ASSESSING THE ABUNDANCE AND NESTING BEHAVIOUR OF GOLDEN-
RUMPED ELEPHANT-SHREWS (GRES) (Rhynchocyon chrysopygus) IN
ARABUKO-SOKOKE FOREST, KILIFI COUNTY, KENYA
ONDORO RAEL NELLY NYANCHAMA
A thesis submitted in partial fulfillment of the requirements for the Degree of Master
of Science in Environmental Science of Pwani University.
May, 2020 ii
DECLARATION
This thesis is my original work and has not been presented in any other University or any other Award.
Ondoro Rael N. Nyanchama
MG20/PU/36233/17
Signature: Date: 08/03/2021
We confirm that the work reported in this thesis was carried out by the candidate under our supervision.
1. Dr. Benards Okeyo
Department of Environmental Science
Pwani University
Signature: Date: 08/03/2021
2. Dr. Colin Jackson,
National Director
A Rocha Kenya, Watamu, Kenya
Signature: ______Date: 08/03/2021
iii
DEDICATION
“Conservation is a cause that has no end. There is no point at which we will say our work is finished.”
Rachel Carson.
iv
ACKNOWLEDGEMENTS
First, I would like to thank the Almighty God for the strength and courage that He bestowed upon me making the completion of this work possible. Many thanks to Nature
Kenya and Mohammed Bin Zayed Funding for Nature for their financial support. My sincere gratitude to my supervisors Dr. Bernards Okeyo and Dr. Colin Jackson for tirelessly providing professional guidance throughout this period.
Special thanks to the Community Forest Association, Forest guards, and research assistants led by Mr. Willy Kombe. I extend my gratitude to Mr. Francis Kagema of
Nature Kenya for providing support in many ways especially on the fieldwork. I am also grateful to my wonderful family for their assistance, encouragement, and support throughout the study. Since I may not be able to mention each person individually, I am sincerely grateful to each person who in one way or another contributed to the success of this thesis.
v
ABSTRACT
The Golden-rumped Elephant-shrew (GRES) (Rhynchocyon chrysopygus) is an insectivorous mammal endemic to Arabuko-Sokoke Forest (ASF) and environs. It is listed as Endangered by the International Union for Conservation of Nature (IUCN) Red
List. Previous population monitoring studies showed that the GRES population has been decreasing from ASF with the decline attributed to a change in the forest quality resulting from demand for forest products. However, these studies were conducted over a decade ago. This study aimed to assess the abundance, and influence of deadwood volume on nest sightings across the Cynometra, Brachystegia and Mixed Forest vegetation types in
ASF. Line transects of 100m length were used to collect data on nesting behaviour while one 10 by 10m quadrat at the centre of each transect was used to collect deadwood data.
Distance sampling method, utilizing detection probability data was used to determine the abundance in each of the vegetation types. Linear regression models were employed to examine the association between the number of nests per transect and deadwood volume while means and proportions were calculated to examine the determinants of GRES nesting sites. A total of 44 transects were laid across the three main vegetation types.
Findings from this study indicate that the GRES population was estimated to be 19,423.
Besides, deadwood volume was not associated with the number of nest sightings. Nests were found in areas with higher canopy cover, litter depth, and vegetation density.
Although findings from this study indicate an increase in the GRES population in ASF, there is a need to continue monitoring their numbers to inform guidelines and conservation measures. There should be continued support to the community to improve their livelihoods to reduce the pressure for forest products and degradation which threatens flora and fauna including GRES with extinction.
vi
TABLE OF CONTENTS
DECLARATION ...... ii
DEDICATION ...... iii
ACKNOWLEDGEMENTS ...... iv
ABSTRACT ...... v
LIST OF FIGURES ...... x
LIST OF TABLES ...... xii
ABBREVIATIONS AND ACRONYMS ...... xiii
CHAPTER 1: INTRODUCTION ...... 1
1.1 Background Information ...... 1
1.2 The Problem Statement ...... 5
1.3 Objectives of the Study ...... 6
1.3.1 General Objective ...... 6
1.3.2 Specific Objectives ...... 6
1.4 Research questions ...... 6
1.5 Significance of the Study ...... 6
1.6 Ethical Considerations in the Study ...... 8
1.7 Limitations of the Study ...... 8
CHAPTER 2: LITERATURE REVIEW ...... 9
2.1 A general overview of small mammals ...... 9
2.2 Distribution of Small Mammals ...... 9
2.3 Elephant Shrews ...... 12
2.3.1 Description of Elephant Shrews ...... 13
2.3.1.1 Golden-rumped Elephant-Shrew ...... 14
vii
2.3.1.2 Other Giant Sengis ...... 15
2.3.1.3 Other types of elephant shrews ...... 15
2.3.2 Distribution of Elephant Shrews ...... 16
2.3.3 Ecology and Behaviours of Elephant-Shrews ...... 17
2.3.3.1 Habitats...... 17
2.3.3.2 Diet and Foraging ...... 18
2.3.3.3 Activity Patterns ...... 19
2.3.3.4 Mating and Courtship ...... 20
2.3.3.5 Reproduction and Parenting ...... 21
2.3.4 Ecological Importance of Elephant Shrews ...... 21
2.3.5 Effect of Habitat Changes on the Ecology of Small Mammals and Elephant-
Shrews………………… ...... 22
2.3.6 Conservation of Small Mammals and Elephant Shrews ...... 24
2.3.7 Socio-Economic Benefits of Small Mammals ...... 26
2.3.7.1 Creation of Jobs ...... 26
2.3.7.2 Creation and Support of local business ...... 26
2.3.7.3 Boost to Tourism ...... 26
2.3.7.4 Development of Infrastructure ...... 27
2.3.7.5 Empowerment of the Local Communities ...... 27
2.3.7.6 Ecosystem Services ...... 27
2.3.7.7 Option Value...... 28
2.3.7.8 Existence Value ...... 28
2.4 Arabuko-Sokoke Forest...... 28
2.4.1 Flora in Arabuko-Sokoke Forest ...... 28
viii
2.4.2 Fauna in Arabuko-Sokoke Forest ...... 29
2.4.3 Status of Arabuko Sokoke Forest ...... 30
2.4.4 Local, National, Regional and Global Value of Arabuko Sokoke Forest ...... 31
2.4.5 Threats to Arabuko Sokoke Forest ...... 31
2.5 Golden-rumped Elephant-shrews in ASF ...... 34
2.6 Status of GRES in ASF ...... 34
2.7 Nesting of GRES ...... 35
2.8 Threats to GRES in ASF ...... 37
2.9 Conservation Status of GRES ...... 38
2.10 Conceptual Framework ...... 38
CHAPTER 3: MATERIALS AND METHODS ...... 40
3.1 Study Area ...... 40
3.1.1 Climatic Conditions ...... 41
3.2 Study Design ...... 43
3.3 Sampling Procedure ...... 43
3.4 Data Collection Procedures and Techniques ...... 44
3.5 Data Analysis Techniques ...... 48
3.5.1 Abundance Estimation ...... 48
2.10.5 Deadwood Calculation ...... 50
2.10.6 Determinants of Nest Location and nest composition ...... 50
CHAPTER 4: RESULTS ...... 52
4.1 Survey Effort ...... 52
4.2 Perpendicular distances ...... 52
4.3 Detection Probability ...... 55
ix
4.4 Effect of covariates on Detection Probability ...... 56
4.5 Density and Abundance Estimate ...... 56
4.6 Deadwood Biomass and Abundance ...... 57
4.7 Determinants of Nesting Site and nest composition ...... 58
4.7.1 Effect of Canopy cover on nesting site ...... 58
4.7.2 Effect of Litter Depth on nesting site ...... 59
4.7.3 Effect of Vegetation Density on nesting site ...... 59
4.7.4 Effect of Canopy Height on nesting site ...... 59
4.7.5 Material Composition of nests ...... 59
CHAPTER 5: DISCUSSION ...... 61
5.1 Introduction ...... 61
5.2 Abundance and Trends of GRES in Arabuko-Sokoke Forest ...... 61
5.3 Deadwood Biomass and Nesting Sites ...... 63
5.4 Factors Influencing Nesting Site ...... 64
CHAPTER 6: CONCLUSION AND RECOMMENDATIONS ...... 66
6.1 Conclusion ...... 66
6.2 Recommendation ...... 66
REFERENCES ...... 68
APPENDICES ...... 82
Appendix 1: Data collection tool ...... 82
Appendix 2: Effect of Covariates on detection probability by vegetation type ...... 85
x
LIST OF FIGURES
Figure 1: Golden-rumped Elephant-shrew...... 14
Figure 2: Conceptual Framework ...... 39
Figure 3: Map of Arabuko-Sokoke Forest ...... 40
Figure 4: Transects sampled in Arabuko Sokoke Forest for this study ...... 48
Figure 5: Distribution of nest sighting distance from transect by vegetation type in
Arabuko-Sokoke Forest ...... 53
Figure 6: Histogram of the estimated MCDS detection function for GRES averaged over the observed covariate values for litter depth, canopy cover, canopy height, and vegetation density in the Cynometra vegetation type...... 54
Figure 7: Histogram of the estimated MCDS detection function for GRES averaged over the observed covariate values for litter depth, canopy cover, canopy height and vegetation density in the Mixed Forest vegetation type ...... 54
Figure 8: Histogram of the estimated MCDS detection function for GRES averaged over the observed covariate values for litter depth, canopy cover, canopy height, and vegetation density in the Brachystegia vegetation type...... 55
Figure 9: Effect of Litter depth on the probability of detection in Cynometra forest .....85
Figure 10: Effect of Litter depth on the probability of detection in the Mixed Forest vegetation Type ...... 86
Figure 11: Effect of litter depth on the probability of detection in the Brachystegia Forest
...... 87
Figure 12: Effect of Canopy Cover on the probability of detection in Cynometra
Forest...... 88
Figure 13: Effect of Canopy Cover on the probability of detection in Mixed forest ...... 89
xi
Figure 14: Effect of Canopy Cover on the probability of detection in Brachystegia forest
...... 90
Figure 15: Effect of Canopy Height on the probability of detection in Cynometra forest ...... 91
Figure 16: Effect of Canopy Height on the probability of detection in Mixed Forest ....92
Figure 17: Effect of Canopy Height on the probability of detection in Brachystegia Forest
...... 93
Figure 18: Effect of Vegetation Density on the probability of detection in Cynometra
Forest...... 94
Figure 19: Effect of Vegetation Density on the probability of detection in Mixed
Forest...... 95
Figure 20: Effect of Vegetation Density on the probability of detection in Brachystegia
Forest...... 96
xii
LIST OF TABLES
Table 1: Distribution of transects by vegetation type ...... 52
Table 2: Estimates of f(0), corresponding 95% confidence intervals (CI), and percentage coefficient of variation (% CV) for the three vegetation types...... 55
Table 3: Effect of covariates on the probability of detection ...... 56
Table 4: Density (D), corresponding 95% confidence interval (CI) and Abundance (N) and its corresponding 95% confidence interval (CI) ...... 57
Table 5: Ordinary Least Square (OLS) Regression between deadwood volume (M3) and the number of nest sightings by vegetation type and overall, across the three vegetation types...... 58
Table 6: Distribution nests’ material composition by vegetation type in Arabuko-Sokoke
Forest...... 60
Table 7: Comparisons of GRES Abundance between 2019 and 2008 ...... 61
xiii
ABBREVIATIONS AND ACRONYMS
ASF: Arabuko-Sokoke Forest
FTES: Four-toed elephant-shrew
GPS: Global Positioning System
GRES: Golden-rumped elephant-shrew
KWS: Kenya Wildlife Service
MCDS Multiple Covariate Distance Sampling
SDGs: Sustainable Development Goals
1
CHAPTER 1: INTRODUCTION
1.1 Background Information
Small mammals can be described as animals that weigh 500g or less up to 1 kg when an adult. Besides, they are the largest Order in class Mammalia, are terrestrial and arboreal in nature (Hoffmann et al., 2010). Small mammals are often hard to see or find mainly due to their characteristics and behaviours, such as nocturnal activity patterns, camouflage, and relative shyness (Cook, 2001). However, like other small animals, they are found in almost every part of the world, where they occupy different types of habitats and vegetation. A global database on the composition of non-volant small mammals developed by (Luza et al., 2019) showed that these animals are found in natural, neglected natural, and human-modified habitats.
Furthermore, Ceballos and Ehrlich noted that certain regions throughout the world had a high diversity of species, with the Amazonian lowlands and Andes of northern and western South America, East Africa, and Southeast Asia reported to have the highest concentrations of species diversity (Ceballos and Ehrlich, 2006). In Eastern Africa, there is scanty information that is available on small mammal faunal diversity, distribution and natural history (Oguge et al., 2004). Nevertheless, there have been studies on small mammals conducted on selected areas such as Mt Kilimanjaro which is Africa’s highest mountain, Mt. Kenya and the montane forests of southeast Kenya (Stanley et al., 2014,
Musila et al., 2019, Oguge et al., 2004). In the mountainous areas of Kilimanjaro and Mt.
Kenya greatest diversity was observed at certain points above sea level.
Elephant-shrews, also referred to as jumping shrews or sengis, are small insectivorous mammals native to Africa (Hanks, 1968). For a very long time, their history was not clearly understood. However, recent studies on evolutionary relationships have shown
2 that sengis are not closely related to the shrews at all but belong to a distinctive group of
African mammals, Afrotheria, which include other animals such as elephants, sea cows, hyraxes, aardvark, and tenrecs (Foundation, 2020, Rathbun, 2020)
They are mouse-like in appearance, have an elongated nose (proboscis) which resembles the elephant trunk and there is an assumption of a close relationship with the true shrews hence their name (Romer, 1972). They are diurnal and feed only on insects in deep leaf litter thus their preference to live in forests. More specifically, closed-canopy woodlands and thickets with a floor that is densely covered by leaf litter provide the ideal habitat for elephant-shrews (Rovero et al., 2013). There are 15 species of elephant-shrews in the order Macroscelidea that are known and are grouped into four genera; Rhynchocyon,
Elephantulus, Macroscelides, and Petrodromus (Skinner, 2005). The Golden rumped elephant-shrew, the species of interest for this study is in the order Rhynchocyon.
Two types of elephant-shrews have been reported in Arabuko-Sokoke Forest (ASF): the
Four-toed Elephant-shrew (Petrodromus tetradactylus) which is widely distributed beyond the forest, and the Golden-rumped elephant-shrew (GRES) (Rhynchocyon chrysopygus), which is limited in distribution and is endemic to Arabuko-Sokoke Forest
(Nicoll and Rathbun, 1990). Existing evidence shows that 90% of the known global population of the GRES is found in ASF (Wekesa, 2017). GRES feeds on small invertebrates such as earthworms, beetles, termites, and ants which are likely found in deadwood (FitzGibbon, 1995). Locating food and cover resources within a close distance is an important aspect for most animals including GRES. Forest fragmentation resulting from logging and fuelwood harvesting contributes to disturbance of the forest ecosystem, hence threatening species existence especially those that are dependent on relatively undisturbed forest areas. The GRES population has been decreasing across all vegetation types in ASF. Analyses from previous studies conducted in 1993, 2000 and 2007 showed
3 a shift in trend in the number of GRES across the three vegetation types as follows; 1)
Mixed Forest 3,900 in 1993, 4,264 in 2000 and 2,080 in 2007, 2) Brachystegia Forest
1,541 in 1993, 1,864 in 2000 and 1,474 in 2007, and 3) Cynometra Forest 16,456 in 1993,
8,228 in 2000 and 9,196 in 2007 (Ngaruiya, 2009b). There was an overall decrease of about 9% in 2007 from the previous survey done in 2000 (Bauer, 2001). This decline in
GRES abundance was attributed to the change in the forest quality resulting from forest degradation (Ngaruiya, 2009b). Additionally, GRES moved to unlikely forest parts like
Brachystegia woodland where there are fewer disturbances (Ngaruiya, 2009b). This species has the most restricted range of all elephant-shrews and is the only one considered endangered (Ngaruiya, 2009b). Inadequate information on the GRES makes it difficult to determine its current conservation status.
