predator - prey imbalance: grizzly bear and wolf extinction affect avian neotropical migrants. Ecological Applications , 11 (4), 947 - 960. Bourgeau - Chavez, L., Endres, S., Battaglia, M., Miller, M. E., Banda, E., Laubach, Z., ... & Marcaccio, J. (2015). Development of a Bi - National Great Lakes Coastal Wetland and Land Use Map Using Three - Sea son PALSAR and Landsat Imagery. Remote Sensing, 7(7), 8655 - 8682. Boydston, E. E., Kapheim, K. M., Szykman, M., & Holekamp, K. E. (2003a). Individual variation in space use by female spotted hyenas. Journal of Mammalogy, 84(3), 1006 - 1018. Boydston, E. E., K apheim, K. M., Watts, H. E., Szykman, M., & Holekamp, K. E. (2003 b ). Altered behaviour in spotted hyenas associated with increased human activity. Animal Conservation , 6 (3), 207 - 219. Cozzi, G., Broekhuis, F., McNutt, J. W., Turnbull, L. A., Macdonald, D. W ., & Schmid, B. (2012). Fear of the dark or dinner by moonlight? Reduced temporal partitioning among Africa's large carnivores. Ecology , 93 (12), 2590 - 2599. Cooper, S. M. (1990). The hunting behaviour of spotted hyaenas (Crocuta crocuta) in a region contain ing both sedentary and migratory populations of herbivores. African Journal of Ecology , 28 (2), 131 - 141. Dickman, A. J., Macdonald, E. A., & Macdonald, D. W. (2011). A review of financial instruments to pay for predator conservation and encourage human carn ivore coexistence. Proceedings of the National Academy of Sciences , 108 (34), 13937 - 13944. Estes, J. A., Terborgh, J., Brashares, J. S., Power, M. E., Berger, J., Bond, W. J., ... & Wardle, D. A. (2011). Trophic downgrading of planet Earth. Science , 333 (604 0), 301 - 306. Estes, J. A., Tinker, M. T., Williams, T. M., & Doak, D. F. (1998). Killer whale predation on sea otters linking oceanic and nearshore ecosystems. Science , 282 (5388), 473 - 476. Gittleman, J. L. (Ed.). (2001). Carnivore conservation (Vol. 5). Ca mbridge University Press. Holekamp, K. E., Smale, L., Berg, R., & Cooper, S. M. (1997b). Hunting rates and hunting success in the spotted hyena (Crocuta crocuta). Journal of Zoology , 242 (1), 1 - 15. Jackson, P., & Nowell, K. (1996). Problems and possible solutions in management of felid predators. Journal of Wildlife Research , 1 , 304 - 314. 17 Kaczensky, P. (1999). Large carnivore depredation on livestock in Europe. Ursus , 59 - 71. Karanth, K. U., Sunquist, M. E., & Chinnappa, K. M. (1999). Long - term monitoring of tigers: lessons from Nagarahole. Riding the tiger: tiger conservation in human - dominated landscapes. Cambridge University Press, Cambridge, United Kingdom , 114 - 122. Kolowski, J. M., & Holekamp, K. E. (2006). Spatial, temporal, and physical characteristi cs of livestock depredations by large carnivores along a Kenyan reserve border. Biological conservation , 128 (4), 529 - 541. Kolowski, J. M., & Holekamp, K. E. (2009). Ecological and anthropogenic influences on space use by spotted hyaenas. Journal of zoology , 277 (1), 23 - 36. Kolowski, J. M., Katan, D., Theis, K. R., & Holekamp, K. E. (2007). Daily patterns of activity in the spotted hyena. Journal of Mammalogy , 88 (4), 1017 - 1028. Kruuk, H. (1972). The spotted hyena: a study of predation and social behavior. Chicago: University of Chicago Press. Lamprey, R. H., & Reid, R. S. (2004). Expansion of human settlement in Kenya's Maasai Mara: what future for pastoralism and wildlife?. Journal of Biogeography , 31 (6), 997 - 1032. Mills, G., & Hofer, H. (1998). Status sur vey and conservation action plan. Hyaenas. IUCN/SSC Hyaena Specialist Group, IUCN, Switzerland . Ndegwa Mundia, C., & Murayama, Y. (2009). Analysis of land use/cover changes and animal population dynamics in a wildlife sanctuary in East Africa. Remote Sensing , 1 (4), 952 - 970. Norton Griffiths, M., Herlocker, D., & Pennycuick, L. (1975). The patterns of rainfall in the Serengeti ecosystem, Tanzania. African Journal of Ecology , 13 (3 4), 347 - 374. Ogada, M. O., Woodroffe, R., Oguge, N. O., & Frank, L. G. (2003). Limiting depredation by African carnivores: the role of livestock husbandry. Conservation Biology , 17 (6), 1521 - 1530. Ogutu, J. O., & Dublin, H. T. (2002). Demography of lions in relation to prey and habitat in the Maasai Mara National Reserve, Kenya. African Journal of Ecology , 40 (2), 120 - 129. Ogutu, J. O., & Dublin, H. T. (2004). Spatial dynamics of lions and their prey along an environmental gradient. African journal of Ecology , 42 (1), 8 - 22. Ogutu, J. O., Piepho, H. P., Dublin, H. T. , Bhola, N., & Reid, R. S. (2009). Dynamics of Mara Serengeti ungulates in relation to land use changes. Journal of Zoology , 278 (1), 1 - 14. Omondi, P. (1994). Wildlife - human conflict in Kenya: integrating wildlife conservation with human needs in the Masai Mara region. 18 Ottichilo, W. K. (2000, March). Wildlife dynamics: an analysis of change in the Masai Mara ecosystem of Kenya. ITC. Packer, C., Loveridge, A., Canney, S., Caro, T., Garnett, S. T., Pfeifer, M., ... & Plumptre, A. (2013). Conserving large carni vores: dollars and fence. Ecology letters , 16 (5), 635 - 641. R Core Team (2015). R: A language and environment for statistical computing. R Foundationfor Statistical Computing, Vienna, Austria. URL http://www.R - project.org/. Rajpurohit, K. S., & Krausman, P. R. (2000). Human - sloth - bear conflicts in Madhya Pradesh, India. Wildlife Society Bulletin , 393 - 399. Reid, R. S., Rainy, M., Ogutu, J., Kruska, R. L., Kimani, K., Nyabenge, M., ... & Lamprey, R. (2003). People, wildlife and livestock in the Mara ecosystem: The Mara Count 2002. International Livestock Research Institute, Nairobi, Kenya . Ripple, W. J., Estes, J. A., Beschta, R. L., Wilmers, C. C., Ritchie, E. G., Hebblewhite, M., ... & t carnivores. Science , 343 (6167), 1241484. Rosenzweig, M. L., Drumlevitch, F., Borgmann, K. L., Flesch, A. D., Grajeda, S. M., Johnson, G., ... & Serrano, K. P. (2012). An ecological telescope to view future terrestrial vertebrate diversity. Evolutionary Ecology Research , 14 (3), 247 - 268. Rudnai, J. (1979). Ecology of lions in Nairobi National park and the adjoining Kitengela Conservation Unit in Kenya. African journal of ecology , 17 (2), 85 - 95. Schuette, P., Creel, S., & Christianson, D. (2013). Coexistence of African lions, livestock, and people in a landscape with variable human land use and seasonal movements. Biological Conservation , 157 , 148 - 154. Serneels, S., & Lambin, E. F. (2001). Impact of land use changes on the wildebeest migration in the northern part of the Serengeti Mara ecosystem. Journal of Biogeography , 28 (3), 391 - 407. Serneels, S., Said, M. Y., & Lambin, E. F. (2001). Land cover changes around a major east African wildlife reserve: the Mara Ecosystem (Kenya). International Journal of Remote Sensing , 22 (17), 3397 - 3420. Singida, I. (1984) Land and population problems in Kajiado and Narok, Kenya. African Studies Revision, 27, 23. Smale, L., Nunes, S., & Holekamp, K. E. (1997). Sexually dimorphic dispersal in mammals: patterns, causes, and conseq uences. Advances in the study of behaviour , 26 , 181 - 251. Ecological meltdown in predator - free forest fragments. Science , 294 (5548), 1923 - 1926. 19 Tilson, R. L., & Henschel, J. R. (1986). Spatial arrangement of spotted hyaena groups in a desert environment, Namibia. African Journal of Ecology , 24 (3), 173 - 180. Treves, A., & Karanth, K. U. (2003). Human carnivore conflict and perspectives on carnivore management wor ldwide. Conservation Biology , 17 (6), 1491 - 1499. Treves, A., & Naughton - Treves, L. (1999). Risk and opportunity for humans coexisting with large carnivores. Journal of Human Evolution , 36 (3), 275 - 282. Treves, A., Jurewicz, R. R., Naughton - Treves, L., Rose, R. A., Willging, R. C., & Wydeven, A. P. (2002). Wolf depredation on domestic animals in Wisconsin, 1976 - 2000. Wildlife Society Bulletin , 231 - 241. Western, D., & Henry, W. (1979). Economics and conservation in third world national parks. BioScience , 29 (7), 414 - 418. Western, D., Russell, S., & Cuthill, I. (2009). The status of wildlife in protected areas compared to non - protected areas of Kenya. PLoS One, 4(7), e6140. Woodroffe, R. (2000). Predators and people: using human densities to interpret declines of large carnivores. Animal conservation , 3 (02), 165 - 173. Woodroffe, R., & Ginsberg, J. R. (1998). Edge effects and the extinction of populations inside protected areas . Science , 280 (5372), 2126 - 2128. 20 CHAPTER 2 DIURNAL MOVEMENT PATTERNS BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RESERVE, KENYA INTRODUCTION Large carnivores are known to require large ranging areas with intact prey communities (Woodroffe, 2000), and they move across landscapes that are spatially heterogeneous ( Kotliar and Wiens 1990) . In fact, habitat heterogeneity plays a key role in determining animal movement patterns more generally, as well as such behavioral and ecological processes as feeding behavior (e.g. Etzenhouser et al. 1998; Bailey and Thompson, 2006). In their pursuit of food reso urces, large carnivores may come into conflict with livestock and humans thus causing serious threats to both (Packer et al., 2013). H Gittleman et al., 2001). 61% of largest carnivore species are listed as threatened by the International Union for Conservation of Nature (IUCN) (Ripple et al., 2014). Large carnivores represent examples of elusive species that alt er their daily movement patterns in response to environmental stimuli, and thus can be used as indicators of the degree of environmental stress caused by anthropogenic influence on ecosystems (Seryodkin et al., 2013). European brown bears ( Ursus arctos ) an d wolves ( Canis lupus ) have been documented to show twilight or nocturnal activity periods to avoid humans (Theuerkauf et al., 2003; Ordiz et al., 2013). Mountain lions ( Puma concolor ) become more nocturnal when human activity increases (Van Dyke et al., 1 986), and coyotes ( Canis latrans ) resume diurnality after human persecution ceases (Kitchen et al., 2000). 21 In Africa, large carnivores are faced with the same challenges as elsewhere in the world today, and they have undergone large population declines ove r the last 30 years (Western et al., 2009). In Tanzania, human respondents significantly viewed large carnivores as more problematic than other species because of the threats they pose to humans and livestock (Dickman, 2014). In Kenya, resident wildlife de clines have been reported in such protected areas as Meru, Nairobi and the Tsavo National Parks (Western et al., 2009). Kenya is also an tourism industry is heavily dependent on national parks and reserv es, which comprise roughly 8% of the total land mass. These protected areas represent a key source of foreign exchange as well as a major source of employment for ai grazing and harassment of wildlife in the Mara is prohibited (Kenya Wildlife Act, 1989), but these regulations may rarely be enforced (Walpole et al., 2003). In the Mara, livestock grazing has been documented to be one of the most important factors affecting space use patterns by the spotted hyena ( Crocuta crocuta ), the most abundant of the large carnivore species in the Reserve (Boydston et al., 2003b). Despite all t he threats facing large carnivores, they are indicators of ecosystem health, and the spotted hyena has been documented to thrive in ecosystems where other species cannot survive. Thus, its disappearance from an ecosystem is an indication of particularly se vere ecosystem degradation (Mills and Hofer, 1998; Woodroffe, 2000). Here, we investigate the effects of anthropogenic activity on the diurnal movement patterns of the spotted hyena, which is found throughout most of sub - Saharan Africa (Kruuk, 22 1972). Spott ed hyenas are strongly dependent on protected areas or zones with minimal human population densities and with adequate prey numbers (Mills and Hofer, 1998). Our study was carried out in the Mara located in south - western Kenya along the border with Tanzania to the South (Fig. 1.1). Details of our study area are as described in Chapter 1. As mentioned in chapter 1, the MMNR is a protected area but it is managed by two different management agencies. The reserve is divided into two areas by the Mara River, an d the area east of this river is managed by the Narok County Government (NCG); this management regime has been in place since the 1960s. The western side of the Reserve falls under the jurisdiction of the Trans - Mara County Government, and is managed by a n on - governmental organization (NGO) called The Mara Conservancy (TMC); TMC has managed the western portion of the Reserve since 2001. Intensive livestock (cattle, sheep and goats) grazing takes place in the eastern side of the reserve, managed by the NCG, b ut no livestock grazing is allowed in the western side of the reserve, managed by TMC. Thus the eastern side of the Reserve is relatively heavily disturbed by anthropogenic activity but the western side is very pristine. Here we studied clans of spotted hy enas located on both sides of the Reserve. The members of the Talek West (TW) clan defend a group territory in the disturbed (livestock grazing allowed) side of the Reserve while the Serena South (SS) and Serena North (SN) clans defend territories in the u ndisturbed (livestock grazing prohibited) side of the Reserve (Fig. 1.2). This gave us the opportunity to make comparisons regarding distance travelled by hyenas living in these two environments. Specifically, we tested the hypothesis that anthropogenic ac tivity affects hyena movements during daylight hours. We expected