<%BANNER%>

Spatial Dynamics of Elephant Impacts on Trees in Chobe National Park, Botswana

Permanent Link: http://ufdc.ufl.edu/UFE0024518/00001

Material Information

Title: Spatial Dynamics of Elephant Impacts on Trees in Chobe National Park, Botswana
Physical Description: 1 online resource (33 p.)
Language: english
Creator: Fullman, Timothy
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: africa, africana, botswana, chobe, distance, effect, elephant, herbivory, loxodonta, national, park, piosphere, river, tree, water
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: SPATIAL DYNAMICS OF ELEPHANT IMPACTS ON TREES IN CHOBE NATIONAL PARK, BOTSWANA Timothy Jon Fullman 714-381-5337 School of Natural Resources and Environment Supervisory chair: Brian Child Master of Science May 2009 Management of the world's largest population of African elephants is under debate in Chobe National Park, Botswana. While the variety of wildlife in Chobe serves as the basis for a booming tourism industry, generating significant jobs and revenues for local communities, the estimated 200,000 elephants are also a source of human-wildlife conflict, raiding crops and sometimes killing people and livestock. There are also concerns that elephants are causing habitat changes and affecting other wildlife, threatening the area's ecological integrity and sustainability of tourism. This project quantifies the impacts of elephants on trees and how these patterns relate to distance from water. Elephant utilization patterns may extend farther than previously considered, suggesting a need for further monitoring to manage shifts in tree species composition in areas far from the river. This information will help improve management decisions in an area where biodiversity conservation, economic growth, and local livelihood concerns are interdependent.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Timothy Fullman.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Child, Brian.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024518:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024518/00001

Material Information

Title: Spatial Dynamics of Elephant Impacts on Trees in Chobe National Park, Botswana
Physical Description: 1 online resource (33 p.)
Language: english
Creator: Fullman, Timothy
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: africa, africana, botswana, chobe, distance, effect, elephant, herbivory, loxodonta, national, park, piosphere, river, tree, water
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: SPATIAL DYNAMICS OF ELEPHANT IMPACTS ON TREES IN CHOBE NATIONAL PARK, BOTSWANA Timothy Jon Fullman 714-381-5337 School of Natural Resources and Environment Supervisory chair: Brian Child Master of Science May 2009 Management of the world's largest population of African elephants is under debate in Chobe National Park, Botswana. While the variety of wildlife in Chobe serves as the basis for a booming tourism industry, generating significant jobs and revenues for local communities, the estimated 200,000 elephants are also a source of human-wildlife conflict, raiding crops and sometimes killing people and livestock. There are also concerns that elephants are causing habitat changes and affecting other wildlife, threatening the area's ecological integrity and sustainability of tourism. This project quantifies the impacts of elephants on trees and how these patterns relate to distance from water. Elephant utilization patterns may extend farther than previously considered, suggesting a need for further monitoring to manage shifts in tree species composition in areas far from the river. This information will help improve management decisions in an area where biodiversity conservation, economic growth, and local livelihood concerns are interdependent.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Timothy Fullman.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Child, Brian.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024518:00001


This item has the following downloads:


Full Text

PAGE 1

1 SPATIAL DYNAMICS OF ELEPH ANT IMPACTS ON TREES IN CHOBE NATIONAL PARK, BOTSWANA By TIMOTHY JON FULLMAN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Timothy Jon Fullman

PAGE 3

3 To my parents, who have encouraged my love of wildlife and enabled me to be where I am today, and to my Lord, who ma de all the animals I love

PAGE 4

4 ACKNOWLEDGMENTS I thank Alecia Brantely, Cerian Gibbes, and Njojo for helping with fieldwork. This project would not have been possible without the assistan ce of the Botswana Departm ent of Wildlife and National Parks and its game scouts. I would espe cially like to thank Mogami, my primary field scout for his hard work and help in the field and to Graham Child for logistical support and training. I thank Brian Child, Mickie Swisher, and Todd Palmer for advice regarding research design and methodology. Amy Cantrell and Megha n Brennan provided st atistical advice. Generous financial support was given by the Cl eveland Metroparks Zoo, the Center for Tropical Conservation and Development at the University of Florida, and Idea Wild. I thank my friends and family who have encouraged my love for animals, endured my excitement in describing what I have learned, and spent many hours review ing proposals. Finally, I thank my wonderful Lord, who has given me my passion for wildlife.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................6LIST OF FIGURES .........................................................................................................................7ABSTRACT ...................................................................................................................... ...............8 CHAPTER 1 INTRODUCTION .................................................................................................................. ..92 MATERIAL AND METHODS ..............................................................................................11Study Area ..............................................................................................................................11Elephant Impacts on Vegetation .............................................................................................12Relative Dung Density ............................................................................................................13Analysis ..................................................................................................................................133 RESULTS ....................................................................................................................... ........164 Discussion .................................................................................................................... ...........24Large-Scale Trends ............................................................................................................ .....24Small-Scale Trends ............................................................................................................ .....26Management Implications ......................................................................................................27REFERENCES .................................................................................................................... ..........30BIOGRAPHICAL SKETCH .........................................................................................................33