ASF is the largest of the intact coastal forests, and has been ranked as the second most important forest for conservation of bird species in addition to the endangered mammal species (Gereau et al., 2014). According to Chira et al., ASF is characterized by distinct vegetation types due to different soil types and rainfall distribution (Chira, 1993). The three main vegetation types in ASF namely are; Mixed Forest, Brachystegia Forest and
Cynometra Forest (Arabuko-Sokoke Forest Management Team, 2002). Mixed Forest is a dense forest type with diverse tree flora, Brachystegia Forest is a more open while
Cynometra Forest is a dense forest or thicket. This forest has been recognized as an international biodiversity hotspot forest due to the high number of endemic and near- endemic species (Chira, 1993).
Globally, among other factors, habitat loss, fragmentation, and overexploitation of resources have been identified as the major threats to species (Groom, 2006). Human impacts to a large extent have led to the destruction of wild habitats, and therefore responsible for the loss of genetic diversity within and among species and in extreme
4 cases species extinction (Lusweti, 2011). Deforestation is the most destructive of human activities which results in fragmentation and poses a greater risk to species that are restricted in their geographic distribution and are not able to move to other forested areas within habitats that are similar to their own (Groom, 2006). Firewood collection, although a “formal exploitation” through Community Forest Association (CFA), is gradually degrading the forest when conducted to unsustainable levels for use in households and commercial purposes as a source of income. A study conducted focusing on the utilization and governance of Arabuko-Sokoke Forest listed firewood collection as the dominant forest-based economic activity estimated at 40% (Ndalilo et al., 2017).
The extraction of timber in the forest dates back to the 1920s which proceeded with minimal planned utilization leading to depleted stocks to support the sawmills.
Furthermore, hunting of bushmeat and exploitation for wood carvings were also practised. Consequently, forest type classification in ASF has been changing over time as a result of the alterations that the forest has been experiencing due to disturbances
(Habel et al., 2017). It was found that factors determining the location of disturbance in the forest include accessibility, vegetation characteristics, and the type of forest. It was noted that people targeted both small and big trees of Cynometra and those that are deeper in the forest, but extraction of poles is mainly from the mixed forest which is near the edge (Waters et al., 2007). Dry fuelwood harvesting in ASF has been reported to reduce invertebrate abundance such as termites hence loss of nest sites for hole- or ground- nesting birds (ASFMT, 2002). ASF is a forest reserve that is managed by the Kenya
Forest Service whose level of protection is somehow weak due to inadequate capacity to patrol and ensure its protection. A slight improvement in relation to pressure as a result of reduced threat was reported in ASF (Ndang'ang'a et al., 2016).
5
Against this backdrop, this study aimed to examine the abundance and nesting behaviour of GRES to 1) provide the most recent evidence on their numbers for appropriate classification by international organisations such as the IUCN, 2) to highlight the likely positive effects of forest conservation initiatives implemented this far, 3) provide a resource for continued monitoring of GRES population status in future studies, and 4) provide locally-relevant evidence for developing and enhancing sound and sustainable forest conservation strategies that can foster the conservation of not only GRES but also other species within ASF and other similar settings.
1.2 The Problem Statement
While Arabuko-Sokoke Forest is recognized internationally as an ecosystem with one of the highest biodiversity richness in Africa, this has largely not protected it from the negative impacts resulting from increasing adjacent human population and the subsequent rise in demand for forest products (Arabuko-Sokoke Forest Management
Team, 2002, Oyugi et al., 2008, Habel et al., 2017). While estimates of GRES population have been carried out, the last one was conducted 13 years ago meaning the forest management has no current data on GRES status in ASF. Additionally, no substantive study has been conducted to identify the determinants of GRES nesting sites. As elephant-shrews depend on nests as a shelter and for survival, it is essential to adequately understand the factors affecting nest construction. One factor, in particular, deadwood availability and its effect on GRES abundance has not been studied. The aim of the study, therefore, was to fill these gaps by examining the abundance of GRES across the three main vegetation types in ASF, establish the distribution of nesting sites, and characterize the nesting materials. In so doing, this would contribute data required to inform GRES conservation interventions and future GRES monitoring studies.
6
1.3 Objectives of the Study
1.3.1 General Objective
This research aimed to assess the abundance and nesting behaviour of Golden-rumped elephant-shrews (Rhynchocyon chrysopygus) in Arabuko-Sokoke Forest, Kilifi County.
1.3.2 Specific Objectives
i. To estimate the current abundance of Golden-rumped elephant-shrews in
Arabuko-Sokoke Forest both overall and by main vegetation type.
ii. To establish the relationship between the number of GRES nest sightings and
deadwood biomass. iii. To assess the determinants of Golden-rumped elephant-shrew nesting sites and
material composition of nests
1.4 Research questions
1. What is the abundance of Golden-rumped elephant-shrew in Arabuko-Sokoke
Forest during the study period?
2. Is there a relationship between the number of GRES nest sightings and deadwood
biomass in Arabuko-Sokoke Forest?
3. What are the determinants of Golden-rumped elephant-shrew nesting sites?
1.5 Significance of the Study
The United Nations Sustainable Development Goals (SDGs) identifies species conservation as one of the main goals – goal 15 (United Nations, 2017). The IUCN has
7 listed GRES as one of the species threatened with extinction. To tailor conservation efforts, it is paramount to continuously monitor biodiversity status and their distribution especially in areas such as ASF where over 90% of the known global GRES population resides. Specifically, the findings of this study sought to determine the current population status of GRES which will be critical in establishing the conservation status of the species. The findings will also be essential in determining whether there are new adaptations in terms of preference to the different vegetation types (Mixed Forest,
Brachystegia Woodland, and Cynometra Forest) in Arabuko-Sokoke Forest. Results on the nesting behaviour of GRES will provide insights on whether there is preference regarding nesting in areas with more deadwood and describe determinants to GRES nesting site, hence assisting in designing and implementing effective environmental enrichment programmes that may be adopted to conserve this endangered species.
Besides, since the study focuses on the conservation of an endemic species, there are numerous benefits for the community. First, the community will benefit from researchers and tourists that will visit the area interested with the species hence generating income directly or indirectly. For instance, evidence from a study in India indicated the huge income that communities can generate from tourism activities (Pillai, 2011). Second, community members, especially, the youth can learn from the researchers who come to study GRES as they are engaged as research assistants and tour guides. This helps impart skills in research and conservation and promote their involvement in empowerment programs. Third, besides income and providing a learning opportunity, community members such as the research assistants create networks with the researchers which are essential for personal growth and connection to future opportunities.
8
1.6 Ethical Considerations in the Study
Ethical approval to conduct this study was sought from the Pwani University Ethics
Review Committee before starting the fieldwork. Further permission to conduct the study in a protected forest was obtained from the Kenya Wildlife Service (KWS) and Kenya
Forest Service (KFS). Throughout the research, the researcher ensured that wild animals had the right of way and that driving and riding at the reserve was kept below the maximum speed limit (40 kph). To ensure minimal disturbance, vegetation was not cleared unless very necessary and baiting was not conducted all through the data collection period. Finally, the intrusion of nesting sites was avoided.
1.7 Limitations of the Study
Weather conditions were the major limitation of the study. During rainy days, it was difficult to access some parts of the forest and the visibility of the nests was also reduced.
This limitation was overcome by shifting the data collection to a drier and thus more favourable period.
9
CHAPTER 2: LITERATURE REVIEW
2.1 A general overview of small mammals
Small mammals are part of a wide range of small animals, including birds, rodents, and insects. With a weight of less than 500g, they form the majority (> 90%) of the mammal population and an essential part of animal diversity (Hoffmann et al., 2010). There are several categories of small mammal species based on aspects, such as animal type, lifespan, and volancy. For example, Luza et al. referred those with a short lifespan, such as the Etruscan shrew and Suncus etrucus as slow-life history mammals and those with a long lifespan like the European hare and Lepus europaeus as fast-life history mammals
(Luza et al., 2019). Small mammals are also known to have varied lifestyles and diets.
Some, like the Atlantic bamboo rat, Kannabateomys amblyonyx have specialized habits and diets while others, like house mouse, Mus musculus are generalists (Luza et al.,
2019). These characteristics make small mammals unique in several aspects including their contribution to animal diversity and forming an essential part of the ecosystem with others such as deer, rabbits and wild pigs providing food to humans and other animals.
2.2 Distribution of Small Mammals
Small mammals are often hard to see or find mainly due to their characteristics and behaviours, such as nocturnal activity patterns, camouflage, and relative shyness (Cook,
2001). However, like other small animals, they are found in almost every part of the world, where they occupy different types of habitats and vegetation. A global database on the composition of non-volant small mammals developed by (Luza et al., 2019) showed that these animals are found in natural, neglected natural, and human-modified habitats. The natural and neglected habitats included natural forests, grasslands, savannahs, and their natural edges. Human-modified habitats were of five types: human-
10 induced forest edges, human-induced grassland edges, tree plantations, crop fields, and clear-cuts. Several ecological studies to determine the occurrence, abundance and preferred habitats of small animals have shown that small mammals have a wide geographical distribution in pine forests, oak forests, grasslands, freshwater marsh, and heathland sites (Cook, 2001). Of these geographical regions, forests remain the primary type of habitat for small animals, including mammals.
There are different types of forests in which animals live; however, small animals are mainly distributed in three major types of forests. These include tropical forests, temperate forests, and boreal forests (Carey and Harrington, 2001). Tropical forests are next to the equator and lack seasonal variations experienced in the Northern and Southern hemispheres occasioned by the earth's tilting on its axis. As a result, they contain the most plentiful and diverse wildlife (Rajpar, 2018). This is in contrast to temperate forests that have highly variable seasons. On the other hand, boreal forests are mainly found in the latitudes and have low biodiversity due to extreme conditions. They are dominated by northern evergreen conifers and the most challenging environment for most animals but suitable for quickly adapting animals, such as migratory birds and hibernating insects.
The characteristics of small animals have also been found to correlate with forest types.
Ceballos and Ehrlich noted that forest type significantly influenced the global distributions and diversity of mammals in terms of patterns of species richness, endemism (restricted-range species), and endangerment (threatened species) (Ceballos and Ehrlich, 2006). Specifically, the researchers showed that species richness of mammals was concentrated in tropical regions throughout the world. Other authors have reported a similar pattern while noting that species' distribution is limited to geographic ranges attributed to various factors (Gaston, 2003, Lomolino et al., 2010). Furthermore,
Ceballos and Ehrlich noted that certain regions throughout the world had a high diversity
11 of species, with the Amazonian lowlands and Andes of northern and western South
America, East Africa, and Southeast Asia reported to have the highest concentrations of species diversity (Ceballos and Ehrlich, 2006).
Similarly, the restricted-range species of small mammals are found in all the continents; however, only a few regions such as islands, peninsulas, and mountain tops have high concentrations (Ceballos and Ehrlich, 2006). These include a large area of Americas from central Mexico to the northern and central Andes and Brazil's Atlantic forests where relative continuous concentrations have been reported. The restricted-range species in
Asia are mainly found in Southern India and Sri Lanka, southwestern China, Vietnam,
Taiwan, Malaysia, Indonesia, Philippines, New Guinea, and northern Australia. In the
African continent, they are common in Cameroon's tropical lowlands in the west, inland, and coastal forests of East Africa, the Ethiopian highlands, and Madagascar. The threatened species are also distributed throughout the world, but higher concentrations are mainly in tropical centers of Africa, Asia, and the Western hemisphere (Ceballos and
Ehrlich, 2006). These regions have high-impact human activities, and to some extent, the occurrence follows species richness patterns.
Given these distribution patterns, it is imperative to note that small mammals' habitats are as diverse as the animals themselves. Forests are also the major habitat of small mammals; thus, they act as a livelihood source for humans and provide habitats for several types of animals, plants, fungi, and bacteria. Small mammals also tend to have varied distribution patterns, with some areas having higher concentrations of certain species than others. For example, (Cook, 2001) noted that rodents were mainly abundant in wetlands and oak forests compared to grasslands, pine forests, and heaths with low numbers.
12
2.3 Elephant Shrews
Elephant shrews or Sengis are small mammals whose origins date back to 50 million years ago (Existence, 2017). For a very long time, their history was not clearly understood with the Western scientists first describing most species in the mid to late 19th century in which they believed that sengis were closely related to true shrews, hedgehogs, and moles of the order Insectivora (Rathbun, 2020, Rathbun, 2009). The researchers later associated sengis with rabbits, primates, and ungulates after concluding that they had no close relationship to any other living mammals’ group. However, recent studies on evolutionary relationships have shown that sengis are not closely related to the shrews at all but belong to a distinctive group of African mammals, Afrotheria, which include other animals such as elephants, sea cows, hyraxes, aardvark, and tenrecs (Foundation, 2020,
Rathbun, 2020).
Sengis have a well-defined taxonomy belonging to the kingdom of Animalia, phylum
Chordata, class Mammalia, order Macroscelidae, and family Macroscelididae. However, the family has two main sub-families: the giant elephant shrews (Rhynchocyoninae) and the soft-furred elephant shrews (Macroscelidinae) (Rathbun, 2020, Amin et al., 2016).
The giant elephant shrews have a single genus (Rhynchocyon) with four extant species; namely, the golden-rumped sengi (Rhynchocyon chrysopygus), black-and-rufous sengi
(Rhynchocyon petersi), gray-faced sengi (Rhynchocyon udzungwensis), and the checkered sengi (Rhynchocyon cirnei) (Sabuni et al., 2011). On the other hand, the soft- furred sengis have three genera (Macroscelides, Elephantulus, and Petrodromus) with 15 species. These include the monospecific Petrodromus, three species of Macroscelides, and 11 species of Elephantulus genus (Rathbun, 2020). Therefore, the family
Macroscelididae has a total of four genera with 19 living species of sengis. However, it appears that the taxonomy of sengis is not well understood. According to (Rathbun,
13
1995), the genus Macroscelides is also monotypic, while Rynchocyon and Elephantulus have three and ten species, respectively. This results in a total of 33 species and sub- species of elephant shrews.
2.3.1 Description of Elephant Shrews
All elephant shrews are characterized by long and thin legs, long rat-like tails, large eyes and external ears, and long, flexible proboscis-like nose (Rathbun, 1979). They derive their name from their small size, extraordinarily long, flexible trunks, long pointed head, and very long, mobile, and trunk-like nose (Foundation, 2020). They use their long legs to move in a hopping fashion like rabbits. Sengis also have a hunchbacked posture and a long scaly tail with a gland underneath, which produces a strong scent used to mark territory and serves as a deterrent against many carnivores. In addition to having a good sense of smell for this purpose, they have well-developed senses of sight and hearing.
Elephant shrews are also semi-digitigrade as they walk on their fingers/toenails (Trust,
2020). Being mammals, they are also endothermic and bilaterally symmetrical. An adult shrew weighs 25g to 700g depending on the species and has a body length ranging between 22 and 30 cm and a tail, approximately a third of the body length (Foundation,
2020). However, these characteristics may slightly vary depending on the species. For example, giant sengis are the largest and most colourful, with adults weighing 350 g to
710g (Rathbun, 2009). The soft-furred species have similar body sizes ranging from about 25g for Macroscelides to about 200g for Petrodromus, while Elephantulus species weigh about 50 to 60g.