the daily movements of cattle and herders in the disturbed side of the Reserve to cause hyenas to change their locations more 23 frequently and thus to travel longer distances than in the undis turbed area. Throughout this to refer to the disturbed side of the reserve (Fig. 1.2). Methods Study animals Spotted hyenas live in fission - fusion social groups called clans, and each clan is composed of multiple adult females, their offspring, and one to several adult immigrant males (Kruuk, 1972; Holekamp et al., 1996). These hyenas may forage either singly or in small groups (Kruuk, 1972; Mills, 1984; Tilson & Henschel, 1986), and members of each clan defend a group territory (Hofer and East, 1993a). Here, all hyenas were known individually by their unique spots (Cooper, 1989). Sex was determined by the dimorp hic glans morphology of the erect phallus (Frank et al., 1990). Females were considered adults when they reached three (3) years (yrs) of age or conceived their first litter, whichever came first (Boydston et al., 2003a). Social ranks of our study animals were known from their positions in a matrix of outcomes of dyadic agonistic interactions (Smale et al., 1993). High - ranking females were considered to be those in the top - ranking females were considered to be those in the bottom third (Green 2015). Between 1 st January and 31 st December 2013, 22 adult female spotted hyenas (>3yrs) were fitted with Global Positioning System (GPS) radio collars (see details below), and subsequently tracked to monitor their diurna l movement patterns in the MMNR (Fig. 2.1). Based on the outcomes of dyadic agonistic interactions between individual - ranking or low - ranking) were individually 24 known by the time we deployed GPS collars on them. In total, we collared 13 high - ranking (H) hyenas (TMC=9, NCG=4) and 9 low - ranking (L) hyenas (TMC=5, NCG=4) (Tables 2.1 and 2.2). Figure 2.1: Hyena spatial movement patterns in the MMNR. 25 Name Collar Id Rank Bart 11526 H Clov 11528 H Digs 11755 H Ema 11530 L Hndy 11522 L Java 11757 H Mtn 11523 L Peep 11529 H Shrm 11111 H Slin 11520 H Taj 11527 L Tnsl 11525 L Wafl 11521 H Zoey 11524 H Table 2.1: Hyenas studied in TMC. 26 Name Collar Id Rank Baez 11414 L Hel 11419 H Hex 11417 L Juno 11413 H Mgta 11412 H Pan 11416 H Roos 12157 L Tilt 11415 L Table 2.2: Hyenas studied in NCG. Collection of spatial data Twenty two adult female spotted hyenas (TMC=14, NCG=8, see table 2.1 and 2.2) were immobilized using tiletamine - zolazepam (6.5 mg/kg; Telazol: W.A Butler Company, Brighton Michigan) administered in a plastic dart via a CO 2 powered rifle (Telinject Inc., Saugus, California), and fitted with GPS collars (Vectronix Aerospace, Berlin, Germany). Collars were programmed to transmit their locations at 10am, 1 pm, and 4 pm every day. The GPS collar technology used here is very precise when compared to earlier studies, which used Very High Frequency (VHF) collars that could only be tracked from vehicles or airplan es. All locations here were digitized using ESRI ArcMap 10.2.2 software (Environmental Systems Research Institute, Redlands, CA, USA). We used a Geographic Coordinate System (GCS) of World Geodetic System of 1984 (WGS 84) then projected the shape files of the GPS coordinates into Universal 27 Transverse Mercator zone 36 South (WGS 84 UTM 36S) (Fig. 2.1). Finally, XY coordinates were added to the GPS coordinates to allow for linear distance calculations in meters (m); these calculations were done using R statis tical software (R core team, 2015). Linear distances were then calculated between each two consecutive locations obtained during daylight hours. Territory sizes were estimated using fixed kernel utilization distributions (UDs) with 95% probabilities for al l hyenas with collars in 2013 (Green, 2015). Minimum and maximum temperatures Earlier studies have shown that spotted hyenas reduce their activity with increases in ambient temperatures (Kruuk, 1972; Cooper, 1990). Therefore, we monitored daily minimum a nd maximum temperatures in 2013 at weather stations located in TMC and NCG. Our study area was situated very close to the equator, such that hours of daylight and darkness were roughly the same year - round. Sunrise and sunset hours at our study site take pl ace between 6:18 - 6:48am and 6:28 - 6:48pm, respectively (Kolowski et al., 2007). Mean monthly daytime and nighttime temperatures in MMNR have been shown to average around 28.3 0 C and 13.8 0 C, respectively (Kolowski et al., 2007). We calculated the mean daily a monthly means. We then calculated the 2013 mean monthly minimum and maximum temperatures to be 14.94 0 C and 28.57 0 C, respectively. Based on our temperature records taken in TMC and NCG, TMC was, on average, a bit warmer than NCG (Fig. 2.2). That is, the mean monthly minimum temperature in TMC was 15.19 0 C, while in NCG it was 14.70 0 C. The mean monthly maximum temperature in TMC was 30.22 0 C, while in NCG it was 26.91 0 C (Fig.2.2). 28 Prey counts Throughout 2013, prey animals within each clan territory were counted in collaboration with Mara Hyena Project (MHP) personnel. Prey were counted twice every month within 100m (fixed widths) on either side of multiple, established 4 - km transects in each territory. Prey counted were then averaged to obtain monthly means. The density of the prey per month (animals/km 2 ) was estimated by dividing the mean number of prey counted in the territories per month by the t ransect area (km 2 ). The monthly prey density totals were added together and divided by 12 to get the mean monthly prey density for 2013. All negative mean monthly statistical analysis (Fig. 2.3). Seasons Rainfall affects ungulate population dynamics in the Mara - Serengeti ecosystem through its effect on vegetation growth and water availability (Ogutu et al., 2008). Similar effects have also been observed in the Laikipia district in Kenya (Georgiadis et al., 2007). Rainfall in the MMNR has been shown to average between 800 - 1200mm annually. The longest dry period takes place between mid - June and mid - October, and a shorter dry season in January and February (Stelfox et al., 1986, Ndegwa and Murayama, 2009). In 2013, daily rainfall was recorded at our camps in TMC and NCG, using a standard metric rain - gauge, and added together to get monthly totals (Fig. 2.4). The monthly totals were added to get the total for the year (2013) which was estimated at 1231.45mm and 1275.5mm at TMC and NCG respectively. The total rainfall recorded in 2013 was divided by 12 to get the mean monthly total rainfall 29 separately for TMC and NCG. All monthly rainfall deviations from the mean were referred to as negative. Reserve management regime Previous studies carried out within the NCG side of the Reserve estimated that up to 6,000 livestock (cattle, sheep and goats) grazed daily within this area. This livestock space - use overlapped with the 95% Utility Di stributions (UDs) of members of the Talek West (TW) clan (Boydston et al., 2003b). Nine ( 9) members of this clan (TW) were known to be killed at bomas next to the Reserve borders during livestock depredation in the period 2001 - 2005. Hyenas and other predat ors have also been reported to be killed indiscriminately by local community members in this area (e.g., 4 spotted hyenas and 1 lion were killed following livestock depredation) (Kolowski and Holekamp, 2006). In the NCG area, the value of livestock lost to depredation has been estimated to be about $6049 in 14 months in one earlier study (Kolowski and Holekamp, 2006). O ur study clan in the NCG area was the Talek West (TW) clan (Fig.1.2), which had a mean monthly clan size of 113 individuals in 2013. The TW territory size was approximately 77.04km 2 . In TMC, we studied two clans, namely Serena North (SN) and Serena South (SS) (Fig. 1.2) with clan sizes averaging 51 and 39 individuals, respectively. Territory size for SN and SS was approximately 42.67km 2 and 2 8.29km 2 respectively. The TW clan defended a territory along the northeastern border of MMNR next to an area densely inhabited by Masai pastoralists (Kolowski & Holekamp, 2006). In addition to exposure to tour vehicles, these hyenas were also exposed to in tensive daily livestock grazing pressure (Kolowski and Holekamp, 30 2009). The SN and SS clans defended territories near the western border of the MMNR, in TMC, where no human activities were allowed except wildlife viewing from tour vehicles. Statistical an alyses R v.3.1.3 statistical software (R core team, 2015) was used for all statistical analysis. Linear distances were calculated between any hyena locations between 10am - 1pm and 1pm - 4pm. We used Generalized Linear Models (GLM) to model the distance travel led by hyenas as a between management regime and rank (coded as rank*management), 7) interaction between management regime and prey (coded as prey*management) and 8) interaction between management regime and season (c oded as season*management). Because the linear distances travelled by the hyenas were not normally distributed, data were log transformed (log base 10) to acquire normality before statistical analysis. Subsequently, our model was as follows: Eqn 1: glm (l dist~rank+time+season+prey+management+rank*management+prey*management+season* management) where ldist was the Log transformed distances (log base 10). In this model, all our fixed factors (rank, management, prey, season, and time) had two We also added the interactions between management regime and rank (rank*management), 31 management regime and prey (prey*management) and management regime and season (season*management) in our model. Other interactions were not included in this mod el since our primary focus was to describe spotted hyena movement patterns in the two sides of the Reserve under different management regimes. In this chapter, sample sizes in our analyses were the daily distances calculated from the three daily locations in 2013 (total locations=6245; am=3164 locations, pm= 3081 locations). Two sample student t - tests were used to test for mean differences betwee n TMC and NCG in total monthly rainfall. Since mean monthly prey density numbers estimated in TMC were not normally distributed, a Wilcoxon Rank Sum test was used to test for prey density differences between TMC and NCG. We presented our means with stand ard errors, and differences between groups were considered significant when P < 0.05. Figure 2.2: Mean monthly minimum and maximum temperatures in TMC and NCG. 0 5 10 15 20 25 30 35 Temperature ( C) Temp (min') TMC Temp (min') NCG Temp' (max') TMC Temp' (max') NCG 32 Figure 2.3: Prey density (animals/km 2 ) in TMC and NCG. Figure 2.4: Total monthly rainfall recorded in the MMNR. 0.00 100.00 200.00 300.00 400.00 500.00 600.00 Density (animals/km ²) TMC NCG 0 50 100 150 200 250 300 350 400 Rainfall (mm) TMC NCG 33 Estimate Std. Error t value Pr(>|t|) (Intercept) 1.570892 0.045662 34.402 <2e - 16*** timepm - 0.326949 0.021200 - 15.422 <2e - 16*** rankL - 0.069507 0.036700 - 1.894 0.058283 managementtmc - 0.173692 0.049453 - 3.512 0.000447*** seasonwet 0.042689 0.049430 0.864 0.387827 preymany 0.034325 0.052001 0.660 0.509232 managementtmc:preymany 0.027157 0.059230 0.459 0.646608 managementtmc:rankL 0.201037 0.045149 4.453 8.63e - 06*** managementtmc:seasonwet 0.003157 0.056029 0.056 0.955070 Table 2.3: GLM table of results for distance moved. Distances have been transformed to log base 10 . Bolded cells indicate significant effects. Results Reserve management regime and distance travelled Our modeling showed that hyenas in the undisturbed side of the Reserve (TMC) travelled significantly shorter distances than did hyenas on the disturbed side (t= - 3.512, p=0.0004) (Table 2.3, Fig. 2.5). In fact, some of the members of the clan defending their territory in the distu rbed side of the reserve (TW) showed that they sometimes travelled all the neighboring town of Talek (Fig.2.1). This result was consistent with our hypothesis that the anthrop ogenic activities (e.g., livestock grazing) in NCG were at least partially responsible for 34 this unique behavior of hyenas. In TMC, hyenas appeared to behave more naturally, and with no disturbance allowed, they moved shorter distances. Time of the day and distance travelled Our results showed that time of day was important in determining how far the hyenas travelled. Hyenas were travelling significantly shorter distances in the afternoon (1pm - 4pm) than in the morning (10am - 1pm) (t= - 15.422, p<0.0001) (Tab le 2.3, Fig. 2.6). This movement pattern was expected because mornings were a bit cooler than afternoons throughout 2013. Figure 2.5: Management regime and average daily distance moved in TMC and NCG. Asterisk represents significant difference and error bars represent standard errors. 