PAGE 6

6 LIST OF TABLES Table page 3-1 Descriptive statistics for the m ost common tree species ...................................................19

PAGE 7

7 LIST OF FIGURES Figure page 2-1 Map of Chobe National Park, Botswana including study area. .........................................153-1 Predictor variables by distance from the Chobe River.. ....................................................203-2 Predictor variables by di stance from all water. ..................................................................213-3 Mean utilization at different distances from the Chobe River. ..........................................223-4 Mean utilization at distances from all water in Chobe National Park, Botswana. ............. 23

PAGE 8

8 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SPATIAL DYNAMICS OF ELEPH ANT IMPACTS ON TREES IN CHOBE NATIONAL PARK, BOTSWANA By Timothy Jon Fullman May 2009 Chair: Brian Child Major: Interdisciplinary Ecology The African elephant ( Loxodonta africana) regulates shifts between different savanna states, primarily through herbivory of woody vege tation. As a water-depende nt herbivore, these impacts on trees are constrained by water availab ility, potentially leadin g to a gradient of degradation known as the piosphere effect. Transects ev aluating vegetation stat us with increasing distance from the Chobe River were conduc ted in Chobe National Park, Botswana, to test whether predictions of the piosphere effect can be applied at multiple scales. Trends varied depending on the type of utilization, with deba rking by elephants decreas ing with distance from the Chobe River and branch herbivory showing a bimodal distribution. Results suggest that piosphere predictions may be applicable over grea ter distances, with important implications for monitoring species changes far from water points. Managers should consider this as they evaluate landscape stability and discuss provisi oning of waterpoints in semi-arid habitats.

PAGE 9

9 CHAPTER 1 INTRODUCTION Herbivory affects the structure and dynam i cs of African savannas (Skarpe, 1992; van Langevelde et al. 2003). Understanding the impacts large mammals have on savannas is critical for management of these systems. The African elephant ( Loxodonta africana (Blumenbach)) has long been seen as one of the principle drivers re gulating shifts between di fferent savanna states (Laws, 1970; Dublin, Sinclair, & McGlade, 1990; Augustine & Mc Naughton, 2004). Often described as a keystone species and ecosyst em engineer (Jones, Lawton, & Shachak, 1994), elephants can have a profound influence on woody vegetation, browsing trees and shrubs, and can impede woodland formation (Pellew, 1983; Holdo, 2007). At larger spatial scales, herbi vore distribution is determined primarily by abiotic factors, such as distance to water, with smaller-scale pr ocesses such as herbivory operating within this framework (Bailey et al. 1996). Concentration of large herbivores near waterpoints may lead to development of a gradient of de gradation that increases with proximity to water, termed the piosphere (Lange, 1969). Th is phenomenon is well documented in Kruger National Park (Thrash et al. 1991; Thrash, 1998; Brits, van Rooyen, & van Rooyen, 2002) and in other arid and semi-arid systems across Africa (e.g., Child, Parris, & Rich, 1971). As a water-dependent species (Redfern et al ., 2003), elephant distri bution is regulated by water availability (Chamaill-J ammes, Valeix, & Fritz, 2007). Th is constraint, along with the tremendous impacts allowed by their body size, creates great potential for el ephants to contribute to the piosphere effect. Indeed, increased utilization of vegetati on around water sources by elephants has been seen in Etosha National Park, Namibia (de Beer et al. 2006) and forest reserves in Tanzania (Afolayan, 1975). Increased understanding of the spatial heterogeneity of elephant impacts on vegetation is needed (Valeix et al. 2007).