14
2.3.1.1 Golden-rumped Elephant-Shrew
Figure 1: Golden-rumped Elephant-shrew
GRES is one of the largest elephant shrews (Files, 2014a). Its scientific name is
Rhynchocyon chrysopugus, and it is also known as golden-rumped sengi. It has a long, flexible snout similar to all elephant shrews; however, their distinguishing characteristic is their bright gold-coloured rump patch and forehead covered by grizzled, stiff gold fur
(Jansa, 1999, Rathbun, 1979, Trust, 2020). The rump patch has an area of thickened skin
(a dermal shield) beneath, and it is thicker in males than females, which is thought to protect the male shrews from the scathing attacks of their fellow hostile males (Jansa,
1999). GRES is also dark amber coloured and has black legs, feet, and ears as well as a sparsely furred tail, which is half the total body length of 20 inches(56 cm) and predominantly black except the white distal third with a black tip (Trust, 2020, Jansa,
1999). The body is covered with a fine, stiff, and glossy fur, but ears are naked (Jansa,
1999). They have a 4 – 5 years lifespan (Trust, 2020, Files, 2014a). AnAge entry database for GRES also shows that the animal has an average adult weight of 540g (19.03 oz) and birth weight of 80g (AnAge, 2017). Moreover, GRES have sexually dimorphic canines, which differ in length between males and females. The males have longer canines of
15 approximately 6.6mm, which is 2mm longer than those of females, and believed to be used by males to attack other males during territory defence (Jansa, 1999).
2.3.1.2 Other Giant Sengis
Black and rufous elephant shrew and checkered elephant shrew are also common. Black and rufous elephant shrew has long proboscises essential for turning over litter and digging up beetles, as well as a long tongue for scooping up bugs (Jacques, 2013). It has a multi-coloured fur with reddish-brown front half and black back half. Adults weigh between 350g and 700g and have a body length of up to 31cm and a tail length of up to
250cm. Their hind limbs are much longer than forelimbs to enable them to move rapidly to avoid predators (Rathbun, 2009, Smit et al., 2011). Conversely, the checkered elephant shrew has several dark stripes on both sides and weigh between 410g and 550g (Gasior,
2006). Also, as one of the largest sengis, they have a similar body size to GRES and black and rufous sengis. Specifically, their body length ranges from 22.9 to 30.5 cm and a tail length of 17.8 to 25.4 cm.
2.3.1.3 Other types of elephant shrews
Other elephant shrews are mainly Elephantus and Macroscelides species. Similar to
GRES and black and rufous sengis, the rufous sengi (Elephantulus rufescens), short- eared or round-eared sengi (Macroscelides proboscideus) and four-toed sengis
(Petrodromus tetraductylus) are also common species (Rathbun, 2020). The less common ones include the eastern rock sengi (Elephantulus myurus), short-snouted sengi
(Elephantulus brachyrhynchus), and bushveld sengi (Elephantulus intufi). These organisms also have varied characteristics. For example, the rufous sengi is a reddish/brown small mouse-like animal weighing 25 to 60mg (Files, 2014b) while round- eared sengi is redder (Schubert, 2011, Dumbacher et al., 2012).
16
2.3.2 Distribution of Elephant Shrews
Elephant shrews have a wide distribution that covers many parts of Africa. Sengis of the order Macroscelidea are restricted to Africa and distributed throughout the continent, with southern and eastern Africa acting as the centre of diversity (Rathbun, 1995).
However, the distribution pattern varies across different species. Some like four-toed sengi, which is among the most widespread species found in the North-eastern corner of
South Africa to Central and Eastern Africa (Foundation, 2020). Conversely, those with restricted distribution are mainly smaller elephant shrew species, including
Macroscelides and Elephantulus, which are confined to Africa, specifically in the upland areas of southern, eastern, and north-western Africa (Rathbun, 1995). For example, eastern rock sengi, which is native to the Ethiopian geographical regions, is primarily distributed across the Southern African sub-region ranging from Mozambique in the north and Orange Free State to the south (Jones, 2002). Specifically, it is found in southern Zimbabwe, western Mozambique, eastern Botswana, and throughout the
Transvaal. In general, Elephantulus species are found in the most significant numbers in
South Africa, followed by east Africa.
Similarly, black and rufous elephant shrews occur only in central and eastern Africa.
Here, they occupy altitudes ranging from sea level to 2,300 metres high, including almost exclusively in the Udzungwa Mountains in Tanzania (Jacques, 2013). The checkered elephant shrews are only found in central Africa, while Macroscelides occur only in southwestern Africa (Rathbun, 2009). For instance, the round-eared sengi occurs in arid and semi-arid regions of South Africa, Namibia, and Botswana (Schubert, 2011).
Furthermore, GRES, which is endangered is endemic to Kenya. They are exclusively found in the coastal areas of Arabuko-Sokoke and Gede forests in Malindi that have remaining fragmented and small pockets of suitable forest (Jansa, 1999, Amin et al.,
17
2016). GRES is also found in a small range in the Dakatcha Woodland forests along the coastal Kenya (Trust, 2020). The areas also harbour black and rufous sengi. On the other hand, the grey-faced shrew is only confined to two forests in Tanzania.
Threatened species of elephant shrews also have limited distribution. These include five taxa of Rhynchocyon and one sub-species of Petrodromus tetradactylus confined to five geographical types of relatively isolated blocks or patches of forest or dense woodland in eastern Africa (Rathbun, 2009). These forest habitats include coastal lowlands in
Kenya and Tanzania, lowland forests islands in the shores of Tanzania, montane forests in Eastern Arc of Tanzania and Kenya and Rift Valley of Malawi, Zambia, and Tanzania; and Inland lowland forests of Uganda and Zaire. The limited distribution of elephant shrews to highly fragmented forests results in restricted populations, limited resource accessibility, and difficulty in finding a mate (Foundation, 2020). Despite the overall wide distribution of elephant shrews and being native to more than six African countries, it is surprising that no elephant shrews are found in western Africa and the vast Sahara region (Foundation, 2020, Rathbun, 2009).
2.3.3 Ecology and Behaviours of Elephant-Shrews
2.3.3.1 Habitats
Elephant-shrews occur in various habitats ranging from forests to woodlands. However, it is still difficult to see them partly attributed to their behavioural habits, such as foraging and resting (Rathbun, 2020). The habitats of elephant shrews also tend to vary with species. Giant sengis and Petrodromus generally inhabit lowland and montane forests, dense or closed-canopy woodlands, and thickets in eastern and central Africa (Rathbun,
2009, Rathbun, 2020). These areas usually have floors densely covered by leaf litter.
GRES is endemic to the moist, dense coastal scrub forest and lowland semi-deciduous
18 forest along the Kenyan coast (Jansa, 1999, Trust, 2020). The black & rufous elephant shrews are found in tropical and terrestrial forest and rainforest; however, like most giant elephant shrews, they also live in the lowland forests and dense woodlands (Jacques,
2013). They live in undisturbed forests, where they keep large territories and make nests on the ground from leaf litter. On the other hand, the checkered elephant shrew lives in an array of habitat ranging from dense forest to open plains (Foundation, 2020).
On the other hand, smaller elephant-shrew species are mainly found in more arid lowlands such as savannahs, scrublands, rocky outcrops, deserts, and uplands areas with dry forests, scrub, savannahs, and open country covered by sparse shrubs of grass
(Foundation, 2020) For example, the eastern rock sengi only inhabits areas with plenty of cracks and crevices where they live primarily in the rocky outcroppings of boulders in hilly terrain as they do not nest or burrow. The preferred inhabit areas within or near the boulders should provide adequate cover from predators with either vegetation or overhanging ledges. Short-snouted sengi is also found in the same region but only among the neighbouring sandy flat ground as their habitats rarely overlap. However, in almost all cases, sengis are found in low densities than many other small mammals (Rathbun,
2020).
2.3.3.2 Diet and Foraging
Elephant shrews eat invertebrates found in leaf litter such as ants, termites, spiders, beetles, millipedes, and earthworms. They also like bugs with black & rufous elephant- shrew relying on bugs found under leaf litter for food (Jacques, 2013). (Rathbun, 1979) noted that although all sengis are invertebrate specialists feeding on anything they find from the leaf litter and soil, most soft-furred species supplement their diet with small fruits, seeds, and green matter. Giant sengis also strictly feed on invertebrates by gleaning
19 leaf litter and soil and appear not to scavenge and poorly adapted to detecting food items not typically encountered in their habitats (Sabuni et al., 2011). GRES is insectivorous animal feeding on various invertebrates, such as earthworms, millipedes, insects, and spiders within its territory (Jansa, 1999, Trust, 2020). This involves the use of their long, flexible snout to overturn leaf-litter in search of food as well as forefeet to dig out the prey (Existence, 2017). Unlike many other small mammals, elephant-shrews feed during the daytime. They are also prone to attack by predators with the main ones, including snakes, raptors, birds of prey, harriers, and various carnivores (Existence, 2017).
2.3.3.3 Activity Patterns
Observational studies have shown that giant sengis are diurnal animals, while soft-furred sengis are mainly nocturnal but have some activity during the day (Amin et al., 2016).
(Rathbun, 1979) noted that GRES are strictly diurnal and most active during the day. Of all the observed activities, foraging was the most common activity at 79.15%, followed by walking (12%), defined as a movement with a particular goal without foraging. They also spent 3.8% of their time resting during certain times of the day, including lying down or dozing or calmly sitting with no indication of being alert of upset on the leaf litter.
They rest in nests constructed on the forest floor, which are mainly used for sleeping at night. However, tree barrows are also used for resting.
The nests are built every 1 – 3 days early in the morning, and by excavating a hollow in the soil and lining it with leaves, and covering the top with dry leaves (Jansa, 1999). The nest construction takes about 2 hours and is reported as the least activity at 2.2%
(Rathbun, 1979). Although Rhynchocyon species build their leaf nests on the forest floor, most soft-furred species use other species' burrows or construct their own (Rathbun,
20
2020). Other species also have specialised sheltering habits, such as living in rock crevices in areas containing boulders.
2.3.3.4 Mating and Courtship
Although all 17 species of sengis are known to be pair-living, detailed information on the social organisation is available for only six species: the golden-rumped sengi, the rufous sengi, the eastern rock sengi, the bushveld sengi, the short-snouted sengi, and the four- toed sengi (Schubert, 2011). GRES are monogamous and live as a couple in a stable monogamous pair in a common territory comprising several acres except where a partner dies necessitating a partner change (Trust, 2020). This includes close home range territories of around four acres each established by males and females as they do not live together in the same area and seldom stay together. Each sex is also responsible for defending their territory, with females chasing intruding females and males chasing off other males (Jansa, 1999).
They are not friendly to strangers and intolerant of close neighbours who would be violently evicted if one trespass into the territory (Foundation, 2020). Protection against an intruder or any threats such as predators involves using aggressive ways such as screaming, sparring, snapping, and kicking, all of which can happen so rapidly that it appears to be a blur of animals tumbling on the forest floor. Tail-slapping involves repeated slap of the forest floor or ground with their tail. The hind legs are also sometimes used while they run in case of a prolonged disturbance (Jansa, 1999). The auditory cues are particularly crucial in warning other elephant shrews of the presence of a predator.
Although the elephant shrews are seldom together, they communicate and check each other using scent markings or tactile and chemical responses with various glands, including perianal, sternal, subcaudal, and foot (Rathbun, 2020).
21
2.3.3.5 Reproduction and Parenting
Elephant shrews have a well-defined reproduction cycle, which is continuous at low latitudes and seasonal in high areas (Rathbun, 2020). Female GRES attain sexual or reproductive maturity at an average age of 38 days (AnAge, 2017). They are viviparous and give birth to an average of one offspring within a gestation period of 42 days.
Checkered elephant shrews also have a gestation period of 45 – 60 days (Foundation,
2020). GRES breed throughout the year and give birth 4 to 5 times a year. The young ones are most vulnerable when leaving their parents; therefore, they remain hidden in the nest for the first three works after birth when they are weaned (Foundation, 2020). They continue to follow the mother on foraging for about a week and become independent between 5 and 20 weeks when they can now leave their parent's territory to establish their own.
2.3.4 Ecological Importance of Elephant Shrews
Like other small mammals, elephant shrews are known to contribute to an ecosystem's well-being by playing a significant role in the environments in which they occur. (Cook,
2001) noted that several species of small mammals contribute to the ecological system's overall diversity, which provides aesthetic value to the environment and human. Besides, small mammals help balance the ecological system by directly influencing the population levels of plant materials and invertebrates they feed on (Foundation, 2020). For example, elephant-shrews feed on insects found in leaf litter; thus, ensuring natural checks on the insect population. Elephant shrews also help control pests and diseases since some of the insects they feed are pests and disease carriers. To other animal species, elephant-shrews have a commensal relationship. For example, red-capped robin-chats (Cossypha natalensis) birds usually follow GRES through the forest and feed on the invertebrates
22 leftovers (Jansa, 1999). Macroscelides also hosts a wide variety of parasites that depend on them for survival (Rathbun, 2020).
Small mammals also act as food sources to other carnivorous animals, such as snakes, lizards, birds of prey, weasel, fox, and coyote (Rathbun, 2020) as well as humans (Jansa,
1999). Other benefits to humans are providing a source of fur and recreational opportunities in hunting and wildlife viewing. Small mammals have also been reported to be beneficial to the flora and environment. They help propagate plants by acting as pollinators and seed dispersers, which help to support forest regeneration, maintain forest health, aerate the soil, and increase plant diversity. Unfortunately, small mammals could act as reservoirs for some infectious diseases that affect man.
2.3.5 Effect of Habitat Changes on the Ecology of Small Mammals and Elephant-
Shrews
The ecology of small mammals, such as abundance and community composition, could be affected by several factors. According to (Cook, 2001), these range from natural elements such as weather, habitats, and predators to human-made effects. A change in habitat, which refers to local environmental conditions in which an animal lives, is the most common factor affecting animal ecology (IUCN, 2020). The common changes include degradation, fragmentation, and loss occasioned by natural or human causes, such as droughts, diseases, fires, and human activities. According to (Rathbun, 1995), human activities that may include rapid expansion of population, logging, farming, hunting, and increasing development are the significant causes of habitat change.
Habitat change has affected animal populations, including the sengis, in several ways.
First, they have had adverse impacts on biodiversity. (Trust, 2020) noted that habitat change, particularly forest habitat loss and fragmentation, is the greatest threat to
23 organisms' biodiversity and survival. (Rathbun, 1995) noted that altered and trapped forests had lower black and rufous sengi and GRES densities than undisturbed forests, which are highly fragmented, small, and disappearing in eastern Africa due to human encroachment.
Similarly, (Yarnell et al., 2008) observed that fire significantly influenced home range, habitat use, and activity patterns of short-snouted sengi. Also, (Avenant, 2003) showed that ecological disturbance significantly influenced the small mammal density, species richness, relative abundance of the component species, and small mammals' diversity characteristics. A healthy and relatively stable ecosystem had many species, high diversity and evenness, specialists, and relatively low contribution of indicator species.
On the contrary, habitats with a lot of disturbance, especially those found at the border of a conservancy, had very few species, low density and evenness, absence of specialists, and small mammal component domination. Other factors such as species interactions, forest-floor characteristics, large coarse woody debris, and varied vegetation composition also affect small mammal characteristics such as composition and species abundance
(Carey and Harrington, 2001).
It has also been noted that elephant shrews could adapt to habitat change. For instance,
(Yarnell et al., 2008) observed that short-snouted sengi could adjust to a new environment after destroying their habitats by fire. Specifically, they noted that the destruction of grasslands mainly used by elephant shrews as habitat resulted in a shift to the use of thickets than grass. The fire-induced habitat modification also restricted the sengi movements to patches of unburned vegetation, with the proportion within the home range in thickets becoming greater post-fire.