0 50 100 150 200 250 TMC NCG Distance (m) Management regime * 35 Figure 2.6: Time of day and average daily distance moved in the MMNR. Asterisk represents significant difference and error bars represent standard errors. Social r ank and distance traveled Earlier studies have shown that low - ranking (L) spotted hyenas hav e little access to food compared to high - ranking (H) ones (Tilson & Hamilton, 1984; Frank, 1986 b ). Therefore, we expected to see significant differences between distances travelled by low - ranking hyenas when compared to the high - ranking ones. As expected, on average, low - ranking hyenas travelled longer distances than did high - ranking hyenas (Fig. 2.7) in 2013. However, our modeling revealed a marginally non - significant difference between the distances travelled by low - ranking and high - ranking hyenas (t= - 1 .894, p=0.0583) (Table 2.3). Moreover, our modeling showed that the low - ranking hyenas in TMC were moving significantly longer distances than 0 50 100 150 200 250 10am - 1pm 1pm - 4pm Distance (m) Time of day * 36 were high - ranking hyenas (t=4.453, p<0.0001) (Table 2.3, Fig. 2.10) in this side of the reserve. However, we did n ot see the same rank - related variation in the NCG side of the Reserve. This might mean that in resources, as it does in pristine habitats. Prey and distance travelled Prey abundance has been documented as one of the factors affecting ranging patterns among large carnivores (Woodroffe, 2000). Therefore, we expected to see hyenas travelling significantly longer distances when prey were scarce than when prey were abundant. However, our results showed no difference in distances travelled during time of prey scarcity and abundance (t= - 0.660; p=0.5092) (Table 2.3, Fig. 2.8). We also expected to see hyenas in the disturbed side of the Reserve travelling longer distances when prey were abund ant due to anthropogenic activity. However, our modelling showed no significant difference between the distance travelled by hyenas in TMC and NCG when prey were abundant or scarce (t=0.459, p=0.6466) (Table 2.3, Fig. 2.11). In order to investigate whethe r prey density recorded in TMC was different from that in NCG, we averaged prey density numbers in TMC recorded per month and compared those to the monthly averages recorded in NCG. Because the mean prey density numbers in TMC and NCG were not normally dis tributed (even after log transformation to base 10 and testing for normality with a Shapiro Wilk test), a Wilcoxon Rank Sum test was used to test for the difference in their means. Our results showed that there was no significance difference in the density of prey recorded in the two sides despite the fact that TMC had slightly higher mean prey density recorded (w=85, p=0.4776) (Fig. 2.3). Average monthly prey density in TMC was 37 138.45 (animals/km 2 ) while that of NCG was 83.42 (animals /km 2 ), but variance w as high on both sides of the Reserve (Fig. 2.3). Seasonality and distance travelled Prey abundance has been shown to be higher in MMNR during the dry than wet season, as most ungulates concentrate during the dry season near watercourses within the Reserve (Ndegwa and Murayama, 2009). We therefore expected to see spotted hyenas travelling significantly shorter distances during the dry season than during the wet season. However, we found no significant differences between the distances moved by hyenas during the wet and dry seasons (t= 0.864, p=0.3878) (Table 2.3, Fig. 2.9). We also expected to see hyenas travelling longer distances during the wet season in NCG than TMC but our results showed no significant difference in this respect between the two sides of the Reserve (t=0.056, p=0.9551) (Table 2.3, Fig. 2.12) In order to investigate whet her total monthly rainfall received in TMC was different from that received in NCG, we compared the total monthly amount of rainfall that fell in the two areas. Data were log transformed to base 10 (tested normal with a Shapiro Wilk test) and tested with a (or with an independent) student t - test. Our results showed no significant difference between the total monthly rainfall received in TMC and NCG (t=0.3079, df=22, p=0.761) (Fig. 2.4). Our results showed NCG received slightly higher rainfall than TMC throu ghout 2013, total rainfall received was 1275.5mm and 1231.45mm in NCG and TMC, respectively. 38 Figure 2.7: Social rank and average daily distance moved in the MMNR. Error bars represent standard errors. Figure 2.8: Prey availability and average daily distance moved in the MMNR. Error bars represent standard errors. 0 50 100 150 200 250 High-ranking Low-ranking Distance (m) Rank 0 50 100 150 200 250 Abundant Scarce Distance (m) Prey 39 Figure 2.9 Seasonality and average daily distance moved in the MMNR. Error bars represent standard errors. Figure 2.10: Social rank and average daily distance moved in TMC and NCG. Asterisk represents significant difference, and error bars represent standard errors. 0 50 100 150 200 250 Wet Dry Distance (m) Season 0 50 100 150 200 250 TMC NCG Distance (m) High-ranking Low-ranking * 40 Figure 2.11: Prey availability and average daily distance moved in TMC and NCG. Error bars represent standard errors. Figure 2.12: Seasonal variation and average daily distance moved in TMC and NCG. Error bars represent standard errors. 0 50 100 150 200 250 TMC NCG Distance (m) Abundant Scarce 0 50 100 150 200 250 TMC NCG Distance (m) Wet 41 Frequency of movement Our results showed decreasing frequency of hyenas movement as distance increased. > 80% of the distances moved during the day were less than 200m and only ~7% of the distances moved were more than 500m (Figure 2.13). See also table 2.14 for average monthly distances moved by hyenas in the year 2013. Figure 2.13: Diurnal frequency of mo vement by spotted hyenas in the MMNR. 0 500 1000 1500 2000 2500 3000 3500 4000 4500 No. of observations Distance moved (m) 42 Month Average distance moved (m) in TMC Average distance moved (m) in NCG Jan 99.64 113.53 Feb 106.06 111.52 March 106.83 181 April 126.92 111.72 May 141.33 135.64 June 123.79 132.69 July 86.21 158.12 Aug 85.77 222.47 Sep 132.41 184 Oct 118.67 281.22 Nov 141.44 219.55 Dec 147.56 175.73 Table 2. 4: Table of average diurnal monthly distances travelled by hyenas in 2013 in the MMNR. Discussion Management regime and movement patterns Previous research in NCG has shown that female spotted hyenas defending their territory in an area with intensive livestock grazing remained active during the morning hours more than did hyenas defending their territory in an area with no livestock grazing allowed (Kolowski et al., 2007). Our results were thus consistent with this study because we did find 43 hyenas travelling significantly shorter distances in the undisturbed area than in the disturbed area (Fig. 2.5). This finding was similar to other previo us studies showing that large carnivores alter their activity patterns in response to anthropogenic activities (e.g. Kitchen, Gese & Schauster, 2000); in the MMNR it was during the day when intensive livestock grazing occurred in 2013 within NCG. Our findi ng that hyenas were traveling longer distances in the morning was consistent with that of Kolowski et al., (2007) (mentioned above). Social rank and distance moved Earlier studies indicated that hyenas of low social rank have low priority of access to r hierarchy in the clan determines its priority of access to food (Tilson & Hamilton, 1984; Frank, 1986b). Earlier studies done in NCG have shown that low - ranking fema le hyenas with no den - dwelling cubs travelled longer distances than did high - ranking female hyenas with no den - dwelling cubs (Boydston et al., 2003a). Another study in Tanzania also found that low - ranking hyenas travelled longer distances in search of food than did high - ranking hyenas during times of low prey abundance (Hofer and East, 1993a). Therefore, we expected significant differences between the distances travelled by low - ranking hyenas and those travelled by high - ranking ones. However, our results sh owed no significant difference between the distances travelled by low and high - ranking hyenas. This result was consistent with those of Kolowski et al., (2007) who found no significant difference between the percent of time high - ranking females spent whil e travelling when compared to the low - ranking females over a 24 - h period in TW clan. Consistent with our prediction and with results from the studies mentioned above, our results revealed a significant difference in distance travelled between low and high - ranking hyenas in 44 the undisturbed side of the reserve (TMC). It may be that we saw no significant effects of rank on the NCG side of the Reserve because hyenas of all ranks on that side are affected similarly by human activity (Fig. 2.7). This hypothesis i s also supported by our finding that high rank individuals from the east side travel further than low rank individuals on the west side. Time of day and movement patterns Earlier studies have shown that spotted hyenas tend to become almost entirely noctur nal in their daily activities in hot, dry habitats (Tilson & Hamilton, 1984; Cooper, 1990). This was consistent with our finding that hyenas were moving longer distances during the cooler morning hours (10am - 1pm) than during the warmer afternoon hours (1pm - 4pm) (Fig. 2.6). Other than temperature differences, our finding may also be due to reduced anthropogenic activity, and thus less disturbance, in TMC. Unfortunately we cannot distinguish between these hypotheses. The implementation of a strict managem ent plan in TMC may be benefiting the hyenas living there. In NCG, livestock grazing takes place (despite being illegal in the country; Kenya Wildlife Act, 1989; Walpole et al., 2003) on a daily basis, and this may be causing the local hyenas to move longe r distances. Our results are consistent with those of earlier researchers who have shown that daily trends in temperature do affect the optimal activity patterns of terrestrial carnivores (Avenant and Nel, 1998). Normally the onset of activities for spotte d hyenas is observed to resume around sunset and end early in the morning (Kolowski et al., 2007; Kruuk, 1972). By contrast, here we find the hyenas traveling even in the afternoon (1pm - 4pm) when temperatures are high. Despite having been shown to be noctu rnal hunters (Cooper, 1990), here we also find 45 spotted hyenas moving about during the daytime, which signifies a change in their normal behavior. Prey availability and movement patterns Prey abundance has been shown to affect space use of large carnivore s (Macdonald, 1983). Large carnivores feed on animals at lower trophic levels than themselves (Treves and Karanth, 2003). Large carnivores regulate numbers of their prey through predation (Estes et al. 1998; Berger et al., 2001; Terborgh et al. 2001). Lion s were found to increase their attacks on humans in areas with few prey except bush pigs (Potamochoerus larvatus) (Packer et al., 2005). D istribution and densities of spotted hyenas were also found to be prey - dependent throughout Africa (Hayward et al., 20 09; Hofer and East 1993b; Mills 1989; Trinkel et al., 2004). Reproductive success in female spotted hyenas is directly correlated with priority of access to food (Holekamp et al., 1996). In east Africa, the diet of spotted hyenas is composed mainly of medi um to large - sized ungulates (Kruuk, 1972). In the MMNR, the resident wildlife are joined annually from July to at least October (Ogutu et al., 2005) by large migratory herds of herbivores from the Serengeti National Park, and this affects prey availabilit y in the Reserve . Therefore, we expected that hyenas would travel longer distances when prey were scarce than when prey were abundant. In contrast to our expectation, our results showed do difference in distance travelled during times of prey scarcity nor during times of prey abundance. We also did not see significant differences in distance travelled on either side of the Reserve when prey were scarce or when prey were abundant (Figs. 2.8 and 2.11). These results were similar to those from a study by Boy dston et al., (2003a), who found that prey availability did not significantly affect space use patterns of 46 female hyenas with den - dwelling cubs in NCG. Earlier studies with the TW clan found that hyenas were actually avoiding prey rich areas, as these were the same areas utilized most heavily by livestock. Hyenas have also been shown to be scared by the herders who guarded their livestock in NCG area (Boydston et al., 2003a; Kolowski and Holekamp, 2009; KE Holekamp unpublished data). In fact, Boydston et al ., (2003b) in her study on TW clan did observe hyenas being displaced from their resting places by cattle, and also observed hyenas emerging from bushy cover immediately after livestock herds passed through an area. Seasonal variation in movement Desert hyenas have been documented to disperse from their territories when their water supply failed (Tilson and Henschel, 1986). Hyenas in the Savuti area of Botswana were observed to travel longer distances in search of drinking water during the dry season (Co oper, 1989). Another study done in Ruaha National Park, Tanzania, revealed a positive correlation between large carnivore distributions and water availability (Abade et al., 2014). Lions were also found to select areas that were within 2km of a waterhole, and they also covered shorter distances than when they were farther from a waterhole (Valeix et al., 2010). However, the MMNR has reliable water sources throughout the year in both TMC and NCG, and water availability seems not to be limited in the Reserve even in the dry season. Therefore we predicted that hyenas would be travelling longer distances during the wet season than they would do during the dry season. This is because it is during the wet season when some prey move out of the Reserve into the surr ounding ranches, but then move back into the Reserve during the dry season (Ndegwa and Murayama, 2009). Here we found that distances travelled by collared MMNR hyenas during daylight hours did not differ significantly between the wet 47 and dry season (Fig. 2 .9). We expected to see hyenas travelling shorter distances during the dry season than during the wet season but we found no seasonal differences in daytime movements. Earlier