PAGE 10

10 Chobe National Park (Chobe), Botswana provides an excellent case to te st piosphere effect theory as it relates to elephants. Botswana is home to the largest known elephant population on the continent with over 150,000 individuals (Blanc et al. 2007). Mosugelo et al. (2002) found wooded areas along the Chobe riverfront have de creased over time. Congruent with piosphere effect predictions, their study f ound decreasing elephant browsing with increasing distance from the river. Previous studies have been conducte d within 10 km inland from the river (Mosugelo et al. 2002; Nellemann, Moe, & Rutina, 2002), but did not evaluate the rest of the park, which extends over 50 km further to the south. This study expands evaluation of spatial dynami cs of elephant impacts on vegetation from the riverfront to the south-eastern border of the park. Vegetation transects evaluated tree utilization by elephants and fire at multiple scales to see whether a traditional piosphere effect is maintained or if elephant utiliz ation of trees is bimodal as ele phants drink and browse near the river and then walk inland to browse again. My results are discussed within the framework of elephant management and the influences brow se patterns might have on the vegetation and wildlife of Chobe and other semi-arid protected areas.

PAGE 11

11 CHAPTER 2 MATERIAL AND METHODS Study Area This study was conducted in Chobe National Park in the northeastern corner of Botswana from May to July 2008 (Figure 2-1). Rainfall averages 600-700 mm per year and occurs primarily between November and March (Chil d, 1968). Average temperatures range between 15.2C 30.2C (Child, 1968). The Chobe River borders the northern part of the park, forming the boundary between Botswana and the Caprivi Strip in Namibia. Ephemeral pans across the park fill during the wet season and slowly disappear as the dry season progresses, providing additional water sources. Several artificial borehole waterpoints have also been established by the Botswana Department of Wildlife and Natio nal Parks near the south of the study area. Vegetation along the riverfront is primarily scrub dominated by Croton megalobotrys (Mll.Arg.), Capparis tomentosa (Lam.), and Combretum mossambicense (Engl.; Herremans, 1995). About 1-2 km from the river the vege tation changes to a shrub-woodland mixture dominated by Baikiaea plurijuga (Harms) and including Burkea africana (Hook.), Croton gratissimus (Burch.), Combretum elaeagnoides (Klotzsch), Baphia massaiensis (Taub.), and Terminalia sericea (Burch ex DC.; Mosugelo et al. 2002; for a more detailed description of vegetation see Simpson, 1975). Farther inland, the vegetation type changes to a mixed Kalahari savanna woodland with domi nant species including Colophospermum mopane (J.Kirk ex Benth.), Combretum apiculatum (Sond.), Burkea africana and Combretum hereroense (Schinz). Child (1968) reported 38 mammal species jack al-sized or larger in Chobe including elephants, buffalo ( Syncerus caffer (Sparrman)), giraffe (Giraffa camelopardalis (L.)), zebra ( Equus burchelli (Gray)), warthog ( Phacochoerus aethiopicus (Pallas)), hippopotamus ( Hippopotamus amphibious (L.)), kudu ( Tragelaphus strepsiceros (Pallas)), sable ( Hippotragus

PAGE 12

12 niger (Harris)), and impala (Aepyceros melampus (Lichtenstein)). Majo r carnivores include lion ( Panthera leo (L.)), leopard ( Panthera pardus (L.)), spotted hyena (Crocuta crocuta (Erxleben)) and African wild dog ( Lycaon pictus (Temminck)). Elephant Impacts on Vegetation Thirty -four transects investigating vegetati on status with increasing distance from the Chobe River were conducted in Chobe Nationa l Park. Transects were located running roughly south following a one-lane dirt track from the Chobe River to the southeastern border of the park (Figure 2-1). The initial transect location was randomly determined and subsequent transects were conducted systematically every 2.5 km. Dist ance between transects was determined using a GPS receiver (Garmin Rino 120, Garmin Ltd., Olathe KS, USA). Two additional transects were conducted farther east along the riverfront to increas e sample size within the thin riparian strip bordering the river. All transects were established at least 50 m from tracks to minimize any track-based effects on vegeta tion or browsing (Mosugelo et al. 2002). A line 100 m long was marked off parallel to the track to define the start point. Heading away from the track, transects were continued until 50 trees were recorded or, in areas of very low tree density, a one hectare (10,000 m2) area was surveyed. Any plant greater than 3 m tall was classified as a tree (Walker, 1976). Each tree was identified to species leve l and spatially georeferenced using GPS. Characteristics including height, DBH (diameter at breast height), number of stems, and percent green vegetation in the crown were recorded, as well as whether the tree was alive or not and the presence of ground and aerial coppicing. Elephant utilization of trees occurred in tw o primary forms, debarking and damage to branches/trunks. Ringbarking occurs when bark is completely removed from a strip spanning the trees circumference and results in death as the tree can no longer tran sport sugars. Debarking herbivory was evaluated using six categories: 0 = no bark removal, 1 = 1-20% of circumference