24
2.3.6 Conservation of Small Mammals and Elephant Shrews
Although predators have been cited as a threat to elephant shrews, habitat loss and fragmentation have been identified as the most critical conservation issues for elephant shrews (Rathbun, 1995). This is mainly for the at-risk species, including all the four species of giant sengis and one sub-species of Petrodromus tetradactylus) which are confined to relatively isolated areas in eastern Africa. GRES is classified as vulnerable, endangered species with the coastal forests they live being cleared for agriculture with the 44 ha of the Gedi Historical Monument in Kenya remaining the only protection site
(Jansa, 1999). Furthermore, the black and rufous sengi has experienced a population decline of 20% to 30% in the past ten years due to a significant percentage of forest loss caused by human and drought-related fires (Foundation, 2020). In general, (Rathbun,
2009) noted that three-quarters of Rhynchocyon species are under threat.
Due to the nature of these threats, most conservation efforts and strategies for elephant shrews have been targeted at the threatened species and not the most common taxa
(Rathbun, 2020). The common species, such as Macroscelides proboscideus,
Petrodromus tetradactylus, and Elephantulus have a wide distribution, habitat diversity, and altitudinal range, and therefore facing little or no danger of extinction. The conservation interventions generally aim to prevent further fragmentation and loss of the restricted or fragmented forest habitats. In particular, most implemented measures focus on balancing human activities and protecting animals from mitigating human-induced habitat change and ensuring the survival of elephant shrews (Trust, 2020). Some of these include proper fencing and monitoring of forests to prevent human encroachment, improve and maintain forest quality, and addition of buffer zone to provide safe areas for local communities to collect firewood and herbs while at the same time keeping off poachers.
25
Emphasis has also been put on community-based conservation. The local communities are engaged in conservation through partnerships with local organisations and authorities to improve their relationships and the communities (Songorwa et al., 2000). This also involves educating the communities about endemic species within their area, the importance of animal conservation, and sustainable ways of using resources, generating income, and investing in livelihoods while protecting the wildlife. These approaches have been useful in conserving the endangered animal species, such as GRES in Kenya
(Ngaruiya, 2009b).
Science and technology are also emerging as a new front of conserving animals, including sengis. In particular, technology is being adopted to help identify critical landscapes that require intervention to set effective and sustainable development plans that improve people's lives while protecting wildlife. For example, the African Wildlife
Foundation uses Geographic Information System (GIS) technology and satellite images to determine forest areas that have been disturbed by human activities (Foundation,
2020). Camera trap monitoring and radio-tracking have also been used to determine the presence and absence of sengis (Sabuni et al., 2011). Other alternative solutions include supporting wildlife health programs, patrol based and intensive zone protections, habitat assessments, animals' relocation, and strategic planning for protected areas and critically endangered species.
In conclusion, conservation is crucial in maintaining the diversity of life on earth by preventing species extinctions, maintaining viable populations, and enabling the recovery of declining and depleted populations (Akçakaya et al., 2018). However, (Rowe and
Terry, 2014) noted that understanding animals' responses to environmental changes could also inform conservation and management efforts. Therefore, the combined approach to
26 wildlife conservation could protect endangered species across the world, including
Kenya.
2.3.7 Socio-Economic Benefits of Small Mammals
2.3.7.1 Creation of Jobs
The presence of small mammals has led to the employment of people as rangers and local guides to tourists and researchers who are interested in them.
2.3.7.2 Creation and Support of local business
Small mammals like other wildlife have led to the creation and support of related local business opportunities. Local businesses such as guest houses, honey making and wildlife-related enterprises such as curio shops and various travel and transport-related initiatives are supported hence acting as a source of income and improving the livelihoods of the local people (Jackson and Naughton-Treves, 2012, Emerton, 1999).
2.3.7.3 Boost to Tourism
The different wildlife components have been found to contribute a substantial and increasing economic returns in different places of the world. In Kenya tourism is quite well developed and wildlife-based tourism has been recognized as one of the country’s leading foreign exchange earner with different stakeholders from the community being recognized in wildlife management such as community enterprises, hoteliers and tour operators, local non-governmental organizations, local wildlife associations among others (Sindiga, 1995).
It has been found that tourists are more attracted to diversified attractions which in turn enhances destination loyalty as a result of tourist satisfaction. The unique contribution of
27 a site was also a factor considered by tourists which can be obtained from small mammals such as the endemic ones (Okello and Grasty, 2009).
2.3.7.4 Development of Infrastructure
Tourism as a result of wildlife including small mammals has led to development and increased access to infrastructure such as roads, communications, water supplies, health, education and security services.
2.3.7.5 Empowerment of the Local Communities
Empowerment impacts as a result of interaction of the local people with people from other communities include; opportunities for institutional development, transfer of skills and participation in local economic decision making (Ashley and Elliott, 2003). There are several projects that have been started as a result of conservation of mammals such as Jamii villa eco-lodge, butterfly farming, bee keeping and tree nursery in the communities around Arabuko-Sokoke Forest (Chiawo et al., 2018).
2.3.7.6 Ecosystem Services
Small mammals have various effects on the environment which includes; on vegetation, soils and other animals. On the vegetation small mammals have impacts on primary productivity through dispersal of seeds, plant species composition, and decomposition rates of plant materials. They also influence both physical and chemical properties of soils. Small mammals’ also prey on insects and therefore modify their environments in such a way as to provide habitat for other animals and increase habitat heterogeneity and biodiversity (Lacher Jr et al., 2019).
28
2.3.7.7 Option Value
This is the benefit that can be obtained as a result of the value of maintaining options available for the future. With the current changes being experienced the greatest value of biodiversity may lie in future opportunities brought to human beings being able to adapt to the local and global changes (Chardonnet et al., 2002).
2.3.7.8 Existence Value
The value for nature including small mammals that form part of it in satisfying human wants is not marketed and most of this value cannot be marketed (Tisdell, 2005). Various studies also argue that wildlife have an intrinsic value to exist irrespective of human attitudes on their existence. Additionally, there are moral principles relating to decisions on their existence (Stevens et al., 1991).
2.4 Arabuko-Sokoke Forest
The range of biodiversity in the coastal forests is dependent on the area, its climatic conditions, and site productivity. ASF is the largest of the intact coastal forests, and has been ranked as the second most important forest for conservation of bird species in addition to the endangered mammal species including those of two taxa that are globally endangered; the GRES and the Sokoke bushy-tailed Mongoose (Gereau et al., 2014).
2.4.1 Flora in Arabuko-Sokoke Forest
There are three major vegetation types in ASF (Arabuko-Sokoke Forest Management
Team, 2002)
Mixed Forest: This is a dense forest type that extends to about 7,000 ha on wetter
coastal sands in the east of Arabuko-Sokoke and occurs on grey sands. It has a diverse
29
tree flora including Afzelia quanzensis, Hymenaea verrucosa, Combretums
chumannii and Manilkara sansibarensis and the cycad Encephalartos hildebrandtii.
Brachystegia Forest: This is a more open forest covering about 7,700 ha, dominated
by Brachystegia spiciformis on drier and infertile white sands which are infertile
through the center of the forest. It is a relatively open habitat that is dominated by
Brachystegia spiciformis.
Cynometra Forest: This is a dense forest or thicket on the north-west side of
Arabuko-Sokoke, covering about 23,500 ha on the red Magarini sands towards the
western side of the forest. It is dominated by trees of Cynometra webberi and
Manilkara sulcata, and the euphorbia species Euphorbia candelabrum, but with
reducing numbers. Brachylaena huillensis also used to be abundant in this zone, but
its numbers have been severely reduced by extraction mainly for the carving trade. A
previous study showed that the occupancy of Rhychocyon was over 80% in
Cynometra in ASF (Amin et al., 2016).
2.4.2 Fauna in Arabuko-Sokoke Forest
Arabuko-Sokoke Forest is home to various species of mammals, birds, reptiles, and amphibians. Some of the mammals within the forest include the African Elephant,
African Buffalo, African Civet, Caracal, Syke’s Monkeys, Yellow Baboons, and Lesser
Galago (or bushbaby). Endemic species such as Sokoke bushy-tailed mongoose and the
Golden-rumped Elephant-shrew and the critically endangered Aders’ Duiker are also found within this coastal forest (Nyunja, 2017).
ASF has a unique and ecologically adapted bird species. Sokoke Scops Owl, Clarke’s
Weaver, Sokoke Pipit, Amani Sunbird, Spotted Ground Thrush and The East Coast
Alakat are globally threatened bird species found within the forest. The birds in ASF
30 have been found to show distinctive distribution among the three habitat types with
Brachystegia woodland holding the richest avifauna (Mulwa, 2017).
According to research conducted by (Nyamache and Malonza, 2017), amphibians were found to be many compared to reptiles, with 25 species of amphibians (3 species of frogs) and 24 reptiles species (3 snakes, 1 tortoise, and 20 lizards) having been recorded. Some of the reptiles and amphibians found in the forest include; Savanna Ridged frog, East
African Puddle Frog, Great plated lizard, Small scaled burrowing asp, Cape Brown house snake, Tree gecko and Puff adder among others (Nyamache, 2017).
2.4.3 Status of Arabuko Sokoke Forest
Currently, biodiversity extinction is one of the greatest challenges that the world is facing. One of the greatest modern-day threats to terrestrial species worldwide is habitat change which includes habitat loss, habitat fragmentation, and habitat degradation
(Bellard et al., 2014). It is worth noting that forest classification in ASF has changed over time as a result of the transformation that the forest is undergoing as a result of disturbance (Habel et al., 2017).
The ASF Cover Survey proved that there are illegal socio-economic activities within the forest leading to disturbances and that logging is still ongoing to much higher levels in a globally significant reserve such as ASF (Waters et al., 2007). Eastern Arc and Coastal
Forests of Tanzania-Kenya are among the forests which cannot afford to lose more habitat with further habitat loss being expected to be detrimental (Brooks et al., 2002).
Previous studies have estimated the global rate of deforestation at 0.7% annually although the rate is different for each country (Abelson, 1996). According to (Watch,
2002), Kenya had lost 7.1% of its humid primary forest between 2002 and 2019. It was found that factors determining the location of disturbance in the forest include
31 accessibility, vegetation characteristics, and the type of forest. It was noted that people targeted both small and big trees of Cynometra and those that are deeper in the forest, but extraction of poles is mainly from the mixed forest which is near the edge (Waters et al., 2007).
2.4.4 Local, National, Regional and Global Value of Arabuko Sokoke Forest
Arabuko Sokoke Forest is used by the community for various purposes it provides; fuel wood to the rural house-holds and was previously a source of pole wood for construction purposes. The forest has been transformed through modernized and coordinated extraction and marketing of coastal forests products. It has been reported that in the year
2001, the communities around Arabuko-Sokoke Forest earned around 37,000 dollars from guiding, bee-keeping and butterfly farming. At a national level Arabuko Sokoke
Forest is an important tourist destination area. Besides, ASF has been developed for ecotourism with walking nature trails have been cleared to attract tourism providing an opportunity to walk; scenery attractions; bird watching; mammals viewing; and butterfly exhibit (Matiku, 2002).
The biodiversity value, research, and potential use are probably the most important global uses and values. The forest contributes in the local and global climate by acting as carbon sinks, have unexploited pharmaceutical potential and is a key research area for international researchers (Matiku, 2002).
2.4.5 Threats to Arabuko Sokoke Forest
2.4.5.1 Deforestation and Forest Degradation
Deforestation and forest degradation are one of the major threats to forests all over the world. Deforestation happens when forests are converted to use by other non-forest issues
32 such as development of infrastructure. On the other hand, forest degradation happens when forest ecosystems are unable to provide essential goods and services to people and nature as they would have without interference. Since forests contain a lot of terrestrial of terrestrial biodiversity their degradation threatens the survival of many species and reduce their ability to provide essential services such as climate regulation, provision of clean air and water among others. Forest fragmentation which refers to the conversion of continuous forest land into patches of separate non-forested lands also leads to forest degradation.
Tropical forests cover less than 10% of the earth’s surface and have been reported to be very diverse, supporting at least two-thirds of the world’s biodiversity (Giam, 2017) .
Despite of this they have are still under pressure from deforestation and degradation activities
There are various agents that lead to deforestation, degradation and fragmentation of forests and they include; commercial farmers, cattle ranchers, livestock herders, loggers, firewood collectors, mining and petroleum industrialists, land settlement planners and infrastructure developers (TEJASWI, 2006).
2.4.5.2 Climate Change
Climate can be referred to as long-term weather patterns that describe a region whereas weather is the state of the atmosphere at a specific time in a specific place. Climate variability refers to variations in the usual state of the climate at temporal and spatial scales. This variability may be brought about by internal processes within the climate system or as a result of variations in natural or human-related factors. Global climate change indicates a change in either the mean state of the climate or in its variability, persisting for several decades or longer (Adedeji, 2014).
33
Climate change can be signified by; increases in average global temperature (global warming); changes in cloud cover and precipitation mainly over land; melting of ice caps and glaciers and reduced snow cover; and increases in ocean temperatures and ocean acidity as a result seawater absorbing heat and carbon dioxide from the atmosphere
(Change, 2007)
The changes in temperatures and rainfall patterns as a result of climate change are expected to produce a strong direct impact on forests. In addition, there are climate change-induced modifications of frequency and intensity of forest wildfires, outbreaks of insects and pathogens, and extreme events such as high winds which also have impacts on forests other than the direct impact of higher temperatures and elevated Carbon (IV) oxide (Kirilenko and Sedjo, 2007).
Climate change has potential effects on the composition, growth rates and regenerative capacity. The vulnerability has been noted to be higher for forests ecosystems that border the oceans and seas such as Arabuko Sokoke Forest because of the predicted rise in sea level (Tarus, 2018).
2.4.5.3 Pests
Currently there are a lot of environmental changes occurring in which aggressive plant and insect species may be able to gain advantage over the native species. This may lead to an increase in their numbers and most likely a greater negative impact on the forests
(Randhir and Erol, 2013).
2.4.5.4 Mining
The energy demand in the world is increasing all over the world with the growing industrialization, living standards and increasing human population. The greatest concern
34 is the finite nature of most high-energy-density fuels such as petroleum and its derivatives which power almost the entire global transport sector and is also used in electricity generation. It is expected that consumption of petroleum will reach the peak a point at which the supplies will begin to decline and biofuels will be the most likely alternative to petroleum. Furthermore, the expansion of biofuel production will be likely to be concentrated in the tropics where the plant grows fastest and the land is typically the cheapest. This will therefore promote the large-scale conversion of tropical forests for growth of biofuel crops since the opportunity cost for conservation will be increased
(Laurance and Peres, 2006). Silica sands for glass manufacture have been mined in
Arabuko-Sokoke Forest in the past (Gachanja and Kanyanya, 2004). Arabuko Sokoke
Forest is believed to contain titanium deposits with many private investors showing interest to start mining (ASFMT, 2002).
2.5 Golden-rumped Elephant-shrews in ASF
GRES can be differentiated from others since it is dark amber in colour with black legs and feet and has a bright yellow rump patch with slightly longer hair (Rovero et al.,
2008). They have the most restricted range of all the elephant-shrews since it has only been reported in the North Kenya coast (FitzGibbon, 1994). GRES are monogamous and breed throughout the year giving birth to single young which is found in a forest with dense woodland and thicket that support dense leaf litter on the ground.