studies have documented that prey are sparsely distributed in MMNR during the we t season (Ndegwa and Murayama, 2009), forcing hyenas to travel further. However we found no significant seasonal variation in daytime movements. Conclusion Human wildlife conflict can lead to species extinctions (Woodroffe and Ginsberg, 1998). Our results suggest that hyenas in the undisturbed side of the Reserve were travelling significantly shorter distances than were hyenas in the disturbed side (Fig. 2.5). We also noted that collared hyenas in the disturbed side of the Reserve were travelling past the nearby town of Talek and into Tanzania (Fig. 2.1). This can be expected to bring them into conflict with humans living around the Reserve (see Kolowski et al., 2006). This situation jeopardizes the survival of this species in this important ecosystem. Ther efore, we are concerned that these could be warning signs of drastic negative changes to this ecosystem in the near future. We suggest that a good management system should be put in place to safeguard the future of hyenas and other large carnivores. Our f indings also showed that hyenas were travelling significantly longer distances in the mornings than they were in the afternoons (Fig. 2.6). This concerns us since the movement patterns investigated here occurred during the time interval each day when spott ed hyenas are generally inactive (Kolowski et al., 2007). Hyenas should be using their inactive time resting or nursing their young ones; so we worry that the patterns seen here could in the long run affect the welfare of hyenas in NCG. 48 We found that there was a greater difference in distances traveled between high and low - ranking hyenas in the undisturbed side than the disturbed side of the Reserve (Fig. 2.10). This result in the undisturbed side of the reserve was expected, but we were surprised to find that high - ranking hyenas in the disturbed side of the Reserve were actually traveling longer distances than were low - ranking hyenas (Fig. 2.10). We suggest that the differences in rank - related variation in movements revealed by our findin gs could potentially serve as early warning signs that the effects of social rank in spotted hyenas are being superseded by effects of anthropogenic disturbance. Clearly we need empirical conservation policies to ensure long - term survival of this and other large carnivore species in the MMNR. The current situation is only expected to become worse, as suggested by previous studies which showed declining trends in resident herbivores in the MMNR (Ottichilo et al., 2000, 2001). These herbivores serve as the pr ey base for hyenas and other predators in the reserve. Just as it has been suggested that protected areas attract human settlement (Wittemyer et al., 2008), human population round the MMNR has been observed to be increasing (Ndegwa and Murayama, 2009). In creased rates of human - wildlife conflict are expected, especially around NCG, just as spotted hyenas have been observed to attack more livestock at villages closer to Reserve borders (see Kolowski et al., 2006). Therefore we suggest an all - inclusive manage ment strategy that could serve as a way forward for the survival of man and wildlife in this ecosystem. Such management strategies would work towards involving local community members to discuss matters of wildlife conservation in and around MMNR, review l and tenure and use systems, consider benefit sharing with the locals and educating them on matters of wildlife conservation. 49 REFERENCES 50 REFERENCES Abade, L., Macdonald, D. W., & Dickman, A. J. (2014a). Using landscape and bioclimatic features to predict the distribution of lions, leopards and spotted Hyaenas in Tanzania's Ruaha Landscape. Avenant, N. L., & Nel, J. A. (1998). Home range use, activity, and density of caracal in relation to prey density. African Journal of Ecology , 36 (4) , 347 - 359. Bailey, H., & Thompson, P. (2006). Quantitative analysis of bottlenose dolphin movement patterns and their relationship with foraging. Journal of Animal Ecology , 75 (2), 456 - 465. Berger, J., Stacey, P. B., Bellis, L., & Johnson, M. P. (2001). A m ammalian predator - prey imbalance: grizzly bear and wolf extinction affect avian neotropical migrants. Ecological Applications , 11 (4), 947 - 960. Boydston, E. E., Kapheim, K. M., Watts, H. E., Szykman, M., & Holekamp, K. E. (2003a). Altered behaviour in spott ed hyenas associated with increased human activity. Animal Conservation , 6 (3), 207 - 219. Boydston, E. E., Kapheim, K. M., Watts, H. E., Szykman, M., & Holekamp, K. E. (2003b). Altered behaviour in spotted hyenas associated with increased human activity. Ani mal Conservation , 6 (3), 207 - 219. Cooper, S. M. (1989). Clan sizes of spotted hyaenas in the Savuti Region of the Chobe National Park, Botswana. Botswana Notes and Records , 21 , 121 - 131. Cooper, S. M. (1990). The hunting behaviour of spotted hyaenas (Crocuta crocuta) in a region containing both sedentary and migratory populations of herbivores. African Journal of Ecology , 28 (2), 131 - 141. Dickman, A. J., Hazzah, L., Carbone, C., & Durant, S. M. (2014). Carnivores, culture and ctors influence perceived problems with carnivores in Biological Conservation , 178 , 19 - 27. Estes, J. A., Tinker, M. T., Williams, T. M., & Doak, D. F. (1998). Killer whale predation on sea otters linking oceanic and nearshore ec osystems. Science , 282 (5388), 473 - 476. Etzenhouser, M. J., Owens, M. K., Spalinger, D. E., & Murden, S. B. (1998). Foraging behavior of browsing ruminants in a heterogeneous landscape. Landscape Ecology , 13 (1), 55 - 64. Frank, L. G. (1986b). Social organization of the spotted hyaena Crocuta crocuta. II. Dominance and reproduction. Animal Behaviour , 34 (5), 1510 - 1527. 51 Frank, L. G., Glickman, S. E., & Powch, I. (1990). Sexual dimorphism in the spotted hyaena (Crocuta crocuta ). Journal of Zoology , 221 (2), 308 - 313. Gittleman, J. L. (Ed.). (2001). Carnivore conservation (Vol. 5). Cambridge University Press. Georgiadis, N. J., Olwero, J. N., & Romañach, S. S. (2007). Savanna herbivore dynamics in a livestock - dominated landsca pe: I. Dependence on land use, rainfall, density, and time. Biological Conservation , 137 (3), 461 - 472. Green, D.S. (2015). Anthropogenic disturbance, ecological change, and wildlife conservation at the edge of the Mara - Serengeti ecosystem. Ph.D. Dissertatio n. Michigan State University. Hayward, M. W., Hayward, G. J., Druce, D. J., & Kerley, G. I. (2009). Do fences constrain predator movements on an evolutionary scale? Home range, food intake and movement patterns of large predators reintroduced to Addo Eleph ant National Park, South Africa. Biodiversity and Conservation , 18 (4), 887 - 904. Hofer, H., & East, M. L. (1993a). The commuting system of Serengeti spotted hyaenas: how a predator copes with migratory prey. I. Social organization. Animal behaviour , 46 (3), 547 - 557. Hofer, H., & East, M. L. (1993b). The commuting system of Serengeti spotted hyaenas: how a predator copes with migratory prey. II. Intrusion pressure and commuters' space use. Animal Behaviour , 46 (3), 559 - 574. Holekamp, K. E., Smale, L., & Szykman , M. (1996). Rank and reproduction in the female spotted hyaena. Journal of Reproduction and Fertility, 108(2), 229 - 237. Holekamp, K. E., Smith, J. E., Strelioff, C. C., Van Horn, R. C., & Watts, H. E. (2012). Society, demography and genetic structure in the spotted hyena. Molecular Ecology , 21 (3), 613 - 632. Kitchen, A. M., Gese, E. M., & Schauster, E. R. (2000). Changes in coyote activity patterns due to reduced exposure to human persecutio n. Canadian Journal of Zoology , 78 (5), 853 - 857. Kolowski, J. M., & Holekamp, K. E. (2006). Spatial, temporal, and physical characteristics of livestock depredations by large carnivores along a Kenyan reserve border. Biological conservation, 128(4), 529 - 541 . Kolowski, J. M., & Holekamp, K. E. (2009). Ecological and anthropogenic influences on space use by spotted hyaenas. Journal of zoology, 277(1), 23 - 36. Kolowski, J. M., Katan, D., Theis, K. R., & Holekamp, K. E. (2007). Daily patterns of activity in the s potted hyena. Journal of Mammalogy, 88(4), 1017 - 1028. Kotliar, N. B., & Wiens, J. A. (1990). Multiple scales of patchiness and patch structure: a hierarchical framework for the study of heterogeneity. Oikos , 253 - 260. 52 Kruuk, H. (1972). The spotted hyena: a study of predation and social behavior. Chicago: University of Chicago Press. Macdonald, D. W. (1983). The ecology of carnivore social behaviour. Nature , 301 (5899), 379 - 384. Mills, G., & Hofer, H. (1998). Status survey and conservation action plan. Hyaenas. IUCN/SSC Hyaena Specialist Group, IUCN, Switzerland. Mills, M. G. (1984). Prey selection and feeding habits of the large carnivores in the southern Kalahari. Koedoe , 27 (2), 281 - 294. Mi lls, M. G. L. (1989). The comparative behavioral ecology of hyenas: the importance of diet and food dispersion. In Carnivore behavior, ecology, and evolution (pp. 125 - 142). Springer US. Ndegwa Mundia, C., & Murayama, Y. (2009). Analysis of land use/cover c hanges and animal population dynamics in a wildlife sanctuary in East Africa. Remote Sensing , 1 (4), 952 - 970. Nowell, K., & Jackson, P. (Eds.). (1996). Wild cats: status survey and conservation action plan (pp. 1 - 382). Gland: IUCN. Ogutu, J. O., Bhola, N ., & Reid, R. (2005). The effects of pastoralism and protection on the density and distribution of carnivores and their prey in the Mara ecosystem of Kenya. Journal of Zoology, 265(3), 281 - 293. Ogutu, J. O., Piepho, H. P., Dublin, H. T., Bhola, N., & Reid, R. S. (2008). Rainfall influences on ungulate population abundance in the Mara Serengeti ecosystem. Journal of Animal Ecology , 77 (4), 814 - 829. Ordiz, A., Støen, O. G., Sæbø, S., Sahlen, V., Pedersen, B. E., Kindberg, J., & Swenson, J. E. (2013). Lasting b ehavioural responses of brown bears to experimental encounters with humans. Journal of Applied Ecology , 50 (2), 306 - 314. Ottichilo, W. K., de Leeuw, J., & Prins, H. H. (2001). Population trends of resident wildebeest [Connochaetes taurinus hecki (Neumann)] and factors influencing them in the Masai Mara ecosystem, Kenya. Biological Conservation , 97 (3), 271 - 282. Ottichilo, W. K., De Leeuw, J., Skidmore, A. K., Prins, H. H., & Said, M. Y. (2000). Population trends of large non migratory wild herbivores and live stock in the Masai Mara ecosystem, Kenya, between 1977 and 1997. African Journal of Ecology , 38 (3), 202 - 216. Packer, C., Ikanda, D., Kissui, B., & Kushnir, H. (2005). Conservation biology: lion attacks on humans in Tanzania. Nature , 436 (7053), 927 - 928. Pac ker, C., Loveridge, A., Canney, S., Caro, T., Garnett, S. T., Pfeifer, M., ... & Plumptre, A. (2013). Conserving large carnivores: dollars and fence. Ecology letters , 16 (5), 635 - 641. 53 R Core Team (2015). R: A language and environment for statistical computi ng. R Foundationfor Statistical Computing, Vienna, Austria. URL http://www.R - project.org/ . Ripple, W. J., Estes, J. A., Beschta, R. L., Wilmers, C. C., Ritchie, E. G., Hebblewhite, M., ... & Wirsing, A. J. (2014). Science , 343 (6167), 1241484. Sindiga, I. (1996). Tourism education in Kenya. Annals of Tourism Research , 23 (3), 698 - 701. Seryodkin, I. V., Kostyria, A. V., Goodrich, J. M., & Miquelle, D. G. (2013). Daily activity patterns of brown bear (Ursus arctos) of the Sikhote - Alin mountain range (Primorskiy Krai, Russia). Russian journal of ecology , 44 (1), 50 - 55. Smale, L., Frank, L. G., & Holekamp, K. E. (1993). Ontogeny of dominance in free - living spotted hyaenas: juvenile rank relations with adult females and immigrant males. Animal Behaviour , 46 (3), 467 - 477. Stelfox, J. G., Peden, D. G., Epp, H., Hudson, R. J., Mbugua, S. W., Agatsiva, J. L., & Amuyunzu, C. L. (1986). Herbivore dynamics in southern Narok, Kenya. The Journal of wildlife management , 339 - 347. Ecological meltdown in predator - fr ee forest fragments. Science , 294 (5548), 1923 - 1926. The Wildlife (Conservation and Management) (Amendment) Act, 1989. Theuerkauf, J., Rouys, S., & Jedrzejewski, W. (2003). Selection of den, rendezvous, and resting sites by wolves in the Bialowieza Forest, Poland. Canadian Journal of Zoology , 81 (1), 163 - 167. Tilson, R. L., & Hamilton, W. J. (1984). Social dominance and feeding patterns of spotted hya enas. Animal Behaviour , 32 (3), 715 - 724. Tilson, R. L., & Henschel, J. R. (1986). Spatial arrangement of spotted hyaena groups in a desert environment, Namibia. African Journal of Ecology , 24 (3), 173 - 180. e conflict and perspectives on carnivore management worldwide. Conservation Biology, 17(6), 1491 - 1499. Trinkel, M., Fleischmann, P. H., Steindorfer, A. F., & Kastberger, G. (2004). Spotted hyenas (Crocuta crocuta) follow migratory prey. Seasonal expansion of a clan territory in Etosha, Namibia. Journal of Zoology, 264(2), 125 - 133. Valeix, M., Loveridge, A. J., Davidson, Z., Madzikanda, H., Fritz, H., & Macdonald, D. W. (2010). How key habitat features influence large terrestrial carnivore movements: waterho les and African lions in a semi - arid savanna of north - western Zimbabwe. Landscape Ecology , 25 (3), 337 - 351. 54 Van Dyke, F. G., Brocke, R. H., Shaw, H. G., Ackerman, B. B., Hemker, T. P., & Lindzey, F. G. (1986). Reactions of mountain lions to logging and huma n activity. The Journal of Wildlife Management , 95 - 102. Walpole, M. J. (2003). Wildlife and People: Conflict and Conversation in Masai Mara, Kenya: Proceedings of a Workshop Series... 