PAGE 13

13 debarked, 2 = 21-40% debarked, 3 = 41-60% debarked, 4 = 6180% debarked, and 5 = 81-100% debarked. Branch herbivory was assessed using five categories: 0 = no utilization, 1 = minor utilization (a few minor branches broken), 2 = moderate utilization (many minor branches broken), 3 = high utilization (main branches br oken), and 4 = main stem utilization (main meristem broken off). Fire damage was also ev aluated using four categories: 0 = no damage, 1 = light damage, 2 = moderate damage, and 3 = heavy damage. Distance of transects to water was determin ed using GIS software (ArcView 9.3, ESRI, Redlands, CA, USA). Water body locations for the Chobe region were obtained from the Botswana Department of Surveys and Mapping in Gaborone and verified against remotely sensed images. Relative Dung Density A dung count was also conducted at each site be ginning at th e start point for the vegetation transect and extending 100 m perpendicular to the track. Mamma lian dung within 5 m on either side of the line was enumerated (for animals that defecate many pellets, one cluster was counted as a single dropping). Dung was identified to species using a guidebook (Stuart & Stuart, 2000) and help from local guides. Dung counts were co nducted to give an estimate of relative animal use of transect sites (Young, Palmer, & Gadd, 2005). Wh ile issues have been raised about the use of dung counts to measure mammal densities (e. g., Fuller, 1991), Barnes (2001) showed that they are as effective as other methods of estimati on for elephants as well as for other vertebrate species. Analysis Data were analyzed us ing SAS software (Ver sion 9.2, SAS Institute Inc., Cary, NC, USA). Linear regression was performed upon utilization categories and potential predictor variables including distance from the Chobe River, distan ce from all water, tree height, tree density,

PAGE 14

14 percent green vegetation in crow n, and relative abundance of ele phant and total dung. Stepwise multiple regression was used to select the best model for changes in utilization based on all potential predictors.

PAGE 15

15 Figure 2-1. Map of Chobe National Pa rk, Botswana including study area.

PAGE 16

16 CHAPTER 3 RESULTS A total of 1600 trees were evaluated. Twenty-five species were identified. Ten species had more than 50 individuals recorded. Combretum apicula tum was the most prevalent (n = 437), followed by Baikiaea plurijuga (n = 166), Terminalia sericea (n = 153), Colophospermum mopane (n = 148), Combretum elaeagnoides (n = 130), Burkea africana (n = 118), Combretum mossambicense (n = 89), Croton megalobotrys (n = 72), Combretum hereroense (n = 61), and Terminalia brachystemma (Welw. Ex Hiern, n = 53; for descri ptive statistics of these species see Table 3-1). Between 32 and 50 i ndividual trees were sampled per transect and species density ranged from two to ten species per transect. To tal utilization of trees by elephants was not correlated to distance from the Chobe River (p = 0.388). Mean tree density increased linearly with distance from the Chobe River (R2 = 0.2333, p < 0.005, n = 34, Figure 3-1A), but showed no relation to distance from water (p = 0.39, Figure 3-1B). Mean tree diameter at breast height (DBH) exhibited a cubic relationship w ith distance from the Chobe River (R2 = 0.7878, p < 0.0001, n = 34, Figure 3-1B), and increased linearly with distance from all water (R2 = 0.3821, p < 0.0001, n = 34, Figure 3-2B). For a summary of relationships between predictor variables and distance to the river and to all water, see Figure 3-1 and Figure 3-2, respectively. Debarking by elephants decreased with distance from the Chobe River (R2 = 0.6217, p < 0.0001, n = 34, Figure 3-3A). When considered in relation to all wate r, debarking patterns exhibited a quadratic distri bution with peaks both near to and far from water (R2 = 0.2986, p < 0.01, n = 34, Figure 3-4A). Debarking density wa s also positively related to mean tree DBH (R2 = 0.2117, p < 0.01, n = 34). Using stepwise multip le regression, distance to river provided the best predictor of debarking (R2 = 0.6217, p < 0.0001, n = 34), with a negative relationship.