2.6 Status of GRES in ASF
The abundance of a species is critical for its conservation and management especially to an endangered and endemic species like GRES. Determining the abundance of GRES will be an important indicator of their extinction risk. Being a difficult animal to actually observe and therefore count, the method that has been employed to estimate population
35 size has been to use the number of nests counted as an index of abundance (Fitzgibbon,
1994). Using this method, Fitzgibbon found GRES population to be 21,897. According to a survey by (Bauer, 2001) the population trends of GRES had declined by 30% from
1993 from an estimated 20,000 to 14,000 individuals. (Ngaruiya, 2009a) estimated the population of GRES in ASF to be about 12,750 individuals by employing the use of nests in distance sampling method, with Cynometra having the highest number of individuals owing to its large size. The mixed forest had a higher density since it has highly favourable habitat characteristics than Brachystegia forest. Further analysis from this study revealed that there was an overall 9% decrease from surveys done by (Bauer, 2001) with specific trends indicating that there was a 12% increase in population in Cynometra
Forest (9,196), a decline of 21% in the Brachystegia forest (1,464) and 50% in the mixed forest (2,080) (Ngaruiya, 2009a)
Another study conducted by use of camera traps suggested that GRES occurrence is associated with habitats that have closed canopy and dense litter cover (Amin et al.,
2016). According to IUCN, the current status of GRES in the forest habitats is unknown and there are no recent accurate estimates on abundance, hence, relying on assumptions that habitat and population continue to reduce due to the various anthropogenic factors identified in the past (FitzGibbon and Rathbun, 2014).
2.7 Nesting of GRES
All the elephant-shrews within the genus Rhynchocyon make nests that are used as a shelter from harsh environmental conditions as well as during breeding. The nests are usually constructed from dead leaves on the forest floor occupying a small depression in the soil and are approximately 10cms deep and 30cms in diameter hence, relatively conspicuous and easy to detect (Fitzgibbon and Rathbun, 1994). To make the nest, the
36 elephant-shrew excavates the forest floor to make a depression by using long scraping sweeps of the forefoot. Dead leaves are then brought together and arranged into a layered lining by rapidly vibrating both front feet on the leaves surface after which the rest of the nest is made by piling additional leaf litter on the top of the nest. Interestingly none of the other Macroscelidinae build or use nests as shelter. Instead they shelter in rock crevices, among boulders at the bases of bushes and thickets with others excavating burrows in soft substrates (Rathbun, 2009). It has been found that GRES uses several nests at any given time, with a study by Fitzgibbon (1994) calculating the number to be six nests per GRES. They maintain the nests for a period of three to four months and research has shown that they usually go back to the unused nests after some time
(FitzGibbon, 1994). Some of the reasons suggested for why GRES builds so many nests include to allow for time for reduced ectoparasite load in the nests, to give size allowance in female size during gestation, to maximize the foraging time in new pre-rich areas and to maintain freshness in nests by avoiding the accumulation of decomposing leaves and discouragement of predators that search for prey in nests (Ngaruiya, 2009a). The nests are also important since this is where the young ones that have been born are confined to for around two weeks. Although previous studies by Fitzgibbon and Rathbun (1994) and
Ngaruiya (2009) have used nesting sites to determine GRES abundance, no studies have been conducted to determine the characteristics of GRES nest sites. Characterizing nesting sites by finding out the nesting materials and litter cover around would provide insights on the preferred nesting sites by GRES. Furthermore, there is a probability that nesting material could have changed to suit the changing forest conditions due to habitat destruction, leading to reduced forest quality, hence, impacting the availability of some forest components.
37
2.8 Threats to GRES in ASF
ASF is adversely affected by logging which is still ongoing up to date, to the levels that are unacceptable to a globally significant reserve (Waters et al., 2007). Factors that were found to influence the location of the disturbance include accessibility, vegetation characteristics, and the type of forest. Habitat loss is one of the main factors affecting the distribution of species. Reduction in the canopy cover and leaf litter affects many aspects of the life of sengis since leaf litter and canopy cover are important in their feeding and nesting that are important aspects for their survival. Although logging results to reduced forest quality, a study by (Bauer, 2001) did not find any direct relationship between logging and GRES, in some cases GRES nests were found at the base of stumps. The level of timber collection had also reduced with the percentage of the cut stem being 15.4 and 0.18 in studies conducted in 1993 and 2000 respectively (Bauer, 2001; FitzGibbon,
1994). This reduction was attributed to increased enforcement by the forest department.
GRES is not preferred as a source of food by the locals since it makes a tasteless dish and it is believed to feed on litter. Besides, some communities did not hunt or eat GRES since the religion of Muslim forbids eating of non-hooved animals. GRES was also associated with witchcraft and pregnant women were forbidden from eating the animal since the effects would be carried to the unborn baby (Ngaruiya, 2009a).
According to a report on the state of ASF in 2007, the habitat was intact with some disturbances having been experienced along the electrical fence area. In the years 2010 and 2011, there were cases of isolates tree logging for timber, carving, firewood, and building poles in addition to mammal poaching. All these activities were going on despite the efforts of KWS and KFS banning non-residential cultivation in the forest and other illegal activities, and improving forest patrols in 2005-2006 (Ndang'ang'a et al., 2016).
Loss of hollow trunks in ASF which are used as refuges for elephant-shrews may also be
38 contributing to the reduction in numbers that have been seen from previous research
(Rathbun and Kyalo, 2000). The effects of deadwood on GRES have not been studied before, although it may have an impact since deadwood hosts saproxylic vertebrates such as beetles that are food to GRES.
2.9 Conservation Status of GRES
For proper ecosystem planning and implementation of strategic adaptive management programs in protected areas, it is crucial to understand the composition and distribution of species in the ecological system. With the ever-increasing impacts from anthropogenic activities and climate change, mammalian species composition and distribution undergo adverse impacts. In addition to that, conservation areas also experience changes in plant biodiversity which can then result in great negative impacts on terrestrial vertebrates. As such biodiversity hotspots such as ASF need frequent monitoring to be incorporated into conservation and management planning (Nyunja, 2017).
According to IUCN, GRES has been listed as endangered because it’s extent of occurrence is less than 5,000km², habitat that is fragmented to a large extent and also the fact that there has been a continuous reduction in its extent of habitat and the number of mature individuals (FitzGibbon and Rathbun, 2015).
2.10 Conceptual Framework
GRES population and their nesting sites are hypothesized to be influenced either directly or indirectly, by a myriad of anthropogenic activities such as logging, agricultural activities, and human population (Figure 2).
39
Figure 2: Conceptual Framework
Logging, firewood collection, increased human population, and agricultural encroachment are factors that may lead to degraded forest quality, contribute to habitat fragmentation or even lead to habitat loss. Degraded forest quality creates an environment that is not conducive for GRES to thrive; for example, reduced cover and litter and interference of nesting sites which may eventually lead to loss of preferred habitat sites, hence, impacts on nesting sites and GRES population. This study focussed on finding out whether dead wood availability plays a role in GRES preferred nesting sites and hence its abundance. Also, it sought to understand factors such as litter cover and nesting material availability in GRES nesting locations.
40
CHAPTER 3: MATERIALS AND METHODS
3.1 Study Area
The study was conducted in Arabuko-Sokoke Forest located in the Coast region of Kenya in Kilifi County. It traverses Kilifi North and Malindi Sub counties at a latitude of 3˚20΄S and longitude 39˚50΄E. The forest covers an area that is approximately 41,600ha and is the largest single block of indigenous coastal forest remaining in East Africa. The forest has three main vegetation types: 1) the Mixed Forest; 2) Brachystegia Forest and; 3)
Cynometra (Arabuko-Sokoke Forest Management Team, 2002). The three vegetation types were included in this study as they were the main vegetation types. The other is
Plantation vegetation which was not included as it included as pockets of planted trees within the other three vegetation types and to allow comparison of the data collected with previous studies. Arabuko Sokoke Forest is the most important site for Golden rumped
Elephant-shrew thus its inclusion as the study site (FitzGibbon, 1994).
Figure 3: Map of Arabuko-Sokoke Forest
41
3.1.1 Climatic Conditions
Rainfall in Arabuko-Sokoke Forest is bimodal with the long rains being between April and July, and short rains between October and December. The other months are usually hot and dry with January and February being the driest months (Pellikka and Clark,
2004). The annual average rainfall varies from less than 600mm north-west part of the forest to over 1000 mm at Gede in the east. The temperatures are usually high throughout the year with a daily mean of 25˚C and little variation from month to month with March being the hottest month (Arabuko-Sokoke Forest Management Team, 2002).
3.1.2 Topography and Soils
Arabuko Sokoke Forest lies on a flat coastal plain at an altitude of about 45 m above sea level on the eastern side. This rises to a plateau of about 60–200 m in the central and western parts of the forest (Arabuko-Sokoke Forest Management Team, 2002). The forest red magarini sands on the western side of the forest, the wetter coastal sands in the eastern part of the forest and the drier, infertile white sands through the center of the forest.
3.1.3 Fauna in ASF
ASF is the largest of the intact coastal forests, and has been ranked as the second most important forest for conservation of bird species in addition to the endangered mammal species (Gereau et al., 2014). Some of the endemic species include the Sokoke bushy- tailed mongoose and the Golden-rumped Elephant-shrew and the critically endangered
Aders’ Duiker are also found within this coastal forest (Nyunja, 2017).
Additionally, the forest has unique and ecologically adapted bird species. Sokoke Scops
Owl, Clarke’s Weaver, Sokoke Pipit, Amani Sunbird, Spotted Ground Thrush and The
42
East Coast Alakat are globally threatened bird species found within the forest. The birds in ASF have been found to show distinctive distribution among the three habitat types with Brachystegia woodland holding the richest avifauna (Mulwa, 2017).
3.1.4 Flora in ASF
There are three major vegetation types in ASF namely; Mixed Forest, Brachystegia
Forest and Cynometra Forest (Arabuko-Sokoke Forest Management Team, 2002). Mixed
Forest is a dense forest type with diverse tree flora including Afzelia quanzensis,
Hymenaea verrucosa, Combretums chumannii and Manilkara sansibarensis.
Brachystegia Forest is a more open dominated by Brachystegia spiciformis. Cynometra
Forest is a dense forest or thicket dominated by trees of Cynometra webberi and
Manilkara sulcata, and the euphorbia species Euphorbia candelabrum, but with reducing numbers. Brachylaena huillensis also used to be abundant in this zone, but its numbers have been severely reduced by extraction mainly for the carving trade.
3.1.5 Population around ASF
Arabuko-Sokoke Forest is surrounded on all sides by villages. There are around 50 villages around the forest. The average population density of the community around the forest has been estimated at 47–60 people per km² (Chiawo et al., 2018). The main ethnic group is the Giriama who displaced the Sanya. The Sanya community were the original forest dwellers and hunters within the forest. At the moment, the forest-adjacent population is dominated by small scale subsistence farmers who utilise the forest for some of their livelihood requirements. The main crops grown by the farmers are maize, cassava and beans. There are also locally grown cash crops which include; coconut, mango, cashew-nuts and sesame. There are no squatters in the forest (Arabuko-Sokoke
Forest Management Team, 2002). The adjacent households have since in the used the
43 forest to meet their needs such as obtaining wild honey and other non-woody products.
Currently, there are conservation initiatives that engage the community hence generating income such as the Kipepeo project in which butterflies are raised and exported
(Gachanja and Kanyanya, 2004).
3.2 Study Design
This study employed a descriptive research design. Descriptive research is designed to describe a population, situation or phenomenon hence answering the what, where, when and how questions (Lans and van der Voordt, 2002). The study design was found to be appropriate since it seeks to describe the characteristics of GRES in terms of the population during the study period and also to provide insights on their nesting behaviour.
3.3 Sampling Procedure
Stratified random sampling was utilized for this study due to natural variability that occurs in ASF, brought about by the different vegetation types. This study, therefore, had three main study sites or strata namely; Mixed Forest (7,000ha), Cynometra Forest
(23,500ha), and Brachystegia Forest (7,700ha). The study sites in proportion are 18%,
62%, and 20% respectively. Line transects were then laid randomly in each of the three vegetation types following a random distance generation using the RANDBETWEEN1 function in excel. The number of transects laid varied with the size of the study site.
According to Buckland (2015); it is advisable to allocate effort in proportion to stratum size rather than assign effort in high-density areas if the objective of the study includes relating animal density to habitat (Buckland et al., 2015).
1 RANDBETWEEN is an excel function that assigns a given number randomly between a given range.
44
3.4 Data Collection Procedures and Techniques
Data was collected with the help of research assistants who were trained on the data collection procedures and techniques. The research team was driven to a new area in a given vegetation type each data collection day. Within each vegetation type, a new area to lay transects was selected each day before starting the fieldwork to ensure that transects were nearly evenly distributed throughout the forest while preventing bias that could have resulted when the areas were selected while in the forest. Furthermore, the number of transects to be laid in an area were estimated before visiting the area with effort being determined by the stratum size. Once an area was selected, a start point (the start of an area stretch along the driving track) was selected and the distance between the start and end of the stretch was used to determine the distance the researchers would move along the driving track to the point to branch into the forest to lay the transect. For instance, if three transects were to be laid in a 6 km stretch, three random numbers (between 0 and
6) were generated using the RANDBETWEEN function in excel and used as the distances the research team would move from the starting point to the point to branch off into the forest. The researchers then drove along the driving track to the nearest first point generated and then branched 300 meters either to the right or left of the point. For consistency purposes, the researchers alternated which side of the track the transect was to be located with the first point always to the right. The study was aimed to ensure that the sightings were adequate for reliable estimation of the detection function. According to Buckland et al., 60-80 detections for line transect are adequate for a reliable approximation of the detection function (Buckland et al., 2004). Data collection was conducted across two months in September and October 2019.
Data was collected in developed datasheets and notes were taken using notebooks to ensure that any further information was recorded.
45
Objective 1: To determine the abundance of Golden-rumped elephant-shrews (GRES) in
Arabuko-Sokoke Forest during the study period.
Line transects of 100 meters’ length were placed randomly in the three study sites; Mixed forest, Brachystegia forest, and Cynometra forest upon branching off as described above.
The start and end GPS location of the line transect were recorded. A single observer walked along the transect line looking on both sides of the transect to locate nests. The same observer was used throughout the study period to reduce observer bias and differences. Once a nest was sighted, the perpendicular distance from the line transect to the nest was measured using a tape measure. GPS coordinates of the nest were also taken.
The nest was then observed to check on the material composition which was classified as either having leaves only, twigs only, or leaves and twigs. The age of the nest was also noted with heaped nests classified as new while one that was flat or almost flat at the top classified as old. Litter depth was measured using a ruler 1m from the nest. An index of vegetation density was estimated using a checkerboard. The researcher moved to four perpendicular points that were located 2 m from the nest and estimated the percentage of boxes on the checkerboard that were not covered by vegetation to enable calculation of the percentage of boxes covered. Canopy cover was also estimated at the four points where vegetation density was estimated by observing and estimating the percentage area of a circular tube of 4.5 cm diameter where the sky was not visible. If the sky was totally visible without any blockage from any vegetation then the canopy cover would be 0%.
Furthermore, canopy height at the nest was estimated by estimating the height of the tallest tree covering the nest location.
46
In summary, the following were recorded:
The occurrence of GRES nests on either side of the transect
The perpendicular distance from transect to the nest sighted.
Exact GPS location, the material composition of the nests, age of the nest, and litter
depth.
The following assumptions were considered for reliable estimation;
All nests directly on the transect were always detected
Distances from the transect to the nest were measured accurately
Nests were identified correctly
Nest's locations were independent of the positions of the transects.
Objective 2: To establish the relationship between deadwood biomass and abundance of
Golden-rumped-elephant-shrews in Arabuko-Sokoke Forest.
Sample plots of 10 m by 10 m located at the centre of each transect were used to investigate deadwood. On the 10 m by 10 m sample plots, all pieces of deadwood with a minimum large-end diameter of 2 cm and length 20 cm were measured (Eräjää et al.,
2010). The information collected included diameter in the minimum section, diameter in the end section, and the length between the minimum and maximum section (De Meo et al., 2017). All measurements were done using a tape measure. Furthermore, for each piece of deadwood, the degree of decay was recorded. The degree of decay was grouped into three classes. Class 1 included solid wood that was recently fallen, with an intact
47 bark or bark that is starting to fall off. Class 2 included non-solid wood that was in poor condition in which a nail could be pushed into the wood by hand. Class 3 included soft, rotten wood which easily collapsed when stepped on (Baker and Chao, 2009).