13 - 16 August 2001. (No. 14). IIED. Western, D., Russell, S., & Cuthill, I. (2009). The status of wildlife in protected areas compared to non - protected areas of Kenya. PLoS One , 4 (7), e6140. Wittemyer, G., Elsen, P., Bean, W. T., Burton, A. C. O., & Brashares, J. S. (2008). Accelerated human population growth at protected area edges. Science , 321 (5885), 123 - 126. Woodroffe, R. (2000). Predators and people: using human densities to interpret declines of large carnivores. Animal conservation , 3 (02), 165 - 173. Woodroffe, R., & Ginsberg, J. R. (1998). Edge effects and the extinction o f populations inside protected areas. Science , 280 (5372), 2126 - 2128. 55 CHAPTER 3 HABITAT USE PATTERNS BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RESERVE, KENYA INTRODUCTION Habitat heterogeneity plays a key role in determining animal movement patterns as well as behavioral and ecological processes, such as feeding behavior (e.g. Etzenhouser et al. 1998; Bailey and Thompson 2006). Behavioral changes in habitat use, are often the first measurable reactions exhibited by animals to human - induced environmental changes (Tuomainen and Candolin, 2011). Animals seldom range randomly (Duncan, 1983), and use certain types of space disproportionally (Samuel et al., 1985). Land cover ch ange is thus an important component of wildlife habitat, as its alteration has important implications for animal species distribution (Balmford et al., 2001; Brooks et al., 2002). Habitat loss and fragmentation have been shown to be among the drivers of de clining populations of large carnivores in the world over the last 100 years (Gittleman et al., 2001). Habitat loss also leads to species extinction or causes species to be threatened with extinction (Brooks et al., 2002). Habitat loss has also been shown to influence animal population dynamics (Ogutu et al., 2009). Large carnivore distribution and habitat selection are mostly determined by prey availability (Hayward et al., 2009; Valeix et al., 2010). Prey availability in turn is affected by vegetation co ver, water availability and elevation (Valeix et al., 2010; Pita et al., 2009; Schadt et al., 2002). The majority (61%) of largest carnivore species are currently listed as threatened by the International Union for Conservation of Nature (IUCN) (Ripple et al., 2014). In Africa, the spotted hyena ( Crocuta crocuta ) has been shown to have altered its space use patterns due to 56 competition for space with livestock (Boydston et al., 2003b). This change in behavior has occurred in the Masai Mara National Reserve food source for herbivores in the MMNR, also provides refuges to lions and spotted hyenas exposed to people and l ivestock (Boydston et al., 2003b; Kolowski and Holekamp, 2009). Vegetative cover in general near the MMNR has declined over the last two decades (e.g. Homewood et al. , 2001; Serneels, Said & Lambin, 2001; Ndegwa and Murayama, 2009). This is due to signific ant expansion in farmland, which now occupies areas that were previously natural grasslands, and were used by wildlife for dispersal, breeding and or calving areas (Ndegwa and Murayama, 2009). Although spotted hyenas have been shown to occupy habitats ran ging from deserts, through grasslands to wooded savannas (Cooper, 1989), their survival is currently faced with many challenges that need to be addressed. With more land expected to be converted into grazing land for livestock, given the global demand for meat (McAlpine, 2009), we fear that survival of this iconic species in the Mara may be jeopardized in the near future. Although three prior studies investigated habitat preferences by spotted hyenas in relation to anthropogenic activities in the MMNR (e.g ., Boydston et al., 2003a, Boydston et al., 2003b, Kolowski and Holekamp, 2009), none of these studies focused specifically on diurnal habitat use patterns, nor did they compare space use between spotted hyena clans inhabiting the same protected area but e xperiencing different management regimes. Therefore, in this chapter we investigate daytime habitat use patterns in the MMNR by adult female spotted hyenas living in three different clans (Fig. 1.2). One of the clans defended a territory in an area 57 where i technology to track the adult female hyenas with Global Positioning System (GPS ) collars (as described in Chapter 2). Here we compared habitat usage between TMC and NCG by investigating patterns of habitat use by 22 adult female spotted hyenas in the MMNR for a period of one year, Jan 1 - Dec 31, 2013. We tested hypotheses suggesting that the two different management styles in TMC and NCG would be associated with different patterns of vegetative cover, and that hyenas in TMC and NCG would use the available habitat types differently. Due to the high levels of anthropogenic activities i n NCG, we expected to see hyenas avoiding habitats where livestock grazing took place but we did not expect to see such behavior in TMC. Methods Study site and study animals Details of our study area and study animals are as described in Chapters 1 and 2. Generating a classified vegetation map of the Masai Mara National Reserve (MMNR), Kenya We used ESRI ArcMap v. 10.2.2 and R v.3.1.3 statistical software (R core team, 2015) to create a classified land cover raster map for the Mara from Google Earth and Landsat images (Landsat 5 TM June 2009). The Landsat image was downloaded from the United States Geological Society (USGS) and we maintained our pixels at a 30m resolution. Landsat images were selected after consideration of cloud cover and other features that might have influenced our final classified image (map). We utilized the seven elements of photo interpretation (shape, pattern, tone, association, size, texture and shadow) in our creation of random forests within the selected 58 Landsat and Google Earth images (see Bourgeau - Chavez et al., 2015). We generated polygons that met our criteria to represent different land cover types and these served as our training data. The polygons were then converted into shape files using ESRI ArcMap v. 10.2.2, and assig ned different numbers to represent specific land cover classes. We made sure that our training data had a representative sample of each vegetation class, and was well distributed within our study clan territories in 2013 (Fig. 1.2). The territories were ge nerated from Kernel density 95% utility distributions (UDS) using all hyenas that had radio collars in the year 2013 (Green, 2015). We then assigned same numbers to similar land cover types, and our classification was as follows; 1 represented buildings, 2 represented grasslands, 3 represented shrub lands, 4 represented bare ground, 5 represented forests, 6 represented rivers, and 7 represented wetlands. All the shape files were assigned a Geographic Coordinate System (GCS) of WGS 84 and projected in Trans verse Mercator UTM Zone 36S, which is the projection system used in the Masai Mara National Reserve. We then used the statistical software R v.3.1.3 (R core team, 2015) to generate the final raster image representing the classified and combined vegetation cover maps for the Mara using the already assigned vegetation classes (Figs. 3.1, 3.2). Because most of our training data were concentrated within the territories of our study clans, we suggest that our final classified image may not be a perfect fit for u se in the entire MMNR, but it is valid within our clan territories, and it is for this reason that our habitat analysis was carried out within the territories defended by our study animals. 59 Statistical analys es R v.3.1.3 statistical software (R core team, 2015) and ESRI ArcMap v.10.2.2 were used in cover map for MMNR (Fig. 3.3; see methods above), ran queries on the projected data, and subsequently used Chi - squared tests in our analysis. A Chi - squared test for independence was used to test whether habitat use in either side of the reserve differed significantly from what was available. Both use and availability were d etermined from counts of pixels. We also used Chi - square tests to compare habitat availability in either side of the Reserve, and habitat use by collared spotted hyenas within and outside of their defended territories (although here we only used the hyena locations). Although we initially classified habitats as 7 different types (Fig. 3.1), we later combined some cover classes in our final habitat analysis. Buildings, which represented only 2% of the total habitat in both TMC and NCG, were excluded in our final analysis. Cover classes that were forests occurred along the rivers . Therefore, we used grasslands, shrub lands, bare ground and riparian vegetation classes in our final analysis (Table 3.1 and 3.2). We also ran queries on the projected data to estimate the number of hyena locations that fell outside of their d efended territories (Fig. 3.4). 60 Results Habitat types availability in TMC and NCG Our results showed different habitat types available to hyenas in TMC and NCG. In TMC, grassland dominated the ecosystem (81.48% of the total pixels), shrub land represent ed 7.02%, while riparian and bare ground represented 6.17% and 5.36%, respectively. In NCG, grassland covered 48.82% of the total pixels, bare ground covered 28.11%, while shrub land and riparian covered 21.34% and 1.72%, respectively. As predicted by our hypothesis, our results showed different habitat types available for hyenas in TMC and NCG. Chi - square test of independence results showed that available habitats differed significantly between TMC and NCG ( 2 =26628.21, df =3, p = 0.0001) (Tables 3.1, 3.2 , and 3.3). Habitat use relative to availability in TMC and NCG Habitat use relative to availability differed between TMC and NCG. Out of the total habitats used in TMC, 82.59% was grassland, 6.96% was shrub land, 6% was bare ground and 4.45% was riparian . Out of the total habitats used in NCG, 25.57% was grassland, 45.55% was shrub land, 13.05% was bare ground and 15.82% was riparian (Tables 3.1 and 3.2). Results from Chi - square tests of independence showed that hyenas in TMC were using habitats in propor tion to their availability ( 2 =0.3236, df=3, p = 0.9555) while habitat use by hyenas in NCG differed significantly from habitat availability ( 2 =32.8575, df =3, p = 0.0001) (Tables 3.4 and 3.5). Grassland habitat use Our results showed that hyenas in TMC were using grassland habitat in proportion to its availability while hyenas in NCG used far less of grassland habitat than was available ( 2 =1603.246, df =1, p < 0.0001) (Table 3.6). 61 Shrub land habitat use Shrub land habitat use relative to availability did not differ significantly between TMC and NCG ( 2 = 1.2114, df =1, p = 0.271), although NCG hyenas used shrub land habitat slightly more than expected (Table 3.6). Bare ground habitat use Our results showed bare ground habitat use relative to availabil ity differed significantly between TMC and NCG. While hyenas in TMC were using bare ground as expected based on its availability, hyenas in NCG used significantly less bare ground habitat than was available to them ( 2 = 599.3769, df =1, p <0.0001) (Table 3.6) . We recognize that the greater abundance of bare ground in NCG (28.11%) than TMC (5.36%) is most likely due to intensive trampling and removal of vegetation by cattle while grazing or traveling to grazing sites. 62 Figure 3.1: Classified habitat map o f the MMNR showing the 7 habitat types . 63 Figure 3.2: Classified habitat map of the MMNR showing the combined habitats . 64 Figure 3. 3 : Classified habitat map of the MMNR showing hyena locations . 65 Figure 3. 4 : Classified habitat map of the MMNR showing habitat use outside defended territories . Riparian habitat use Our results showed that riparian habitat use relative to availability differed significantly between TMC and NCG. While hyenas in TMC used riparian habitat in proportion to its availability, N CG hyenas significantly used riparian habitat more than it was available ( 2 = 654.8744, df =1, p <0.0001) (Table 3.6) . 66 Habitat use outside the clan territories Habitat use by hyenas outside of their defended territories differed between TMC and NCG. External habitat use represented 2.82% and 3.72% of all hyena locations in TMC and NCG respectively (Tables 3.7, 3.8 , and 3.9; Fig. 3.4). Results of a Chi - square test of independence showed that NCG hyenas used habitats outside of their defended territory significantly more than did TMC hyenas ( 2 = 6.2353, df =1, p = 0.0125) (Table 3.9, Fig.3.4) . Habitat Available pixels Used pixels % available % used Grassland 62860 5350 81.48% 82.59% Shrub land 5418 451 7.02% 6.96% Bare ground 4135 389 5.36% 6% Riparian 4758 288 6.17% 4.45% Table 3.1: Habitat classification in TMC hyena territories . Habitat Available pixels Used pixels % available % used Grassland 41603 905 48.82% 25.57% Shrub land 18191 1612 21.34% 45.55% Bare ground 23961 462 28.11% 13.05% Riparian 1470 560 1.72% 15.82% Table 3.2: Habitat classification in NCG hyena territor y. 67 Habitat availability TMC NCG Chi - square Grassland available pixels 62860 41603 2 = 26628.21, df = 3, p<0.0001 Shrub land available pixels 5418 18191 Bare ground available pixels 4135 23961 Riparian available pixels 4758 1470 Table 3.3: Chi - square test results for differences in available habitats in TMC and NCG. Bolded cells represent significant effects . Habitat % habitat available % habitat used Chi - square Grassland 81.455469 (82.021344) 82.587218 (82.021344) 2 = 0.3236, df = 3, p= 0.9555 Shrub land 7.020772 (6.991399) 6.962025 (6.991399) Bare ground 5.358230 (5.681585) 6.004940 (5.681585) Riparian 6.165529 (5.305673) 4.445817 (5.305673) Table 3.4: Chi - square test results for habitat use in TMC. Expected values are presented in parenthesis. 