PAGE 17

17 The best predictors of debarking (in descending order) were distance to ri ver, total dung density, and mean tree DBH (R2 = 0.7203, p < 0.0001, n = 34). Dung density and DBH were positively correlated with debarking while tree density was negatively correla ted. A similar coefficient of determination is obtained by a model using ju st distance to river and total dung density (R2 = 0.6838, p < 0.0001, n = 34). Branch herbivory peaked closest to the river and again at the fa rthest point from the river, best fitting a cubic regression (R2 = 0.4923, p = 0.0001, n = 34, Figure 3-3B). When the heavily utilized riverfront areas were excluded from analysis, transect s greater than 4 km from the river showed a linear increase in mean branch herb ivory with increasing distance from the river (R2 = 0.4919, p < 0.0001, n = 30, Figure 3-3C). Finer sc ale analyses considering all water found that branch herbivory decreased linearly with distance from water (R2 = 0.1834, p < 0.05, n = 34, Figure 3-4B). Branch herbivory is also negatively related to tree height (R2 = 0.2071, p < 0.01, n = 34). Overall, branch u tilization is best predicted by elephant dung density (R2 = 0.3048, p < 0.001, n = 34), exhibiting a positive relationshi p. The best model is predicted by elephant dung density, mean tree height, and distan ce from river, in descending order (R2 = 0.5622, p < 0.0001, n = 34). Tree height exhi bited a negative relationship w ith branch herbivory and the other predictors a positive relatio nship. For the reduced sample of transects greater than 4 km from the river, branch utilization is be st predicted by distan ce from the river (R2 = 0.4919, p < 0.0001, n = 34) and the overall best model cont ained distance from the river and total dung density (R2 = 0.6127, p < 0.0001, n = 30). Fire damage residual diagnostics showed devi ation from normality assumptions. A squareroot transformation was used to correct this fo r analysis. Fire damage exhibited a quadratic distribution, lowest near the rive r and southern border of the pa rk, and highest about 40 km from

PAGE 18

18 the river (R2 = 0.3438, p < 0.01, n = 34, Figure 3-3D). In general, fire damage increased with distance from water (R2 = 0.1453, p < 0.05, n = 34, Figure 3-4C). Fire damage was best predicted by elephant dung density, exhibiting a negative relationship (R2 = 0.1949, p < 0.01, n = 34). The best model contained, in descending order, elephant dung density, mean percent green vegetation in crown, tree de nsity, and distance to water (R2 = 0.5205, p < 0.001, n = 34). There was a positive relationship to distance from water and a negative relationship with the other three predictor variables.

PAGE 19

19Table 3-1. Descriptive statistics for the most common tree species (n 50) Species n Mean DBH (cm) Mean height (m) Mean percent green vegetation in canopy Mean debarking Mean branch herbivory Mean fire damage Mean distance from river (km) Mean distance from all water (km) Burkea africana 118 23.43 7.19 57.92 1.09 1.99 1.55 19.87 2.81 Colophospermum mopane 148 13.79 4.00 69.83 0.54 3.80 0.41 49.61 1.57 Combretum mossambicense 089 08.89 3.98 11.73 0.84 3.42 0.17 22.76 0.48 Combretum apiculatum 437 11.79 4.63 43.12 0.35 2.90 0.57 37.62 1.87 Combretum elaeagnoides 130 09.18 3.99 47.62 0.20 3.38 0.40 46.26 1.82 Combretum hereroense 061 22.11 5.50 65.80 0.36 2.31 0.56 38.53 2.02 Croton megalobotrys 072 14.69 4.86 49.88 1.32 3.15 0.11 00.44 0.28 Terminalia brachystemma 053 12.99 4.05 61.32 0.57 3.23 1.09 33.59 1.95 Baikiaea plurijuga 166 29.13 7.55 82.52 0.58 1.65 0.84 15.04 3.33 Terminalia sericea 153 09.94 3.97 26.75 0.28 3.05 1.82 39.19 2.02

PAGE 20

20 Figure 3-1. Predictor variable s by distance from the Chobe River. A) tree density (R2 = 0.2333, p < 0.005, n = 34), B) mean diameter at breast height (DBH; R2 = 0.7878, p < 0.0001, n = 34), C) mean tree height (R2 = 0.1704, p < 0.05, n = 34), D) mean percent green vegetation in crown of tree (R2 = 0.0863, p = 0.09, n = 34), E) total dung density (R2 = 0.4792, p < 0.0001, n = 34), and F) elephant dung density (R2 = 0.3918, p < 0.0005, n = 34).

PAGE 21

21 Figure 3-2. Predictor variab les by distance from all water. A) tree density (R2 = 0.0235, p = 0.39, n = 34), B) mean diameter at breast height (DBH; R2 = 0.3821, p < 0.0001, n = 34), C) mean tree height (R2 = 0.6009, p < .0001, n = 34), D) mean percent green vegetation in crown of tree (R2 = 0.06, p = 0.14, n = 34), E) total dung density (R2 = 0.1904, p < 0.05, n = 34), and F) elephant dung density (R2 = 0.0382, p = 0.27, n = 34).