Objective 3: To assess the determinants of Golden-rumped elephant-shrew nesting sites and nest composition.
We computed means, medians, percentages, chi-square tests and conducted t-tests to examine this objective. We first restricted the data to include those transects where at least one nest was observed. Then, for each vegetation type we examined the distribution of these nests across several independent variables, namely: - the age of the nest (whether
Old or New), canopy cover (%), vegetation density and measurements of litter depth
(centimetres) 1 m from the nests were taken. The material composition of the nests was then examined across the three vegetation types by examining the distribution of the material compositions of nests categorised as Leaves only or Twigs only or Leaves and
Twigs.
48
Figure 4: Transects sampled in Arabuko Sokoke Forest for this study
3.5 Data Analysis Techniques
3.5.1 Abundance Estimation
The perpendicular distances from the line transect to the sighted nests were used to model a detection function stratified by the vegetation type. Different detection functions for each vegetation type were estimated using the Distance program version 7.3 (Thomas et al., 2010) to calculate GRES abundance. To estimate density and abundance the following formulae was used using the Multiple Covariate Distance Sampling (MCDS) engine in the Distance software (Buckland et al., 2004, Buckland et al., 2015);
푛 퐷̂ = . f(0) 2퐿
49
Where; 퐷̂=Density n= objects detected at perpendicular distances f (0) = the probability density function of the perpendicular distances L = Length of the transect
MCDS detection function models were fitted by applying the half-normal or hazard rate key-functions and cosine, simple polynomial, or Hermite polynomial series expansion with stratification by vegetation type and using litter-depth, vegetation density, canopy cover, and canopy height as covariates. These covariates were assumed to be a priority to influence the detection probability. They were entered into the model as non-factors and examined their effect on the detection probability. The model with the least Akaike information criterion (AIC) was then selected and presented. Additionally, the
Kolmogorov-Smirnov and the Cramér-von Mises goodness-of-fit tests were used. The
Kolmogorov-Smirnov goodness-of-fit test focusses on the largest difference between the observed and expected distances where a p-value of less than 0.05 is interpreted as a poor fit while a p-value greater or equal to 0.05 interpreted as evidence for a good fit. On the other hand, the Cramér-von Mises goodness-of-fit tests use the overall departure
(difference) between the observed and expected distances to indicate whether there are significant problems. In this test, a p-value of less than 0.05 indicates significant problems in the departures between data and fitted models (Buckland et al., 2004,
Burnham et al., 1980, Marques et al., 2007).
The resulting estimated densities for each vegetation type was then divided by a correction factor of 0.49 to take into account the multiple nest use by GRES and following a similar approach by (Ngaruiya, 2009a).
50
2.10.5 Deadwood Calculation
This study employed the Huber’s formula to calculate deadwood volume. The Huber’s formula was selected as evidence from (Filho et al., 2000), who indicated that the
Huber’s formula was superior compared to other methods of calculating deadwood volume.
Huber’s formula;
V= Coarse Woody Debris Volume
L= Length of the log
dm= mid-diameter of the log
A linear regression model (Seber and Lee, 2012) was fitted to relate the number of nests and deadwood volume observed per transect for each vegetation type and overall for
ASF. The linear regression was chosen due to the continuous nature of the outcome variable (number of nests).
2.10.6 Determinants of Nest Location and nest composition
Different approaches were used to examine the characteristics of the areas where nests were found in each of the vegetation types. First, means were computed for continuous variables (canopy cover, canopy height, litter depth, and vegetation density) while percentages were calculated for categorical variable such as nest material composition
(whether a nest was composed of leaves only, leaves and twigs or twigs only). Second, the Student t-test was computed to assess whether there were significant differences in
51 the means of the continuous variables by comparing different vegetation types. Third, the
Chi-square test was used to determine whether there was a significant association between nest material composition and vegetation types.
52
CHAPTER 4: RESULTS
4.1 Survey Effort
A total of 44 transects covering a distance of about 4,400 m across the three main vegetation types in ASF were covered. Table 1 summarises the distribution of the transects, the maximum distance of sighting a nest, transect length, and the number of nests observed in each vegetation type. Most nests were observed in the Cynometra (62) vegetation type while Brachystegia had the least (23). However, sighting distance was longest in the Brachystegia (15.8 m) vegetation type but the shortest in Cynometra (8 m).
Table 1: Distribution of transects by vegetation type
Total The maximum Number of Number of Forest Type Transect distance of sighting Transects Nests length (m) a nest (m) Cynometra 24 2,400 62 8 Brachystegia 11 1100 23 15.8 Mixed 9 900 27 10 Forest Total 44 4,400 112 N/A
4.2 Perpendicular distances
On average, nests were sighted at a longer distance from the transect in the Brachystegia vegetation type (5.98 m) compared to 3.35 m (p-value =0.0005) and 3.76 m (p-value
=0.0437) in the Cynometra and Mixed forest respectively (Figure 3). Across the three vegetation types, the maximum distance of detecting a nest was 15.8 m and the median distance was 3.3 m. Figure 3 shows the distribution of the perpendicular distance data by vegetation type.
53
Figure 5: Distribution of nest sighting distance from transect by vegetation type in Arabuko-Sokoke Forest
Models with the half-normal cosine key-function had the least Akaike information criterion (AIC) (518.48) value compared to those with a hazard-rate key-function
(530.65), hence, detection functions from the half-normal cosine models are presented.
An examination of the perpendicular distance from the transect to a nest in each of the vegetation types did not indicate any apparent rounding issues (Figure 4 to 6). Rounding of distances to favoured values usually results to an unreliable detection function hence indicating a problem with the data. Furthermore, the Kolmogorov-Smirnov goodness-of- fit statistic indicated a very good fit across the three-vegetation type where Dn = 0.1479
(p-value=0.1329) in Cynometra Forest, Dn = 0.1085 (p-value=0.9081) and Dn = 0.1388
(p-value=0.7671) in the Mixed Forest and Brachystegia vegetation types respectively.
This also provides evidence of no rounding of observed distances in the data.
54
Again, the Cramer-von Mises goodness-of-fit test showed no significant departures between the observed data and the fitted model across all three vegetation types.
1.2
1.0
0.8
0.6
0.4
0.2
0.0 0 2 4 6 8 10 12 14 16 Perpendicular distance in meters
Figure 6: Histogram of the estimated MCDS detection function for GRES averaged over the observed covariate values for litter depth, canopy cover, canopy height, and vegetation density in the Cynometra vegetation type.
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0 0 2 4 6 8 10 12 14 16 Perpendicular distance in meters
Figure 7: Histogram of the estimated MCDS detection function for GRES averaged over the observed covariate values for litter depth, canopy cover, canopy height and vegetation density in the Mixed Forest vegetation type
55
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0 0 2 4 6 8 10 12 14 16 Perpendicular distance in meters
Figure 8: Histogram of the estimated MCDS detection function for GRES averaged over the observed covariate values for litter depth, canopy cover, canopy height, and vegetation density in the Brachystegia vegetation type.
4.3 Detection Probability
Nests were more likely to be detected in the Cynometra forest followed by mixed-forest vegetation and least in the Brachystegia forest, however, the difference was not significant (95% CIs were overlapping). Table 2 shows the estimates of the detection probability for the function (f (0)).
Table 2: Estimates of f(0), corresponding 95% confidence intervals (CI), and percentage coefficient of variation (% CV) for the three vegetation types.
95% CI Vegetation type f(0) % CV Lower Upper Cynometra 0.212 0.169 0.268 11.55 Mixed Forest 0.187 0.122 0.287 20.8 Brachystegia 0.137 0.043 0.431 58.99
56
4.4 Effect of covariates on Detection Probability
Included covariates had a different effect on the probability of detecting GRES nest across the three-vegetation types. For instance, a centimetre increase in litter depth was associated with a reduction in the probability of detection by 0.106 and 0.905 in the
Cynometra vegetation type and Brachystegia vegetation but was associated with an increase in the probability of detection by 0.292 in the Mixed Forest Vegetation. Canopy height was the only covariate that had a consistent effect directly on the probability of detection across the three vegetation types where a percentage increase in canopy height was associated with a reduction in the probability of detection (Table 3 and Appendix 2
– Figures 7 to 18).
Table 3: Effect of covariates on the probability of detection
Cynometra Mixed forest Brachystegia Standard Standard Standard Estimate Estimate Estimate error error error Litter -0.106 0.206 0.292 0.300 -0.905 2.010 depth Canopy 0.006 0.007 -0.009 0.017 0.058 0.072 Cover Canopy -0.031 0.033 -0.022 0.058 -0.164 0.140 Height Vegetation 0.001 0.008 -0.010 0.011 0.187 0.363 Density 4.5 Density and Abundance Estimate
Estimates of the number of GRES per square kilometre (density) varied across the three vegetation types with the density being highest in the Mixed Forest vegetation type (57 individuals [95% CI: 29 – 113]) whereas the Brachystegia vegetation type had the least density at 29 individuals [95% CI: 8 – 103] per square kilometre (Table 4).
Overall, the estimated population of GRES across the three vegetation types of ASF was approximately 19,423 individuals. Most individuals were found in Cynometra at about
57
13,160 [95% CI: 8, – 19,939] while the least were estimated to exist in the Brachystegia vegetation type 2,248 [95% CI: 640 – 7,899].
Table 4: Density (D), corresponding 95% confidence interval (CI) and Abundance (N) and its corresponding 95% confidence interval (CI)
Habitat Vegetation Density 95% CI % CV Abundance 95% CI area type (D) for D for D (N) for N (Km2) 8,939 – Cynometra 56.0 38.0 – 82.4 235 13,160 19.4 19,372 Mixed 2,033 – 57.4 29.0 – 13.3 70 4,015 Forest 33.3 7,927 640 – Brachystegia 29.19 8.3 – 102.6 77 2,248 67.4 7,899 Total 382 19,423 4.6 Deadwood Biomass and Abundance
Deadwood volume was on average higher in the Brachystegia vegetation type (about
0.13 m3) compared to 0.12 m3 and 0.09 m3 in the mixed forest and Cynometra respectively. Table 5 shows the output from ordinary least square (OLS) regression models between deadwood volume (cubic meters) and the number of nest sightings in each transect (abundance) by vegetation type and overall. Although there was no evidence of an association between deadwood volume and abundance – all coefficients have P-values >0.05 – abundance was positively associated with deadwood volume only in the Cynometra vegetation type where a cubic meter increase in deadwood volume was associated with a nearly double increase in abundance (1.96 [95% CI: -3.71 – 7.64; P- value=0.48]).
58
Table 5: Ordinary Least Square (OLS) Regression between deadwood volume (M3) and the number of nest sightings by vegetation type and overall, across the three vegetation types.
Cynometra Coefficients P-value Lower 95% Upper 95%
Deadwood Volume 1.964 0.48 -3.707 7.636 R Square 0.023 Observations 24 Mixed Forest Deadwood Volume -3.096 0.649 -18.511 12.319 R Square 0.031 Observations 9 Brachystegia Deadwood Volume -3.112 0.432 -11.673 5.448 R Square 0.07 Observations 11 Overall Model Deadwood Volume -0.699 0.734 -4.826 3.428 R Square 0.003 Observations 44
4.7 Determinants of Nesting Site and nest composition
4.7.1 Effect of Canopy cover on nesting site
There was variation in the canopy cover where nests were found across the three vegetation types. Canopy cover was estimated to be on average highest in the Mixed
Forest vegetation (64.97% [95% CI 58.69 – 71.26]) compared to the Brachystegia Forest
(55.74% [95% CI 49.14 – 62.33], p-value=0.0503) and lowest in the Cynometra forest
(47.64% [95% CI 43.51 – 51.77], p-value<0.001).
59
4.7.2 Effect of Litter Depth on nesting site
Nests were generally found in areas with litter depths of nearly 1.41 cm (95% CI 1.26 –
1.55) on average. Although no significant difference was observed in litter depth comparisons across the three vegetation types, nests found in Brachystegia Forest had marginally higher litter depth at 1.67 cm (95% CI 1.35 – 1.97) while that in the mixed forest was the least at about 1.26 cm (95% CI 0.90 – 1.62).
4.7.3 Effect of Vegetation Density on nesting site
Vegetation density was significantly higher in the nests found in the Cynometra forest
(21.77% [95% CI 16.41 – 27.13]) compared to Mixed forest (10.07% [95% CI 2.65 –
17.48], p-value= 0.015) and Brachystegia forest (9.84% [95% CI 5.07 – 14.62], p- value=0.010).
4.7.4 Effect of Canopy Height on nesting site
Nests were found in areas with varying estimated Canopy heights across the three vegetation types. Canopy height was on average highest among the nests found in the
Brachystegia Forest (16.17 m [95% CI 13.82 – 18.52] followed by that from the mixed forest at 14.50 m [95% CI 12.23 – 16.77, p-value=0.297] and least in the Cynometra forest (9.76 m [95% CI 8.76 – 10.77], p-value<0.001).
4.7.5 Material Composition of nests
Across all vegetation types, over 60% of nests were comprised of both leaves and twigs.
However, some nests were found to be made of twigs only in the Cynometra forest unlike in the other vegetation types (Table 6). A Chi-square test of association between nest
60 material composition and vegetation type showed strong evidence of association (p- value=0.014).
Table 6: Distribution nests’ material composition by vegetation type in Arabuko-Sokoke Forest
Mixed Cynometra Brachystegia Overall forest n (%) n (%) n (%) Material Composition n (%) Twigs only 5 (8) 0 (0) 0 (0) 5 (4) Leaves only 7 (11) 9 (33) 9 (38) 25 (22) Leaves and Twigs 50 (81) 18 (67) 15 (62) 83 (74)
61
CHAPTER 5: DISCUSSION
5.1 Introduction
This study provides abundance estimates from September to October 2019. It also builds up to information that is available regarding Golden-Rumped Elephant-shrew IUCN Red
List (2015) in which there were no population estimates for GRES since 2008 (Ngaruiya,
2009a).
5.2 Abundance and Trends of GRES in Arabuko-Sokoke Forest
The total population of GRES in ASF was found to be 19,423 individuals which is an increase from the previous study by Ngaruiya (2009) in which the total population in the study area was 12,750. Cynometra Forest still had the highest number of individuals and mixed forest had the highest density (about 57.4) similar to findings reported by
(Ngaruiya, 2009a). GRES in Cynometra, Mixed Forest and Brachystegia had increased by about 43%, 93% and 53% respectively. Table 7 below shows the specific numbers per vegetation type between 2019 (current study) and 2008;
Table 7: Comparisons of GRES Abundance between 2019 and 2008
Vegetation Type Current Study (2019) Ngaruiya, 2009
Cynometra Forest 13,160 9, 196
Mixed Forest 4,015 2, 080
Brachystegia Forest 2,248 1,474
62
Although declines in GRES abundance of about 30% and 9% were reported between
1993 – 2000 and 2000-2008 respectively (Bauer, 2001, Ngaruiya, 2009a), this study finds
GRES population to have increased by nearly 52% since 2008.
The high number of GRES individuals across all vegetation types especially in the
Cynometra habitat can mainly be attributed to the habitat characteristics since the
Cynometra habitat type is very dense with small diameter trees hence the GRES nests are not easily visible by predators. This finding corroborates the finding that vegetation density was higher among nests found in Cynometra forest. It is expected that a species selection of a given microhabitat is dependent on its ability to coexist, use that habitat and ultimately survive (Jorgensen, 2004).
On the other hand, Brachystegia Forest is very open with trees that are tall and with large diameter making it possible for predators to easily see GRES hence the low number even though there is an increase. Although GRES is not usually targeted by trappers since they are difficult to catch and have been associated with unpleasant taste they are caught in traps, usually snares that are meant for other animals (FitzGibbon, 1994).