68 Habitat % habitat available % habitat used Chi - square Grassland 48.815488 (37.193842) 25.57220 (37.193842) 2 = 32.8575, df = 3, p= 0.0001 Shrub land 21.344676 (33.447133) 45.54959 (33.447133) Bare ground 28.114990 (20.584762) 13.05454 (20.584762) Riparian 1.724846 (8.774263) 15.82368 (8.774263) Table 3.5: Chi - square test results for habitat use in NCG. Expected values are presented in parenthesis. Bolded cells represent significant effects. Habitat TMC NCG Chi - square Grassland available pixels 62860 (64356.484) 41603 (40106.516) p<0.0001 Grassland used pixels 5350 (3853.516) 905 (2401.486) Shrub land available pixels 5418 (5397.3676) 18191 (18211.632) p=0.271 Shrub land used pixels 452 (471.6324) 1612 (1591.368) Bare ground available pixels 4135 (4391.0009) 23961 (23704.9991) p<0.0001 Bare ground used pixels 389 (132.9991) 462 (718.0009) Riparian available pixels 4758 (4441.2787) 1470 (1786.7213) p<0.0001 Riparian used pixels 288 (604.7231) 560 (243.2787) Table 3.6: Chi - square test results for differences in habitat use compared to available habitats in TMC and NCG. Expected values are presented in parenthesis. Bolded cells represent significant effects. 69 Name Inside territory locations Outside territory locations Total locations Bart 316 1 317 Clov 387 0 387 Digs 542 8 550 Ema 816 119 935 Hndy 307 6 313 Java 947 1 948 Mtn 368 16 384 Peep 108 0 108 Shrm 938 6 944 Slin 317 7 324 Taj 1005 19 1024 Tnsl 369 5 374 Wafl 290 7 297 Zoey 18 0 18 Table 3. 7: Hyena locations inside and outside their defended territories in TMC. 70 Name Inside territory locations Outside territory locations Total locations Baez 357 29 386 Hel 957 0 957 Hex 195 8 203 Juno 358 1 359 Mgta 319 0 319 Pan 237 3 240 Roos 26 3 35 298 Tilt 883 62 945 Table 3. 8 : Hyena locations inside and outside their defended territor y in NCG. TMC NCG Chi - square Outside locations 195 (216.8729) 138 (116.1271) 2 = 6.2353, df = 1, p=0.0125 Inside locations 6728 (6706.1271) 35 69 ( 3590.8729) Table 3.9: Chi - square test results for habitat use within and outside defended territories in TMC and NCG. Expected values are presented in parenthesis. Bolded cells represent significant effects . 71 Discussion Non - uniform use of habitat has been demonstrated amongst a number of terrestrial mammal species (Gates, 1979; Duncan, 1983; Samuel et al., 1985). Habitat use studies have also been shown to guide management policies (Ingram and Rogan, 2002) . Successful management of predator and prey populations requires detailed knowledge of regulatory factors within populations (Cooper, 1990). The wildebeest migration into the MMNR from the Serengeti National Park has been shown to be influenced by the veget ation conditions in the Reserve (Ottichilo, 2000). Here we found that habitat use by spotted hyenas differed a great deal between parts of the Reserve managed by TMC and NCG. Clearly the anthropogenic disturbance and intensive grazing on the NCG side of th e Reserve are having strong effects, both on the types of habitat that are available for use by the hyenas and on their choice of habitat types in which to spend their time. Grassland habitat use Most livestock (especially cattle) prefer grassland habitat s in the Reserve and this may increase conflict with hyenas in NCG. Even though grassland was the most abundant habitat in the TW clan territory covering 48.82% (Table 3.2), our results showed hyenas used significantly less of this habitat than was availab le. This result was consistent with that of Boydston et al., (2003b), who showed female hyenas in the TW clan avoiding the prey rich central short grass plains due to intensive use by livestock. Our results were also consistent with those of Kolowski and H olekamp (2009) who found TW hyenas avoiding open grasslands more than was expected based on availability. 72 Shrub land habitat use Shrub land was the third most abundant habitat type available in NCG (21.34%), but hyenas there used this habitat more than ex pected. We suggest this pattern may result from livestock grazing and potential harassment by herders. This condition is therefore forcing hyenas to spend much of their time in the limited available shrub land habitats on this side of the Reserve. Our resu lts were consistent with those of Kolowski and Holekamp, (2009) who found TW hyenas having a greater preference for shrub land vegetation than was expected based on availability . V egetation class was also found to be one of the most important variables det ermining space use by spotted hyenas during the daily livestock grazing period in this area (Kolowski and Holekamp, 2009). Shrub land use in TMC showed no difference from expected values, so we suggest the strict management plan observed in the TMC side o f the reserve is favoring the welfare of hyenas there. Bare ground habitat use Although there was far more bare ground available to hyenas in NCG (28.11%) than TMC (5.36%), NCG hyenas used significantly less of the available bare ground habitat than expe cted there. We suggest that bare ground areas in NCG are just not particularly useful to hyenas because the livestock that graze daily inside the Reserve have ruined the vegetation in those areas so no natural prey for the hyenas occur there. Habitat loss has been shown to lead to the decline of endangered Africa wild dogs ( Lycaon pictus ) (Gorman et al., 1998). Therefore, the high percentage of bare ground we found in this side of the Reserve was of grave concern to us. 73 Riparian habitat use Riparian habit at was the least represented in NCG (1.72%), yet we see hyenas there using significantly more of this habitat than expected. Our results suggest that, as with shrub land habitat, hyenas in NCG may prefer riparian habitats, which are characterized by dense vegetative cover, to avoid harassment by humans attending their cattle. Habitat use outside defended territories As hyenas in NCG have been shown to avoid areas where grazing occurs within the clan territories (see Boydston et al., 2003b), we expected to see significant differences between NCG and TMC with respect to the proportion of hyena locations found outside de fended territories. Anthropogenic activities have been shown to have direct effects on wildlife (See Sinclair et al., 2007). Thus, our observation was expected that hyenas in NCG were using habitats outside their defended territories significantly more tha n were TMC hyenas. We suggest that the anthropogenic activities occurring in NCG may be driving hyenas outside their territories more than is the case in TMC. Conclusion Concerted conservation efforts are required to avoid species extinctions (Brooks et a l., 2002). H uman population increase has also been shown to be one of the greatest drivers of global environmental change (Tilman et al., 2001). Therefore there is urgent need to balance wildlife protection with the welfare of local people. Such efforts ar e proposed here including r emoval of grazing pressure if sound management practices are to be employed in this important ecosystem. Such practices are known to aid in recovery of overgrazed or over - browsed habitats. Similar studies in other parts of the w orld have shown positive correlations 74 between plant species composition and palatable plant cover recovery following cessation of grazing pressure (Frank et al., 2014). Although balancing the economic development of local people and wildlife conservation r emains a big challenge (Tosun, 2001), we suggest that, if good management policies are implemented. Such management policies would involve strict law enforcement to ensure no livestock grazing in NCG, and raising awareness amongst local communities regardi ng the importance of wildlife conservation. Through such policies and many others, we could ensure that wildlife protection is enhanced while promoting the welfare of local people . Just as other studies have suggested, we need to promote habitat conservati on so as to ensure coexistence of man and wildlife. Tourists pay a considerable amount of money to see wildlife in the MMNR. Therefore, if the patterns of habitat use observed here in NCG by spotted hyenas are similar to those of other large carnivores, th en cattle grazing in the Reserve may be negatively affecting tourism in this ecosystem by forcing the predators into thick vegetative cover, out of sight of tourists. 75 REFERENCES 76 REFERENCES Bailey, H., & Thompson, P. (2006). Quantitative analysis of bottlenose dolphin movement patterns and their relationship with foraging. Journal of Animal Ecology , 75 (2), 456 - 465. Balmford, A., Moore, J. L., Brooks, T., Burgess, N., Hansen, L. A., Williams, P., & Rahbek, C. (2001). Conservation conflicts across Africa. Science , 291 (5513), 2616 - 2619. Bourgeau - Chavez, L., Endres, S., Battaglia, M., Miller, M. E., Banda, E., Laubach, Z., ... & Marcaccio, J. (2015). Development of a Bi - National Great Lakes Coastal Wet land and Land Use Map Using Three - Season PALSAR and Landsat Imagery. Remote Sensing, 7(7), 8655 - 8682. Boydston, E. E., Kapheim, K. M., Szykman, M., & Holekamp, K. E. (2003a). Individual variation in space use by female spotted hyenas. Journal of Mammalogy, 84(3), 1006 - 1018. Boydston, E. E., Kapheim, K. M., Watts, H. E., Szykman, M., & Holekamp, K. E. (2003b). Altered behaviour in spotted hyenas associated with increased human activity. Animal Conservation, 6(3), 207 - 219. Brooks, T. M., Mittermeier, R. A., Mittermeier, C. G., Da Fonseca, G. A., Rylands, A. B., Konstant, W. R., ... & Hilton Taylor, C. (2002). Habitat loss and extinction in the hotspots of biodiversity. Conservation biology , 16 (4), 909 - 923. Cooper, S. M. (1990). The hunting behaviour of spotte d hyaenas (Crocuta crocuta) in a region containing both sedentary and migratory populations of herbivores. African Journal of Ecology , 28 (2), 131 - 141. Cooper, S. M. (1989). Clan sizes of spotted hyaenas in the Savuti Region of the Chobe National Park, Bots wana. Botswana Notes and Records , 21 , 121 - 131. Duncan, P. (1983). Determinants of the use of habitat by horses in a Mediterranean wetland. The Journal of Animal Ecology , 93 - 109. Etzenhouser, M. J., Owens, M. K., Spalinger, D. E., & Murden, S. B. (1998). Fo raging behavior of browsing ruminants in a heterogeneous landscape. Landscape Ecology , 13 (1), 55 - 64. Frank, A. S., Wardle, G. M., Dickman, C. R., & Greenville, A. C. (2014). Habitat - and rainfall - dependent biodiversity responses to cattle removal in an arid woodland - grassland environment. Ecological Applications , 24 (8), 2013 - 2028. GATES, S. (1979). A study of the home ranges of free ranging Exmoor pon ies. Mammal Review , 9 (1), 3 - 18. Gittleman, J. L. (Ed.). (2001). Carnivore conservation (Vol. 5). Cambridge University Press. 77 Gorman, M. L., Mills, M. G., Raath, J. P., & Speakman, J. R. (1998). High hunting costs make African wild dogs vulnerable to kl eptoparasitism by hyaenas. Nature , 391 (6666), 479 - 481. Green, D.S. (2015). Anthropogenic disturbance, ecological change, and wildlife conservation at the edge of the Mara - Serengeti ecosystem. Ph.D. Dissertation. Michigan State University. Hayward, M. W., H ayward, G. J., Druce, D. J., & Kerley, G. I. (2009). Do fences constrain predator movements on an evolutionary scale? Home range, food intake and movement patterns of large predators reintroduced to Addo Elephant National Park, South Africa. Biodiversity a nd Conservation , 18 (4), 887 - 904. Homewood, K., Lambin, E. F., Coast, E., Kariuki, A., Kikula, I., Kivelia, J., ... & Thompson, M. (2001). Long - term changes in Serengeti - Mara wildebeest and land cover: Pastoralism, population, or policies?. Proceedings of t he National Academy of Sciences , 98 (22), 12544 - 12549. Ingram, S. N., & Rogan, E. (2002). Identifying critical areas and habitat preferences of bottlenose dolphins Tursiops truncatus. Marine Ecology Progress Series , 244 (29), 247 - 255. Kolowski, J. M., & Hole kamp, K. E. (2009). Ecological and anthropogenic influences on space use by spotted hyaenas. Journal of zoology, 277(1), 23 - 36. McAlpine, C. A., Etter, A., Fearnside, P. M., Seabrook, L., & Laurance, W. F. (2009). Increasing world consumption of beef as a driver of regional and global change: A call for policy action based on evidence from Queensland (Australia), Colombia and Brazil. Global Environmental Change , 19 (1), 21 - 33. Ndegwa Mundia, C., & Murayama, Y. (2009). Analysis of land use/cover changes and a nimal population dynamics in a wildlife sanctuary in East Africa. Remote Sensing , 1 (4), 952 - 970. Ogutu, J. O., Piepho, H. P., Dublin, H. T., Bhola, N., & Reid, R. S. (2009). Dynamics of Mara Serengeti ungulates in relation to land use changes. Journal of Z oology, 278(1), 1 - 14. Ottichilo, W. K., De Leeuw, J., Skidmore, A. K., Prins, H. H., & Said, M. Y. (2000). Population trends of large non migratory wild herbivores and livestock in the Masai Mara ecosystem, Kenya, between 1977 and 1997. African Journal of Ecology , 38 (3), 202 - 216. Pita, R., Mira, A., Moreira, F., Morgado, R., & Beja, P. (2009). Influence of landscape characteristics on carnivore diversity and abundance in Mediterranean farmland. Agriculture, Ecosystems & Environment , 132 (1), 57 - 65. R Core Te am (2015). R: A language and environment for statistical computing. R Foundationfor Statistical Computing, Vienna, Austria. URL http://www.R - project.org/ . 