PAGE 22

22 Figure 3-3. Mean utilization at different distances from the Chobe River. A) debarking (R2 = 0.6217, p < 0.0001, n = 34), B) branch herbivory (R2 = 0.4923, p = 0.0001, n = 34), C) reduced branch herbivory, greater than 4 km from river (R2 = 0.4919, p < 0.0001, n = 30), and D) fire damage (R2 = 0.3438, p = 0.0015, n = 34).

PAGE 23

23 Figure 3-4. Mean utilization at distances from all water in Chobe National Park, Botswana. A) debarking (R2 = 0.2986, p = 0.0041, n = 34), B) branch herbivory (R2 = 0.1834, p = 0.0115, n = 34), and C) fire damage (R2 = 0.1453, p = 0.0261, n = 34).

PAGE 24

24 CHAPTER 4 DISCUSSION Patterns of herbivory have im portant ramifi cations for overall landscape stability and resilience, especially for keystone species such as elephants. Effective management plans rely on an understanding of the major ecological processe s that create these patterns. In this study, I consider the effects of water on tr ee utilization in Chobe National Park, Botswana to test whether predictions of the piosphere effect can be applied on a larger scale. Large-Scale Trends Distribu tion of elephant impact varied dependi ng on the type of utilization and on the scale considered. At a broad scale, debarking decrease d with distance from the Chobe River (Figure 3-3A), aligning with piosphere predictions. Bran ch herbivory also initially decreased with distance, confirming findings from previous studies (e.g., Mosugelo et al. 2002), as well as piosphere effect predictions (Figure 3-3B). It is interesting to note, however, that after this initial decrease, mean branch u tilization increased again. In fact, if the highly utilized sections around the riverfront are removed from analysis, br anch herbivory beyond about 4 km correlates strongly to a linear increase with distance from the river (Figure 3-3C ). This is in contrast with the piosphere effect, though a recent study in Kr uger National Park found a similar pattern of increasing tree utilizatio n by elephants with distance from water (Shannon et al. 2008), albeit at smaller spatial scales (up to 4 km from water). They suggested these trends might be explained by tree density, terrain ruggedness, or soil depth. Tree density in Chobe also increased linearly with distance from the Chobe River (Figure 3-1A), but was not correlated with branch herbivory suggesting some other factor is responsible for the appare nt deviation from piosphere predictions.

PAGE 25

25 These large-scale trends in elephant utilization may be related to broader climatic conditions and seasonality, rather than variati on in tree characteristics During the wet season, temporary water pans across the park fill and anim als disperse away from the river (see Skarpe et al., 2004 for an example of this in buffalo). During this time, ther e is an abundance of vegetation on trees leading ele phants that have moved farther away from the river while dispersing to browse mostly on branches, leading to the increasing pattern of branch herbivory with distance from the Chobe River that was obs erved. In the dry season, animals move back to the permanent water of the Chobe River. The reduc tion in available forage at this time of year pushes a shift towards increased debarking to pr ovide water and sugars, leading to increased debarking with proximity to the river. There is also browsing on what vegetation is available near the river, causing the elevated branch herb ivory levels within four kilometers of the riverfront. Unfortunately, this hypo thesis cannot be tested with th e current dataset as data were recorded at only a single time period, necessitating furt her study on this issue. Fire damage exhibits trends opposite those seen by elephant herb ivory, with low fire damage at either end of the distance spectrum (Figure 3-3D). The nega tive relationship between fire damage and elephant dung density suggests that elephant impacts may reduce chances of fire. This result contrasts with studies that have found that elephant utilization of trees, particularly debarking, may increase tree sus ceptibility to fire (Beuchner & Dawkins, 1961; Holdo, 2007; Moncrieff, Kruger, & Midgley, 20 08). Higher elephant im pact, however, could also remove potential fuel load, reducing fire inte nsity in the event of a burn. Additionally, fire is actively suppressed in Chobe National Park and the patterns observed may simply reflect anthropogenic activity, and not be indicative of natural herbivore-fire interactions.