The increase in the numbers may reflect the conservation measures established such as participatory forest management (Matiku et al., 2013), fencing of the forest in 2006/07 as well as livelihood improvement measures such as the Kipepeo project (Okeyo, 2013).
However, as indicated in other studies abundance is not a good indicator of habitat quality since there are other factors that can sustain the numbers such as immigration (Weldy et al., 2019).
(FitzGibbon, 1994) noted that GRES did occur in some scrub and degraded woodland habitats although at low densities. From personal interviews GRES was also found to be existing in disturbed areas such as Dakatcha woodlands. Additionally, GRES nests were
63 also observed near the forest edge which is associated with more logging and cut stumps
(Bauer, 2001). (Yarnell et al., 2008) also observed that the short-snouted sengi could adjust to a new environment after their habitats have been destroyed by fire. Specifically, they reported that the destruction of grasslands mainly used by elephant shrews as habitat resulted in a shift to the use of thickets than grass.
Likewise, a study conducted on small mammals in the Amazon showed increased abundance with habit features indicative of disturbed areas and this pattern was associated with increased resource abundances in these areas with insect biomass and number of fruiting trees showing similar relationships (Lambert et al., 2006).
The variation in population estimates could also be attributed to methodological differences from previous studies in which the detection distances on either side of the transects were limited to 3m and more than one observer were involved in searching of the nests within the identified area (Bauer, 2001, FitzGibbon, 1994, Ngaruiya, 2009a).
This therefore does not meet the assumptions in distance sampling which this study met.
In distance sampling truncation of data is allowed for analysis purposes in which some distances beyond a certain point are discarded (Buckland, 1993). Also, it has been noted that precision was better for untruncated data for the hazard-rate model (Buckland, 1993).
In addition, the study by (Bauer, 2001) did not focus on the abundance but instead on the impact of commercial and subsistence practices using GRES as an indicator species.
5.3 Deadwood Biomass and Nesting Sites
There was no evidence of an association between deadwood volume and abundance.
Deadwood was highest in Brachystegia which had the lowest number of GRES individuals and least in Cynometra which recorded the highest numbers of GRES individuals. This suggests that so long as GRES was within a forested habitat, availability
64 of dead wood did not have an impact on nest location and consequently its presence although further research should be conducted to confirm this. (Lambert et al., 2006) reported that species abundance had a negative relationship with the mean log size, number of logs and the volume of downed wood. This is in line with a study in which potential conservation threats such as grazing, stone collection, litter collection and timber extraction, fire, feral dog and road were observed in captured sites of small mammals hence showing that small mammals respond positively to anthropogenic disturbances (Norbu et al., 2017). A study in the coniferous forest also indicated different responses to coarse woody debris(CWD) by different species where by a strong negative association was recorded between CWD and the abundance if golden-malted ground squirrels, yellow-pine and long-eared chipmunks, and a strong positive association with deer mice (Sollmann et al., 2015). This contradicts a study in which the overall abundance of small mammals was positively correlated with the cover of logs (coarse woody debris) (Ecke et al., 2001).
5.4 Factors Influencing Nesting Site
Findings from this study indicated that vegetation density, litter depth, and canopy cover were higher in areas where nests were found. This could be attributed to several reasons.
First, leaf litter is a main component of nests thus GRES are more likely to construct nests where leaves are in abundance. Second, in another study, canopy cover, and leaf litter (litter depth) reduction were shown to reduce the GRES population as these provided cover against likely predators (Rathbun and Kyalo, 2000). Furthermore, leaves and twigs form a larger composition of litter in ASF and it was no surprise that these were the majority materials used for nest construction. According to (Sponchiado et al.,
2012) most species richness and abundance variability were associated with environments that offer more resources that is, food, water and shelter as was evidenced
65 by the study on rodents. Small mammals are also sensitive to habitat structure (Novillo et al., 2017).
66
CHAPTER 6: CONCLUSION AND RECOMMENDATIONS
6.1 Conclusion
In conclusion, first, the population of GRES in ASF has increased. The increase could either be attributed to conservation measures that have been implemented that discourage human activities within the forest or it could indicate that GRES is adapting to changes that are occurring within the forest. Second, no significant association between deadwood and nest sightings was found. Last, higher canopy cover, litter depth, and vegetation density were associated with nest sightings.
6.2 Recommendation
Consequently, there is a need to maintain already existing conservation measures such as the KIPEPEO project which improves the community’s livelihoods whilst implementing other measures to discourage human activities within the forest which threatens flora and fauna including GRES with extinction. Furthermore, forest management teams
(community members and forest guards) should periodically conduct forest patrols to monitor logging which reduces vegetation density and canopy cover that is associated with GRES nesting sites. Finally, further studies should be conducted to understand if there exist patterns of habitat selection and to continuously monitor the abundance of
GRES to inform proper management of the species. Besides future studies should look at the variation in and abundance of invertebrates that are consumed by GRES. With changing deadwood volume within habitats, there is a chance that the abundance of invertebrates is also influenced, consequently impacting food availability of GRES and other mammals/animals. Further, this should be linked to the examination of the diet composition of GRES and how this is influenced by seasons and forest type. For instance,
GRES have also been anecdotally reported in Dakatcha Forest but no comparisons have
67 been made with regards to habitat preference between the two forests as well as the food preferences.
68
REFERENCES
ABELSON, P. 1996. Valuing Tropical Forest in Kenya. Project Appraisal and Valuation
of the Environment. Springer.
ADEDEJI, O. 2014. Global climate change. Journal of Geoscience and Environment
Protection, 2, 114.
AKÇAKAYA, H. R., BENNETT, E. L., BROOKS, T. M., GRACE, M. K., HEATH, A.,
HEDGES, S., HILTON‐TAYLOR, C., HOFFMANN, M., KEITH, D. A. &
LONG, B. 2018. Quantifying species recovery and conservation success to
develop an IUCN Green List of Species. Conservation Biology, 32, 1128-1138.
AMIN, R., AGWANDA, B. R., OGWOKA, B. & WACHER, T. 2016. Status and
behavioural ecology of sengis in the Boni-Dodori and Arabuko-Sokoke forests,
Kenya, determined by camera traps. Journal of East African Natural History,
105, 223-235.
ANAGE. 2017. AnAge entry for Rhynchocyon chrysopygus [Online]. Available:
https://genomics.senescence.info/species/entry.php?species=Rhynchocyon_chry
sopygus [Accessed].
ARABUKO-SOKOKE FOREST MANAGEMENT TEAM 2002. Arabuko-Sokoke
strategic forest management plan 2002–2027. Kenya Forest Services Nairobi.
ASFMT 2002. Arabuko-Sokoke Strategic Forest Management Plan 2002–2027.
Arabuko-Sokoke Forest Management Team. Unpublished report by Forest
Department—Ministry of Environment and Natural ….
ASHLEY, C. & ELLIOTT, J. 2003. 'Just wildlife?'or a source of local development? ODI
Natural resource perspectives.
69
AVENANT, N. L. 2003. The use of small mammal community characteristics as an
indicator of ecological disturbance in the Korannaberg Conservancy. ACIAR
MONOGRAPH SERIES, 96, 95-98.
BAKER, T. & CHAO, K. 2009. Manual for coarse woody debris measurement in
RAINFOR plots.
BAUER, C. R. 2001. Impact of commercial and subsistence practices on the Arabuko-
Sokoke Forest in coastal Kenya, using an endemic mammal as an indicator
species. Eastern Kentucky University.
BELLARD, C., LECLERC, C., LEROY, B., BAKKENES, M., VELOZ, S., THUILLER,
W. & COURCHAMP, F. 2014. Vulnerability of biodiversity hotspots to global
change. Global Ecology and Biogeography, 23, 1376-1386.
BROOKS, T. M., MITTERMEIER, R. A., MITTERMEIER, C. G., DA FONSECA, G.
A., RYLANDS, A. B., KONSTANT, W. R., FLICK, P., PILGRIM, J.,
OLDFIELD, S. & MAGIN, G. 2002. Habitat loss and extinction in the hotspots
of biodiversity. Conservation biology, 16, 909-923.
BUCKLAND, S. T., ANDERSON, D. R., BURNHAM, K. P., LAAKE, J. L.,
BORCHERS, D. L. & THOMAS, L. 2004. Advanced distance sampling, Oxford
University Press Oxford.
BUCKLAND, S. T., ANDERSON, D.R., BURNHAM, K.P. AND LAAKE, J.L. 1993.
Distance Sampling: Estimating Abundance of Biological Populations, London,
Chapman and Hall.
BUCKLAND, S. T., REXSTAD, E. A., MARQUES, T. A. & OEDEKOVEN, C. S. 2015.
Distance sampling: methods and applications, Springer.
70
BURNHAM, K. P., ANDERSON, D. R. & LAAKE, J. L. 1980. Estimation of density
from line transect sampling of biological populations. Wildlife monographs, 3-
202.
CAREY, A. B. & HARRINGTON, C. A. 2001. Small mammals in young forests:
implications for management for sustainability. Forest Ecology and
Management, 154, 289-309.
CEBALLOS, G. & EHRLICH, P. R. 2006. Global mammal distributions, biodiversity
hotspots, and conservation. Proceedings of the National Academy of Sciences,
103, 19374-19379.
CHANGE, U. N. F. C. O. C. 2007. Impacts, vulnerabilities and adaptation in developing
countries. UNFCCC Bonn,, Germany.
CHARDONNET, P., CLERS, B. D., FISCHER, J., GERHOLD, R., JORI, F. &
LAMARQUE, F. 2002. The value of wildlife. Revue scientifique et technique-
Office international des épizooties, 21, 15-52.
CHIAWO, D. O., KOMBE, W. N. & CRAIG, A. J. 2018. Conservation and human
livelihoods at the crossroads: Local needs and knowledge in the management of
Arabuko Sokoke Forest. African Journal of Ecology, 56, 351-357.
CHIRA, M. 1993. Ecological study of herpetofauna in the Arabuko-Sokoke and Gede
Coastal forests of Kenya.
COOK, R. P. 2001. Small Mammals, Big Role. 2001 Cape Cod National Seashore.
DE MEO, I., AGNELLI, A. E., GRAZIANI, A., KITIKIDOU, K., LAGOMARSINO,
A., MILIOS, E., RADOGLOU, K. & PALETTO, A. 2017. Deadwood volume
assessment in Calabrian pine (Pinus brutia Ten.) peri-urban forests: Comparison
between two sampling methods. Journal of Sustainable Forestry, 36, 666-686.
71
DUMBACHER, J. P., RATHBUN, G. B., SMIT, H. A. & EISEB, S. J. 2012. Phylogeny
and taxonomy of the round-eared sengis or elephant-shrews, genus Macroscelides
(Mammalia, Afrotheria, Macroscelidea). PLoS One, 7, e32410.
ECKE, F., LÖFGREN, O., HÖRNFELDT, B., EKLUND, U., ERICSSON, P. &
SÖRLIN, D. 2001. Abundance and diversity of small mammals in relation to
structural habitat factors. Ecological Bulletins, 165-171.
EMERTON, L. 1999. Community conservation research in Africa: principles and
comparative practice. Global Environmental Change Programme. Paper, 9.
ERÄJÄÄ, S., HALME, P., KOTIAHO, J. S., MARKKANEN, A. & TOIVANEN, T.
2010. The volume and composition of dead wood on traditional and forest fuel
harvested clear-cuts. Silva Fennica, 44.
EXISTENCE, E. O. 2017. Golden-rumped Sengi [Online]. Available:
http://www.edgeofexistence.org/species/golden-rumped-sengi/. [Accessed].
FILES, T. A. 2014a. Golden-Rumped Elephant Shrew [Online]. Available:
https://www.theanimalfiles.com/mammals/elephant_shrews/golden_rumped_ele
phant_shrew.html. [Accessed].
FILES, T. A. 2014b. Rufous Elephant Shrew [Online]. Available:
https://www.theanimalfiles.com/mammals/elephant_shrews/rufous_elephant_sh
rew.html. [Accessed].
FILHO, A. F., MACHADO, S. A. & CARNEIRO, M. R. A. 2000. Testing accuracy of
log volume calculation procedures against water displacement techniques
(xylometer). Canadian journal of forest research, 30, 990-997.
FITZGIBBON, C. & RATHBUN, G. 2014. Rhynchocyon chrysopygus. The IUCN Red
List of Threatened Species. Version.
72
FITZGIBBON, C. & RATHBUN, G. B. 2015. Rhynchocyon chrysopygus. The IUCN
Red List of Threatened Species 2015: e.T19705A21287265.
https://dx.doi.org/10.2305/IUCN.UK.2015-2.RLTS.T19705A21287265.en.
IUCN.
FITZGIBBON, C. D. 1994. The distribution and abundance of the golden-rumped
elephant-shrew Rhynchocyon chrysopygus in Kenyan coastal forests. Biological
Conservation, 67, 153-160.
FITZGIBBON, C. D. 1995. Comparative ecology of two elephant‐shrew species in a
Kenyan coastal forest. Mammal Review, 25, 19-30.
FITZGIBBON, C. D. & RATHBUN, G. B. 1994. Surveying Rhynchocyon elephant‐
shrews in tropical forest. African Journal of Ecology, 32, 50-57.
FOUNDATION, A. W. 2020. Elephant Shrew [Online]. Available:
https://www.awf.org/wildlife-conservation/elephant-shrew. [Accessed October
27 2020].
GACHANJA, M. & KANYANYA, E. 2004. Conservation of Kenya’s Coastal Forests:
Social Economic Resources’. WWF East African Coastal Forest Ecoregion
Programme, Nairobi, Kenya.
GASIOR, R. 2006. Rhonchocyon cirnei [Online]. Animal Diversity Web. Available:
https://animaldiversity.org/accounts/Rhynchocyon_cirnei/. [Accessed].
GASTON, K. J. 2003. The structure and dynamics of geographic ranges, Oxford
University Press on Demand.
GEREAU, R. E., KARIUKI, M., NDANG'ANG'A, P. K., WEREMA, C. & MUORIA,
P. 2014. Biodiversity status and trends in the Eastern Arc Mountains and coastal
73
forests of Kenya and Tanzania region, 2008-2013, BirdLife International, African
Partnership Secretariat.
GIAM, X. 2017. Global biodiversity loss from tropical deforestation. Proceedings of the
National Academy of Sciences, 114, 5775-5777.
GROOM, M. J. 2006. Threats to biodiversity. Principles of conservation biology, 63,
109.
HABEL, J. C., CASANOVA, I. C. C., ZAMORA, C., TEUCHER, M., HORNETZ, B.,
SHAURI, H., MULWA, R. K. & LENS, L. 2017. East African coastal forest
under pressure. Biodiversity and Conservation, 26, 2751-2758.
HANKS, G. C. J. 1968. A revision of the elephant-shrews, family Macroscelididae.
Bulletin of the British Museum (Natural History) Zoology, 16, 47-111.
HOFFMANN, A., DECHER, J., ROVERO, F., SCHAER, J., VOIGT, C. & WIBBELT,
G. 2010. Field methods and techniques for monitoring mammals. Manual on field
recording techniques and protocols for all taxa biodiversity inventories, 8, 482-
529.
IUCN. 2020. IUCN Definitions - English [Online]. Available:
https://www.iucn.org/sites/dev/files/iucn-glossary-of-definitions_en.pdf.
[Accessed].
JACKSON, M. M. & NAUGHTON-TREVES, L. 2012. Eco-bursaries as incentives for
conservation around Arabuko-Sokoke Forest, Kenya. Environmental
Conservation, 39, 347-356.
JACQUES, A. 2013. Rhynchocyon petersi [Online]. Animal Diversity Web. Available:
https://animaldiversity.org/accounts/Rhynchocyon_petersi/ [Accessed October
06 2020].