78 Ripple, W. J., Estes, J. A., Beschta, R. L., Wilmers, C. C. , Ritchie, E. G., Hebblewhite, M., ... & Science , 343 (6167), 1241484. Samuel, M. D., Pierce, D. J., & Garton, E. O. (1985). Identifying areas of concentrated use within the home range. The Journal of Animal Ecology , 711 - 719. Serneels, S., Said, M. Y., & Lambin, E. F. (2001). Land cover changes around a major east African wildlife reserve: the Mara Ecosystem (Kenya). International Journal of Remote Sensing, 22(17), 3397 - 3 420. Schadt, S., Revilla, E., Wiegand, T., Knauer, F., Kaczensky, P., Breitenmoser, U., ... & Trepl, L. (2002). Assessing the suitability of central European landscapes for the reintroduction of Eurasian lynx. Journal of Applied Ecology , 39 (2), 189 - 203. Sinclair, A. R. E., Mduma, S. A., Hopcraft, J. G. C., Fryxell, J. M., Hilborn, R. A. Y., & Thirgood, S. (2007). Long Term Ecosystem Dynamics in the Serengeti: Lessons for Conservation. Conservation Biology , 21 (3), 580 - 590. Tilman, D., Fargione, J., Wolff, B., D'Antonio, C., Dobson, A., Howarth, R., ... & Swackhamer, D. (2001). Forecasting agriculturally driven global environmental change. Science , 292 (5515), 281 - 284. Tosun, C. (2001). Challenges of sustainable tourism development in the developing world: th e case of Turkey. Tourism Management , 22 (3), 289 - 303. Tuomainen, U., & Candolin, U. (2011). Behavioural responses to human induced environmental change. Biological Reviews , 86 (3), 640 - 657. Valeix, M., Loveridge, A. J., Davidson, Z., Madzikanda, H., Fritz, H., & Macdonald, D. W. (2010). How key habitat features influence large terrestrial carnivore movements: waterholes and African lions in a semi - arid savanna of north - western Zimbabwe. Landscape Ecology , 25 (3), 337 - 351. 79 CHAPTER 4 ENERGY COST OF LOCOMOT ION BY SPOTTED HYENAS ( Crocuta crocuta ) IN THE MASAI MARA NATIONAL RESERVE, KENYA INTRODUCTION Animals may travel for several reasons but foraging is one of the most basic activities that mammals undertake (Garland, 1983). Predators, in particular, have t o overcome enormous energetic constraints that strongly affect their ecology and evolution (Van Valkenburgh et al., 2004; Gorman et al., 1998; Carbone et al., 1999; Gittleman, 1985). The energetic cost of transport has been expressed as a function of body mass in various vertebrates (Schmidt - Nielson, 1972). Large carnivores range widely (Woodroffe, 2000; Packer et al., 2013) and their energy requirements while moving are expected to be high compared to those of many other mammals. Large carnivores specializ e on feeding on medium and large - sized vertebrates (Carbone et al., 2007). Garland (1983) found that carnivores have greater daily movement distances than do other mammals, and their costs of movement are expected to be higher as well. Members of the orde r Carnivora move an average of 4.4 times farther than other mammals on a daily basis (Garland, 1983) so their energy expenditure should be relatively high. In nature, many animals move at speeds that climax at their maximum rate of oxygen consumption (VO 2 m ax) (see Kruuk, 1972; Seeherman et al., 1981; Thompson, 1980; Taylor et al., 1981). Because VO 2 max is in most cases 10 times greater than the resting metabolic rate in mammals (Taylor et al., 1981), locomotion can be a costly activity. Taylor et al. (1981) pointed out that it is only in Africa where one can currently find a diversity of large wild animals. One place in Africa where animal diversity is extraordinarily high 80 Kenya. Here we estimated daytime energy use during locomotion by 22 adult female spotted hyenas ( Crocuta crocuta ) in the Mara in 2013 (see methods below). Spotted hyenas may scavenge on carrion (Kruuk, 1972; Cooper, 1990) , which is less energetically costly than hunting live prey, but most of their food (roughly 3/4) is obtained from hunting (Cooper, 1990). Spotted hyenas generally prefer medium to large size prey (Henschel and Skinner, 1990, Kruuk, 1972). This preference for large prey maximizes their energy gain but also requires extra energy to pursue and subdue large prey (Carbone et al., 2007). As previous research has shown, dietary requirements are key determinants of animal energetics (Altmann, 1987). Here we asked whether management regime also affects the energy expenditure by wild spotted hyenas. As d side of the Reserve using more energy during locomotion than hyenas in the undisturbed side of the Reserve during the daytime (10am - 4pm). Methods Study area, and study animals Details of our study area and study animals are as explained Chapters 1 and 2. Estimating the energy cost of locomotion Hyena energy use during locomotion was determined by using GPS collar technology to monitor movement patterns of twenty two adult female spotted hyenas in the MMNR (TMC=14, NCG=8, see tables 2.1 and 2.2). The hye nas were immobilized using tiletamine - zolazepam (6.5 mg/kg; Telazol: W.A Butler Company, Brighton Michigan) administered in a 81 plastic dart via a CO 2 powered rifle (Telinject Inc., Saugus, California), and fitted them with GPS collars (Vectronix Aerospace, Berlin, Germany) (Fig. 1.3). Each GPS collar was fitted with a temperature sensor to record the ambient temperature corresponding to hyena locations. The daily temperatures recorded by the sensors in the GPS collars were averaged each month separately for TMC and NCG (Table 4.1). Using the local Kenyan cell phone network, which is handy for acquiring the locations of the hyenas, movement patterns across space and time were recorded daily. The GPS locations were received at a central point using the local cell phone network, and the n downloaded into computers. Collars were programmed to transmit their locations at 10am, 1 pm, and 4 pm every day. All locations were digitized using ESRI ArcMap 10.2.2 software (Environmental Systems Research Institute, Redlands, CA, USA). Shape files of the downloaded GPS coordinates (locations) were given a Geographic Coordinate System (GCS) of World Geodetic System of 1984 (WGS 84) then projected into Universal Transverse Mercator zone 36 South (WGS 84 UTM 36S). Finally XY coordinates were added to all ow for linear distance calculations in meters (m) which were done using R v. 3.1.3 statistical software (R core team, 2015). Linear distances in meters (m) were calculated between each two consecutive locations. Territory sizes were estimated using fixed k ernel utilization distributions with 95% probabilities for all hyenas with collars in 2013 (Green, 2015). During immobilization, body mass (kg) for each hyena was recorded, and the body mass was used to estimate their energy expenditure (Tables 4.2 and 4. 3) . If a single hyena was darted twice during our study period (2013) or earlier during adulthood, her body mass (kg) as used in the energy calculation was the latest weight recorded (Table 4.3) . Distance travelled in meters 82 (m) and time taken in seconds ( s) between any two locations at the three time - points each day (10am, 1pm, and 4pm) were calculated in R statistical software (R core team, 2015). The cost of locomotion has been shown to be a function of body mass (Taylor et al., 1982), so we calculated e nergy spent in locomotion as a function of body mass (kg) and speed of movement (v o ) expressed in m/s. The distance travelled, time taken, speed of movement, and body mass of each animal were used to determine the energy used by that animal during locomoti on via the formula Vo 2 /Mb = 0.533Mb - 0.316 .v o +0.300 Mb - 0.303 (Taylor et al., 1982). Here, Mb represents body mass, and Vo 2 /Mb is the energy use per kg of body mass expressed in units mlO 2 /s/kg. v o (speed) was expressed in m/s, Mb was in kg, and Vo 2 /Mb was c onverted to Joules (J) by using the energetic equivalent 1mlo 2 =20.1 Joules (J) (Taylor et al. 1982). The energy in Joules (J) was then multiplied by the total time (seconds) and total weight (kg) of each hyena to determine her total energy use during loco motion and thereafter converted into mega joules (MJ). This conversion was done by dividing the calculated energy (J) by 1,000,000. Energy units (MJ) were calculated per day and thereafter averaged per month to get monthly estimates (Table 4.2). Throughout this Chapter (as we also did in Chapter 2), we will be using morning to refer to the time between 10am - 1pm and afternoon to refer to the time between 1pm - 4pm. 83 Month TMC temperature ( ° C) NCG temperature ( ° C) Jan 36.83 32.99 Feb 37.64 32.14 March 36.11 32.43 April 37.55 31.97 May 33.99 31.43 June 33.48 31.1 July 32.99 31.14 Aug 34.03 32.7 Sep 36.07 33.42 Oct 37.12 33.68 Nov 36.32 33.51 Dec 36.93 32.19 Table 4.1 : Table of mean monthly temperatures in TMC and NCG. Temperatures were recorded dai ly by sensors fitted on GPS collars and thereafter averaged into monthly records. 84 Month Average monthly energy cost of locomotion (MJ) in TMC Average monthly energy cost of locomotion (MJ) in NCG Jan 1.157 1.133 Feb 1.162 1.131 March 1.171 1.142 April 1.169 1.108 May 1.174 1.101 June 1.169 1.098 July 1.143 1.111 Aug 1.142 1.121 Sep 1.148 1.113 Oct 1.140 1.117 Nov 1.142 1.115 Dec 1.143 1.104 Table 4.2 : Table of average diurnal monthly energy cost of locomotion by hyenas in TMC and NCG. 85 TMC hyenas NCG hyenas Bart 56.2 Baez 55.2 Clov 63.6 Hel 55.2 Digs 50.3 Hex 60.4 Ema 61.4 Juno 48.6 Hndy 56.2 Mgta 64.8 Java 66.3 Pan 63.4 Mtn 54.6 Roos 53.8 Peep 55.6 Til t 57.8 Shrm 57.6 - - Slin 47.8 - - Taj 68.8 - - Tnsl 55.6 - - Wafl 64.9 - - Zoey 57.6 - - Statistical analyses R v.3.1.3 statistical software (R core team, 2015) was used for all statistical analysis. Linear distances were calculated between any two given sets of coordinates between 10am - 1pm and 1pm - 4pm. We used Generalized Linear Models (GLM) to model the energy u sed by hyenas as a function of ten predictor variables: 1) management regime referred to as 86 rce prey), 4) social rank interaction between management regime and rank (coded as management*rank), 8) interaction between management regime and prey (coded as management*prey), 9) an interaction between management regime and season (coded as management*season) and 10) an interaction between standardized temperature and management r egime (coded as management*tempstd). Subsequently, our model was as follows: Eq1: glm(LMJ~time+rank+management+season+prey+tempstd+prey*management+rank*managem ent+season*management+tempstd*management). Where, LMJ refers to the log transformed (base10) ene rgy cost of locomotion in Mega Joules (MJ). In this model, five of our main factors (rank, management, prey, season and time) had two levels with rank being either high (H) or Low (L) (Table 2.1 and 2.2). Management regime th main factor) was treated as a continuous variable and the standardization was done from the mean of all the temperature values recorded by the 22 hyenas (Table 2.1 and 2.2) in the MMNR in 2013. We also included the interactions between management regime and rank (management*rank), management regime and prey 87 ( management*prey), management regime and season (management*season) and management regime and standardized temperature (management*tempstd) in our model. Speed of movement (v o ) was calculated by dividing the linear distance traveled in meters by time spent doing so in seconds; we thus expressed speed (v o ) in m/s. The body mass (kg) measured during immobilization (Table 4.2) and the calculated speed (m/s) were then used to calculate the energy spent in locomotion using the method above (recall; Vo 2 /Mb = 0.533Mb - 0.316 .v o +0.300 Mb - 0.303 ). We presented our means with standard errors, and differences between groups were considered significant when P < 0.05. Results Reserve management regime and energy use Our results showed that hyenas in the undisturbed side of the Reserve significantly used more energy than did hyenas in the disturbed side (t=2.775, p=0.0055 ) (Table 4.4, Fig. 4.2). This result was shocking to us because our results based on movement data showed hyenas in NCG travelled significantly long er distances than did hyenas in TMC. Time of day and energy use Our results showed hyenas were using significantly less energy in the afternoon (1pm - 4pm) than in the morning (10am - 1pm) (t= - 10.668, p<0.0001) (Table, 4.4, Fig. 4.2). This result was expect ed, as our movement results showed hyenas moving significantly shorter distances in the afternoon than in the morning. Social rank and energy use Although our results showed low - ranking (L) hyenas used slightly more energy than did high - ranking (H) hyena s (Fig.4.3), this difference was not statistically significant (t= 1.298, 88 p=0.1945) (Table 4.4). This result was not surprising, as we found no significant difference in distance moved between low - ranking hyenas and high - ranking hyenas in our movement mode ling results. However, low - ranking hyenas in TMC used significantly more energy than did high - ranking hyenas in TMC (t=8.681, p<0.0001). Surprisingly, such a difference was not observed in NCG. (Table 4.4, Fig. 4.6). Prey availability and energy use Our results showed hyenas were using significantly less energy when prey were abundant than when prey were scarce (t= - 2.121, p=0.0339) (Table 4.4, Fig. 4.4). This result was expected, as predators are expected to need less energy in acquiring food when prey are abundant. However we did not see significant differences between energy use in locomotion by hyenas in TMC and NCG when prey was abundant or scarce (t= - 1.501, p=0.1334), (Table 4.4, Fig. 4.7). Seasonality and energy use Our results showed no signif icant difference in energy use between wet and dry seasons (t= - 0.399, p =0.6897) (Table 4.4, Fig. 4.5). This result was expected, as we also found no significant difference in the distance moved by hyenas between dry and wet seasons. We did not find signi ficant differences between energy use in locomotion when comparing hyenas living in TMC and NCG during the wet and dry seasons (t=0.550, p=0.5827) (Table 4.4, Fig. 4.8). This result was expected because we got similar results with our movement modeling. Standardized temperature and energy use Our results showed hyenas were using significantly more energy in locomotion as the temperature increased (t=2.728, p=0.0064). This result was expected, as increases in 89 temperature have been shown to increase metabolic energy so animals are expected to use more energy as ambient temperatures increase (Knut, 1997). Our results also showed no difference in energy cost of locomotion between the undisturbed side of the Reserv e (TMC) and the disturbed side of the Reserve (NCG) (t= - 0.475, p=0.6347) with increase in temperature. This result was somewhat surprising to us because we expected to see hyenas using significantly higher amounts of energy in the warmer TMC side than the cooler NCG side of the Reserve. Discussion Management regime and energy use We suspect our finding that hyenas in TMC were using significantly more energy in locomotion than NCG hyenas might have been due to differences in their body masses (Fig. 4.9). On average, TMC hyenas weighed more than NCG hyenas but moved shorter distances. The average weight for TMC hyenas was ~60.2kg while that of NCG hyenas was ~56.77kg (Table 4.3). Since our energy calculation was expressed as a function of body mass, the bia s we observed here may be due to body mass differences. . . Social rank and energy use We were somewhat surprised by our finding that the difference in energy use between low - ranking and high - ranking hyenas in TMC was significantly more than was the di fference in NCG (Fig. 4.6). This means that both high - ranking and low - ranking hyenas in NCG are spending similar amounts of energy in locomotion (Fig. 4.6). We fear that the high stress levels that hyenas in this side of the Reserve seem to be facing may h ave negative fitness consequences in the long run. 90 Prey and energy use Our results showed that spotted hyenas used significantly more energy in locomotion when prey were scarce than when they were abundant. This result was consistent with our expectatio n that hyenas should be using more energy while looking for prey when prey density is low in the Reserve. However, we did not see the same patterns of energy use in relation to prey availability in TMC and NCG. This result was not consistent with our expe ctation that hyenas in NCG should be using more energy during locomotion due to extra movement resulting from livestock grazing within their territories. We are concerned that the decreasing ungulate population trends in the MMNR (Ottichilo et al., 2000, S erneels and Lambin, 2001) and its environs may continue to impose further energetic costs on hyenas. Season and energy use We did not find significant differences in the energy costs of locomotion to hyenas between dry and wet seasons. This was not cons istent with our expectation that hyenas should be using less energy in locomotion during the dry period. This result may have been due to the fact that it is during the dry season that most prey concentrate in the greener pastures near water courses in the Reserve as opposed to during the wet season when they may even move out of the Reserve onto the neighboring group ranches. We also found no significant differences in energy use between hyenas in TMC and those in NCG in locomotion during the dry and wet s easons. Other factors like distribution of the prey may play a part here since we did not look at prey distribution in this research. Other prey also develop defensive mechanisms when they are in groups, so hyenas may be forced to move around the Reserve l ooking for easy prey to catch despite the fact that lots of prey may be present during the dry period. 91 Standardized temperature and energy use Our results were consistent with earlier findings that energy use increases with ambient temperature (Knut, 1997). However, our finding that there was no difference in energy use in locomotion between hyenas in TMC and NCG hyenas was unexpected because we saw hyenas traveling significantly shorter distances in TMC than they did in NCG. Conclusion Our results we re surprising in light of the greater anthropogenic disturbance in NCG than TMC. Based on this, we had expected to see more costly locomotion in NCG than TMC, but we observed the opposite. We suggest this pattern might be due either to differences in body mass between hyenas in TMC and NCG. Our results also showed that the differences we observed in energy cost of locomotion was due to differences in management regimes but not differences in temperatures. 92 Estimate Std. Error t value Pr(>|t|) (Intercept) 1.132751 0.005040 224.740 <2e - 16*** timepm - 0.0104763 0.0009821 - 10.668 <2e - 16*** rankL 0.0019274 0.0014853 1.298 0.19447 managementtmc 0.0056330 0.0020298 2.775 0.00553** seasonwet - 0.0007992 0.0020013 - 0.399 0.68967 preymanyp - 0.0044752 0.0021097 - 2.121 0.03394* tempstd 0.0030949 0.0011346 2.728 0.00639** managementtmc:preymany - 0.0036071 0.0024031 - 1.501 0.13341 rankL: managementtmc 0.0158660 0.0018277 8.681 <2e - 16*** managementtmc:seasonwet 0.0012487 0.0022724 0.550 0.58268 managementtmc:tempstd - 0.0005673 0.0011939 - 0.475 0.63468 Table 4.4: GLM table of results for energy cost of locomotion. Energy values have been transformed to log base 10. Bolded cells represent significant effects. 93 Fig ure 4.1: Management regime and average daily energy cost of locomotion in TMC and NCG. Asterisk represents significant difference, and error bars represent standard errors. 1 1.05 1.1 1.15 1.2 1.25 TMC NCG Energy (MJ) Management * 94 Figure 4.2: Time of day and average daily energy cost of locomotion in the M MNR. Asterisk represents significant difference, and error bars represent standard errors. Figure 4.3: S ocial rank and average daily energy cost of locomotion in the MMNR. Error bars represent standard errors 1.05 1.1 1.15 1.2 1.25 10am-1pm 1pm-4pm Energy (MJ) Time of day 1.05 1.1 1.15 1.2 1.25 High-ranking Low-ranking Energy (MJ) Rank * 95 Figure 4.4: Prey availability and average daily energy cost of locomotion in the MMNR. Asterisk represent significant difference, and error bars represent standard errors. F igu re 4.5: Seasonality and average daily energy cost of locomotion in the MMNR. Error bars represent standard errors. 1.05 1.1 1.15 1.2 1.25 Abundant Scarce Energy (MJ) 1.05 1.1 1.15 1.2 1.25 Wet Dry Energy (MJ) 96 Figure 4.6: Social rank and average daily energy cost of locomotion in TMC and NCG. Asterisk represents significant difference, and error bars represent standard errors. Figu re 4.7: Prey availability and average daily energy cost of locomotion in TMC and NCG. Error bars represent standard errors. 1 1.05 1.1 1.15 1.2 1.25 TMC NCG Energy (MJ) High-ranking Low-ranking 1 1.05 1.1 1.15 1.2 1.25 TMC NCG Energy (MJ) Abundant Scarce * 97 Figure 4.8: Seasonal ity and average energy cost of locomotion in TMC and NCG. Error bars represent standard errors. Fig ure 4.9 : The relationship between hyena body mass (kg) and energy cost of locomotion (MJ). 1 1.05 1.1 1.15 1.2 1.25 TMC NCG Energy (MJ) Wet Dry 0.95 1 1.05 1.1 1.15 1.2 1.25 1.3 45 50 55 60 65 70 Energy (MJ) Body mass (kg) 98 GENERAL CONCLUSIONS FROM THIS THESIS This study has shown that during the daytime, hyenas were moving significantly longer distances in the morning (10am - 1pm) than they did in the afternoon (1pm - 4pm). We have also seen that hyenas inhabiting the disturbed side of the Reserve (NCG) were moving significantly longer distances than were hyenas in the undisturbed side of the reserve (TMC). We also saw movement patterns. In addition we found that hyenas in the undi sturbed side of the Reserve were travelling significantly shorter distances than they did in the disturbed side of the Reserve as temperatures increased. Hyenas in the disturbed side of the Reserve may be forced to move about more as temperatures increase while trying to look for dense vegetation which is scarcely available. Through this study, we have shown that the anthropogenic activities that take place in NCG may be causing serious behavioral change in hyenas during the daytime, and we believe these e ffects should be addressed very soon by Reserve managers. In our study of habitat use we found that hyenas in TMC significantly preferred open vegetation (grassland and bare ground) whereas hyenas in NCG preferred habitat characterized by dense vegetativ e cover (shrub land and riparian). Our habitat results also showed that hyenas in NCG were found significantly outside their defended territory than expected. We suggest that the livestock grazing that takes place in this side of the Reserve may be forcing the hyenas here to seek alternative resources outside their defended territory. Finally, in Chapter 4 we obtained the surprising result that TMC hyenas spend more energy on locomotion than do NCG hyenas despite that fact that NCG hyenas travelled sign ificantly longer distances. We found here that hyenas experienced significantly higher 99 energetic costs in the morning than they did in the afternoon. Our results also showed that seasonal variation in rainfall (dry or wet) does not lead to significant diff erences in the cost of significantly fewer benefits in NCG than is the case in TMC. We found that movement was less costly energetically to hyenas when prey were abund ant than when prey were scarce. Finally, our results showed that differences in temperatures between TMC and NCG do not lead to significant differences in energy cost of locomotion. . We conclude by suggesting that addressing the anthropogenic activities i n NCG in a better way will not only benefit he hyenas but also other animals in this key ecosystem for biodiversity conservation. 100 REFERENCES 101 REFERENCES Altmann, S. A. (1987). The impact of locomotor energetics on mammalian foraging. Journal of Zoology , 211 (2), 215 - 225. Carbone, C., Mace, G. M., Roberts, S. C., & Macdonald, D. W. (1999). Energetic constraints on the diet of terrestrial carnivores. Nature , 402 (6759), 286 - 288. Carbone, C., Teacher, A., & Rowcliffe, J. M. (2007). The costs of carnivory. PLoS biology , 5 (2), e22. Cooper, S. M. (1990). The hunting behaviour of spotted hyaenas (Crocuta crocuta) in a region containing both sedentary and migrato ry populations of herbivores. African Journal of Ecology , 28 (2), 131 - 141. Garland Jr, T. (1983). Scaling the ecological cost of transport to body mass in terrestrial mammals. American Naturalist , 571 - 587. Gittleman, J. L. (1985). Carnivore body size: eco logical and taxonomic correlates. Oecologia , 67 (4), 540 - 554. Gorman, M. L., Mills, M. G., Raath, J. P., & Speakman, J. R. (1998). High hunting costs make African wild dogs vulnerable to kleptoparasitism by hyaenas. Nature , 391 (6666), 479 - 481. Green, D.S. (2015). Anthropogenic disturbance, ecological change, and wildlife conservation at the edge of the Mara - Serengeti ecosystem. Ph.D. Dissertation. Michigan State University. Henschel, J. R., & Skinner, J. D. (1990). The diet of the spotted hyaenas Crocuta crocuta in Kruger National Park. African Journal of Ecology , 28 (1), 69 - 82. Knut Schmidt - Nielsen. (1997). Animal physiology: adaptation and environment . Cambridge University Press. Kruuk, H. (1972). The spotted hyena: a study of predation and social behav ior. Ottichilo, W. K. (2000, March). Wildlife dynamics: an analysis of change in the Masai Mara ecosystem of Kenya. ITC. Packer, C., Loveridge, A., Canney, S., Caro, T., Garnett, S. T., Pfeifer, M., ... & Plumptre, A. (2013). Conserving large carnivores: dollars and fence. Ecology letters , 16 (5), 635 - 641. R Core Team (2015). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R - project.org/ . 102 Schmidt - Nielsen, K. (1972). Locomotion: energy cost of swimming, flying, and running. Science , 177 (4045), 222 - 228. Seeherman, H. J., Taylor, C. R., Maloiy, G. M., & Armstrong, R. B. (1981). Design o f the mammalian respiratory system. II. Measuring maximum aerobic capacity. Respiration physiology , 44 (1), 11 - 23. the northern part of the Serengeti Mara ecosy stem. Journal of Biogeography, 28(3), 391 - 407. Taylor, C. R., Heglund, N. C., & Maloiy, G. M. (1982). Energetics and mechanics of terrestrial locomotion. I. Metabolic energy consumption as a function of speed and body size in birds and mammals. Journal of Experimental Biology , 97 (1), 1 - 21. Taylor, C. R., Maloiy, G. M., Weibel, E. R., Langman, V. A., Kamau, J. M., Seeherman, H. J., & Heglund, N. C. (1981). Design of the mammalian respiratory system. III. Scaling maximum aerobic capacity to body mass: wi ld and domestic mammals. Respiration physiology , 44 (1), 25 - 37. Thompson, S. D., MacMillen, R. E., Burke, E. M., & Taylor, C. R. (1980). The energetic cost of bipedal hopping in small mammals. Tilson, R., Blottniz, F. V., & Henschel, J. (1980). Prey selec tion by spotted hyaena (Crocuta crocuta) in the Namib Desert. Madoqua , 12 , 41 - 49. Van Valkenburgh, B., Wang, X., & Damuth, J. (2004). Cope's rule, hypercarnivory, and extinction in North American canids. Science , 306 (5693), 101 - 104. Woodroffe, R. (2000). Predators and people: using human densities to interpret declines of large carnivores. Animal conservation , 3 (02), 165 - 173