PAGE 26

26 Small-Scale Trends At a sm aller scale, debarking initially follows piosphere effect predictions, decreasing with distance from water up to about 3 km (Figure 3-4A). Beyond this, however, it increases again, contrasting with expected pios phere theory, but aligning with my proposed bimodal distribution. One explanation for this trend is the increase in mean diameter at breast height (DBH) of trees with distance from water. Debarking increased with DBH in Chobe, as has been seen in other areas (e.g., Afolayan, 1975). It is possible that around 3 km from water a threshold is reached where tree size is large enough that debarking is profitable for th e elephants and so prevalence increases. More research is needed to elucidate these fine-scale trends. Branch herbivory trends also differed across scales. Although la rge-scale branch utilization contrasts with piosphere predicti ons, small-scale patterns consideri ng distance from all water fit well within a piosphere effect framework (Figure 3-4B). This trend seems further confirmed by patterns of tree size and distan ce from water. Both mean tree height and DBH increase with distance from water (Figure 3-2B,C). A study in semi-arid grazing lands in Australia found that distance to water did not influen ce plant characteristics (Foran et al. 1982). That this was not found in my study site suggests el ephants may be preferentially browsing species near water, creating a traditional piosphere effect. The re duction of elephant branch herbivory with increasing tree height, presumably because bran ches become less accessible on higher trees, may also contribute to the overall reduction in branch utilization with distance from water. Fire damage increased with distance from water (Figure 3-4C). Areas farther from waterpoints tend to be drier and burn more th an those close to water (Larsen, 1997; Wallenius et al. 2004). In light of this, small-scale fire patt erns are likely to be an effect of habitat characteristics rather than influence of megaherbivores.

PAGE 27

27 Management Implications Previous studies have typically considered piosphere effects at distances of up to about 10 km from water (Child et al. 1971; Thrash et al. 1991; Thrash, 1998; Thrash & Derry, 1999; Brits et al. 2002). My large-scale findings for both deba rking and branch herb ivory suggest that these utilization trends may continue far beyond th at which has been previously suggested. This has important implications for the stability of la ndscapes in semi-arid systems. Instead of just influencing a sacrifice area and utilization zone several kilometers around a waterpoint (Brits et al. 2002), elephants may be affecting trees across the landscape. This may result in unexpected shifts in land cover and species co mposition if these more distant areas are not monitored. One trend observed during the course of this study suggests these sh ifts may already be happening. Chobe National Park was famous for its Acacia woodlands up until the 1960s (Child, 1968; Simpson, 1975; Skarpe et al. 2004). In my study, however, only three individuals of a single species were recorded out of 1600 total trees. This phenomenon has been observed by other studies in the area as well (e.g., Lewin, 1986; Skarpe et al., 2004; Wolf, 2008. Masters Thesis University of Florida. Gainesville, Florid a, USA.). It is generally accepted that elephants have played a role in this dec line, perhaps in conjunction with other species, such as impala, which prevented woodland regeneration and growth of seedlings (Lewin, 1986; Rutina, 2004. Impalas in an elephant-impacted woodland: browser-driven dynamics of the Chobe riparian zone, northern Botswana. PhD Thesis Agricultural University of Norway. s, Norway.). Skarpe et al. (2004) suggest that in light of these influences, species such as Acacia may require local refuges to persist. The findings from my study in dicate inland areas farther from the river may not provide these refuges and that further ev aluation is needed to better understand the mechanisms behind species reductio ns and what steps could be ta ken to prevent future losses.

PAGE 28

28 It is possible that the expanded piosphere effect seen in this study reflects the large size of the Chobe River and the influence it has on wild life movements and dynamics. In places with smaller rivers or more abundant alternative water sources, different utilization patterns may occur. Further research should test whether an ex tended piosphere effect is seen for elephants in other semi-arid systems and whether this is applicable both to artificial waterpoints as well as large natural sources such as rivers. Multiple regression models evaluating predicto rs of elephant utilization suggest more efficient strategies for monitoring through vegeta tion surveys. Researchers and managers must balance minimizing time and cost while maximizi ng information yield. All three types of tree utilization were predicted by dung density as well as distance to the river or water. While predictions could be improved by adding measur ements of the trees themselves, dung counts may provide a quick and coarse method to estimate levels of elephant impact for an area. For the manager or researcher surveying wide areas w ith limited resources, spat ially located dung counts combined with geographic information systems software containing water locations may be the most efficient broad survey method, to be follo wed up with more detailed tree evaluations in areas of concern. The characteristics of the study location cons trained tree evaluation to areas accessible from tracks. While only small dirt tracks were ut ilized and all transects were located at least 50 m from tracks to avoid potential negative effects, there is still a chance that the presence of these structures influenced my results. Unfortunate ly, the nature of the park and the high density of elephants necessitated proximity to an area accessible to a ve hicle. Future work will use satellite remote sensing to investigate how vegeta tion trends vary over wider areas of the park to test further application of my findings, as well as the potential effect of tracks.

PAGE 29

29 This study has examined the utility and applic ation of the piosphere effect at multiple scales in southern Africa. While the piosphere effect generally seems to be upheld near waterpoints, over a larger scale this is more que stionable and seems to re late to the type of herbivory that is occurring. It is possible that in some cont exts piosphere effects may extend much farther than previously suggested. Further study into the dynamics of the piosphere effect on woody vegetation is needed to understand thes e complex trends and see how far they may extend, as well as considering what other factors are influencing patterns of vegetation utilization by elephants.