74
JANSA, S. 1999. Rhynchocyon chrysopygus [Online]. Animal Diversity Web. Available:
https://animaldiversity.org/accounts/Rhynchocyon_chrysopygus/ [Accessed
October 06 2020].
JONES, J. 2002. Elephantulus myurus [Online]. Animal Diversity Web. Available:
https://animaldiversity.org/accounts/Elephantulus_myurus/ [Accessed October
11 2020].
JORGENSEN, E. E. 2004. Small mammal use of microhabitat reviewed. Journal of
Mammalogy, 85, 531-539.
KIRILENKO, A. P. & SEDJO, R. A. 2007. Climate change impacts on forestry.
Proceedings of the National Academy of Sciences, 104, 19697-19702.
LACHER JR, T. E., DAVIDSON, A. D., FLEMING, T. H., GÓMEZ-RUIZ, E. P.,
MCCRACKEN, G. F., OWEN-SMITH, N., PERES, C. A. & VANDER WALL,
S. B. 2019. The functional roles of mammals in ecosystems. Journal of
Mammalogy, 100, 942-964.
LAMBERT, T. D., MALCOLM, J. R. & ZIMMERMAN, B. L. 2006. Amazonian small
mammal abundances in relation to habitat structure and resource abundance.
Journal of Mammalogy, 87, 766-776.
LANS, W. & VAN DER VOORDT, D. 2002. Descriptive research. Ways to study and
research urban, architectural and technical design.
LAURANCE, W. F. & PERES, C. A. 2006. Emerging threats to tropical forests,
University of Chicago Press.
LOMOLINO, M., RIDDLE, B., WHITTAKER, R. & BROWN, J. 2010. Biogeography
(Sinauer, Sunderland, MA).
LUSWETI, A. 2011. Biodiversity Conservation in Kenya: Institute of Economic Affairs.
75
LUZA, A. L., GRAHAM, C. H. & HARTZ, S. M. 2019. A global database on non-volant
small mammal composition in natural and human-modified habitats. Data in
Brief, 23, 103842.
MARQUES, T. A., THOMAS, L., FANCY, S. G. & BUCKLAND, S. T. 2007.
Improving estimates of bird density using multiple-covariate distance sampling.
The Auk, 124, 1229-1243.
MATIKU, P. 2002. The Coastal Forests of Kenya. A national synthesis report for the
development of the WWF-EARPO Eastern Africa Coastal Forests Ecoregion
Programme, 1-4.
MATIKU, P., CALEB, M. & CALLISTUS, O. 2013. The impact of participatory forest
management on local community livelihoods in the Arabuko-Sokoke forest,
Kenya. Conservation and Society, 11, 112-129.
MULWA, R., & MUTUNGA, J. 2017. Birds of Arabuko Sokoke Forest. Research Gate,
34-39.
MUSILA, S., CHEN, Z.-Z., LI, Q., YEGO, R., ZHANG, B., ONDITI, K., MUTHONI,
I., HE, S.-W., OMONDI, S. & MATHENGE, J. 2019. Diversity and distribution
patterns of non-volant small mammals along different elevation gradients on Mt.
Kenya, Kenya. Zoological research, 40, 53.
NDALILO, L., MBUVI, M. & LUVANDA, A. 2017. Utilization and Governance of
Arabuko Sokoke Forest. Biodiversity Status of Arabuko Sokoke Forest, 47-59.
NDANG'ANG'A, P. K., BARASA, F. M., KARIUKI, M. N. & MUORIA, P. 2016.
Trends in forest condition, threats and conservation action as derived from
participatory monitoring in coastal Kenya. African Journal of Ecology, 54, 76-
86.
76
NGARUIYA, G. 2009a. Assessment of the Range and Population of Golden-rumped
Elephant-shrew (Rhynchocyon chrysopygus) in the Northern Coastal forests of
Kenya. Biological Sciences, University of Nairobi. Unpublished MA Thesis.
NGARUIYA, G. W. 2009b. Assessment of the range and population of golden rumped
elephant-shrew (Rhynchocyon Chrysopygus) In The Northern Coastal Forests Of
Kenya. University of NAIROBI.
NICOLL, M. & RATHBUN, G. 1990. African Insectivora and elephant-shrews: an
action plan for their conservation: IUCN. Gland, Switzerland.
NORBU, L., THINLEY, P., PHURPA, U. D. & TSHERING, P. 2017. Diversity and
seasonal abundance of small mammals in Bumdeling Ramsar Site, Trashiyangtse,
Eastern Bhutan.
NOVILLO, A., CUEVAS, M. F., OJEDA, A. A., OVEJERO, R. J., TORRES, M.,
EUGENIA, M. & OJEDA, R. A. 2017. Habitat selection and coexistence in small
mammals of the southern Andean foothills (Argentina). Mammal Research, 62,
219-227.
NYAMACHE, J., & MALONZA, P. K. 2017. Amphibians and Reptiles of Arabuko.
Research Gate, 41-45.
NYAMACHE, J. & MALONZA, P. K. 2017. Amphibians and Reptiles of Arabuko.
Biodiversity Status of Arabuko Sokoke Forest, Kenya. Kenya Forestry Research
Institute.
NYUNJA, J., OGWOKA, B., & NJERI, L. 2017. Composition and Distribution of Small
and Large Mammals in Arabuko Sokoke Forest. Biodiversity Status of Arabuko
Sokoke Forest, Kenya.
77
OGUGE, N., HUTTERER, R., ODHIAMBO, R. & VERHEYEN, W. 2004. Diversity
and structure of shrew communities in montane forests of southeast Kenya.
Mammalian Biology, 69, 289-301.
OKELLO, M. M. & GRASTY, K. 2009. The role of large mammals and protected areas
to tourist satisfaction in the Northern Circuit, Tanzania. Tourism Analysis, 14,
691-697.
OKEYO, R. O. 2013. IMPACTS OF KIPEPEO CONSERVATION PROJECT ON
LIVELIHOODS AND CLIMATE CHANGE MITIGATIONS AMONG ARABUKO
SOKOKE COMMUNITIES, KILIFI COUNTY, KENYA. Kenyatta University.
OYUGI, J. O., BROWN, J. S. & WHELAN, C. J. 2008. Effects of human disturbance on
composition and structure of Brachystegia woodland in Arabuko‐Sokoke Forest,
Kenya. African Journal of Ecology, 46, 374-383.
PELLIKKA, P. & CLARK, B. 2004. Taita Hills and Kenya, 2004: Seminars, Reports
and Journal of a Field Expedition to Kenya, University of Helsinki, Department
of Geography.
PILLAI, R. 2011. Benefit of tourism to local community: Result or residue. Indian J. of
Tourism & Hospitality Management, 2, 149-161.
RAJPAR, M. N. 2018. Tropical Forests Are An Ideal Habitat for Wide Array of Wildlife
Species. Tropical Forests: New Edition, 37.
RANDHIR, T. O. & EROL, A. 2013. Emerging Threats to Forests: Resilience and
Strategies at System Scale.
RATHBUN, G. & KYALO, S. 2000. Golden-rumped elephant-shrew.
RATHBUN, G. B. 1979. The social structure and ecology of elephant-shrews.
78
RATHBUN, G. B. 1995. Conservation issues and strategies for elephant‐shrews.
Mammal Review, 25, 79-86.
RATHBUN, G. B. 2009. Why is there discordant diversity in sengi (Mammalia:
Afrotheria: Macroscelidea) taxonomy and ecology? African Journal of Ecology,
47, 1-13.
RATHBUN, G. B. 2020. Sengis (Elephant-Shrews) [Online]. Available:
https://www.sengis.org/synopsis.php. [Accessed October 27 2020].
ROMER, A. S. 1972. The vertebrate as a dual animal—somatic and visceral.
Evolutionary biology. Springer.
ROVERO, F., COLLETT, L., RICCI, S., MARTIN, E. & SPITALE, D. 2013.
Distribution, occupancy, and habitat associations of the gray-faced sengi
(Rhynchocyon udzungwensis) as revealed by camera traps. Journal of
mammalogy, 94, 792-800.
ROVERO, F., RATHBUN, G., PERKIN, A., JONES, T., RIBBLE, D. O., LEONARD,
C., MWAKISOMA, R. & DOGGART, N. 2008. A new species of giant sengi or
elephant‐shrew (genus Rhynchocyon) highlights the exceptional biodiversity of
the Udzungwa Mountains of Tanzania. Journal of Zoology, 274, 126-133.
ROWE, R. J. & TERRY, R. C. 2014. Small mammal responses to environmental change:
integrating past and present dynamics. Journal of Mammalogy, 95, 1157-1174.
SABUNI, C., BEDDETTI, A. & LEIRS, H. 2011. Can giant sengis (genus Rhynchocyon)
be captured with baited traps. Afrotherian Conservation-Newsletter of the IUCN-
SCC Afrotheria Specialist Group, 8, 9-12.
79
SCHUBERT, M. 2011. A summary of the social system of the round-eared sengi.
Afrotherian Conservation—Newsletter of the IUCN-SCC Afrotheria Specialist
Group, 8, 12-13.
SEBER, G. A. & LEE, A. J. 2012. Linear regression analysis, John Wiley & Sons.
SINDIGA, I. 1995. Wildlife based tourism in Kenya: Land use conflicts and government
compensation policies over protected areas. Journal of Tourism Studies, 6, 45.
SKINNER, J., & CHIMIMBA, C. 2005. The Mammals of the South African Sub-Region
Family Macroscelididae. Cambridge University Press.
SMIT, H., JANSEN VAN VUUREN, B., O'BRIEN, P., FERGUSON‐SMITH, M.,
YANG, F. & ROBINSON, T. 2011. Phylogenetic relationships of elephant‐
shrews (Afrotheria, Macroscelididae). Journal of Zoology, 284, 133-143.
SOLLMANN, R., WHITE, A. M., GARDNER, B. & MANLEY, P. N. 2015.
Investigating the effects of forest structure on the small mammal community in
frequent-fire coniferous forests using capture-recapture models for stratified
populations. Mammalian Biology, 80, 247-254.
SONGORWA, A. N., BÜHRS, T. & HUGHEY, K. F. 2000. Community-based wildlife
management in Africa: a critical assessment of the literature. Natural Resources
Journal, 603-643.
SPONCHIADO, J., MELO, G. & CÁCERES, N. 2012. Habitat selection by small
mammals in Brazilian Pampas biome. Journal of Natural History, 46, 1321-1335.
STANLEY, W. T., ROGERS, M. A., KIHAULE, P. M. & MUNISSI, M. J. 2014.
Elevational distribution and ecology of small mammals on Africa’s highest
mountain. PloS one, 9, e109904.
80
STEVENS, T. H., ECHEVERRIA, J., GLASS, R. J., HAGER, T. & MORE, T. A. 1991.
Measuring the existence value of wildlife: what do CVM estimates really show?
Land Economics, 67, 390-400.
TARUS, G. K. 2018. Projecting the impacts of climate change on tree biomass in
Arabuko-Sokoke Forest, Kenya. Journal of Climate, 1.
TEJASWI, P. B. G. 2006. MANUAL ON DEFORESTATION, DEGRADATION, AND
FRAGMENTATION USING REMOTE SENSING AND GIS.
THOMAS, L., BUCKLAND, S. T., REXSTAD, E. A., LAAKE, J. L., STRINDBERG,
S., HEDLEY, S. L., BISHOP, J. R., MARQUES, T. A. & BURNHAM, K. P.
2010. Distance software: design and analysis of distance sampling surveys for
estimating population size. Journal of Applied Ecology, 47, 5-14.
TISDELL, C. 2005. Economics, Ecology and the Environment. Working paper.
TRUST, W. L. 2020. Golden-rumped elephant shrew [Online]. Available:
https://www.worldlandtrust.org/species/mammals/golden-rumped-elephant-
shrew/. [Accessed].
UNITED NATIONS, U. 2017. The sustainable development goals report.
WATCH, G. F. 2002. Global forest watch. World Resources Institute, Washington, DC
Available from http://www. globalforestwatch. org (accessed March 2002).
WATERS, J. J., JACKSON, C. & JACKSON, R. G. 2007. Forest Cover Survey Abuko-
Sokoke Forest, Kenya. A Rocha Kenya Occassional Report.
WEKESA, C. 2017. Biodiversity Status of Arabuko Sokoke Forest, Kenya.
81
WELDY, M. J., EPPS, C. W., LESMEISTER, D. B., MANNING, T., LINNELL, M. A.
& FORSMAN, E. D. 2019. Abundance and ecological associations of small
mammals. The Journal of Wildlife Management, 83, 902-915.
YARNELL, R. W., METCALFE, D. J., DUNSTONE, N., BURNSIDE, N. & SCOTT,
D. M. 2008. The impact of fire on habitat use by the short-snouted elephant shrew
(Elephantulus brachyrhynchus) in North West Province, South Africa. African
Zoology, 43, 45-52.
82
APPENDICES
Appendix 1: Data collection tool
RESEARCH ON GRES IN ARABUKO-SOKOKE FOREST Data Collection Form Data Collectors: Date: Transect Details:
82
Transect Name/Number: ………………………………………………………… Length (m): …………………………………………… Start Coordinates: ………………………………………End Coordinates: ………………………………………………………………… Vegetation Type: Cynometra Forest ( ) Brachystegia Forest ( ) Mixed Forest ( ) Plantation ( ) Nest Details:
NO. GPS Co- Distance Material Composition Nest Age Litter Canopy Canopy Vegetation Proximate tree ordinates from (Based on visual depth Cover Height Density species>5cm transect observation) (cm) (%) (m) Chequerboard diameter 2m 1m (%) radius from the Twigs Leaves Twigs New Old from nest Only Only and nest Leaves
83
NO. GPS Co- Distance Material Composition Nest Age Litter Canopy Canopy Vegetation Proximate tree ordinates from (Based on visual depth Cover Height Density species>5cm transect observation) (cm) (%) (m) Chequerboard diameter 2m 1m (%) radius from the Twigs Leaves Twigs New Old from nest Only Only and nest Leaves
8
3
84
Dead Wood Details: Degree of decay Current State 1. GL = green leaves (very fresh) 1. Standing 2. BL = brown leaves (fresh) 2. On the ground 3. SB = small branches remain (new) 3. Leaning/suspended 4. LB = large branches remain (old) 5. R = rotting Standing
Sample Piece No Current Mode of death Diameter 1 Diameter 2 Length of piece Degree in decay Plot No. State
Natural Cut Elephant
8
4
85
Appendix 2: Effect of Covariates on detection probability by vegetation type a) Litter Depth
8
5
Figure 9: Effect of Litter depth on the probability of detection in Cynometra forest
86
8
6
Figure 10: Effect of Litter depth on the probability of detection in the Mixed Forest vegetation Type
87
8
7
Figure 11: Effect of litter depth on the probability of detection in the Brachystegia Forest
88
b) Canopy Cover
8
8
Figure 12: Effect of Canopy Cover on the probability of detection in Cynometra Forest
89
8
9
Figure 13: Effect of Canopy Cover on the probability of detection in Mixed forest
90
90
Figure 14: Effect of Canopy Cover on the probability of detection in Brachystegia forest
91
c) Canopy Height
91
Figure 15: Effect of Canopy Height on the probability of detection in Cynometra forest
92
92
Figure 16: Effect of Canopy Height on the probability of detection in Mixed Forest
93
93
Figure 17: Effect of Canopy Height on the probability of detection in Brachystegia Forest
94
d) Vegetation Density
94
94
Figure 18: Effect of Vegetation Density on the probability of detection in Cynometra Forest
95
95
Figure 19: Effect of Vegetation Density on the probability of detection in Mixed Forest
96
96
Figure 20: Effect of Vegetation Density on the probability of detection in Brachystegia Forest