PAGE 30

30 REFERENCES AFOLAYAN, T. A. (1975) Effects of elephant activities on forest plantations in the Kilimanjaro forest-game reserve in northern Tanzania. Oikos 26, 405-410. AUGUSTINE, D. J. & MCNAUGHTON, S. J. ( 2004) Regulation of shrub dynamics by native browsing ungulates on Ea st African rangeland. J. Appl. Ecol. 41, 45-58. BAILEY, D. W., GROSS, J. E., LACA, E. A., RITTENHOUSE, L. R., COUGHENOUR, M. B., SWIFT, D. M. & SIMS, P. L. (1996) Mechanis ms that result in la rge herbivore grazing distribution patterns. J. Range Mgmt 49, 386-400. BARNES, R. F. W. (2001) How reliable are dung counts for estimating elephant numbers? Afr. J. Ecol. 39, 1-9. BEN-SHAHAR, R. (1996) Woodlan d dynamics under the influence of elephants and fire in Northern Botswana. Plant Ecol. 123, 153-163. BEUCHNER, H. K. & DAWKINS, H. C. (1961) Vegetation change induced by elephants and fire in Murchison Falls National Park, Uganda. Ecology 42, 752-766. BLANC, J. J., BARNES, R. F. W., CRAIG, G. C., DUBLIN, H. T., THOULESS, C. R., DOUGLAS-HAMILTON, I. & HART, J. A. (2007) African Elephant Status Report 2007: an update from the African Elephant Database. In: Occasional Paper Series of the IUCN Species Survival Commission IUCN/SSC African Elepha nt Specialist Group, Gland, Switzerland. BRITS, J., VAN ROOYEN, M. W. & VAN ROOYEN, N. (2002) Ecological impact of large herbivores on the woody vegetation at selected watering points on the eastern basaltic soils in the Kruger National Park. Afr. J. Ecol. 40, 53-60. CHAMAILL-JAMMES, S., VALEIX, M. & FRITZ, H. (2007) Managing heterogeneity in elephant distribution: intera ctions between elephant popula tion density and surface-water availability. J. Appl. Ecol. 44, 625-633. CHILD, G. (1968) An Ecological Survey of Northeastern Botswana FAO, #TA2563. Rome, Italy. CHILD, G., PARRIS, R. & RICH, E. (1971) Use of mineralised water by Kalahari wildlife and its effects on habitats. East Afr. Wildl. J. 9, 125-142. DE BEER, Y., KILIAN, W., VE RSFELD, W. & VAN AARDE, R. J. (2006) Elephants and low rainfall alter woody vegetation in Etosha National Park, Namibia. J. Arid Environ. 64, 412-421. DUBLIN, H. T., SINCLAIR, A. R. E. & MCGLADE J. (1990) Elephants and fire as causes of multiple stable states in th e Sereng eti-Mara woodlands. J. Anim. Ecol. 59, 1147-1164.

PAGE 31

31 FORAN, B. D., BASTIN, G., REMENGA, E. & HYDE, K. W. (1982) The response to season, exclosure, and distance from water of three central Australian past ure types grazed by cattle. Aust. Rangel. J. 4, 5-15. FULLER, T. K. (1991) Do pellet counts index wh ite-tailed deer numbers and population change? J. Wildl. Mgmt 55, 393-396. HERREMANS, M. (1995) Effects of woodl and modification by African elephant Loxodonta africana on bird diversity in northern Botswana. Ecography 18, 440-454. HISCOCKS, K. (1999) The impact of an increasing elephant population on the woody vegetation in southern Sabi Sand Wildtuin, South Africa. Koedoe 42, 47-55. HOLDO, R. M. (2007) Elephants, fire, and frost can determine co mmunity structure and composition in Kalahari woodlands. Ecol. Appl. 17, 558-568. LANGE, R. T. (1969) The piosphere: Sheep track and dung patterns. J. Range Mgmt 22, 396-400. LARSEN, C. P. S. (1997) Spatial and temporal va riations in boreal fore st fire frequency in northern Alberta. J. Biogeogr. 24, 663-673. LAWS, R. M. (1970) Elephants as agents of ha bitat and landscape change in East Africa. Oikos 21, 1-15. LEWIN, R. (1986) In ecology, change brings stability. Science 234, 1071-1073. MONCRIEFF, G. R., KRUGER, L. M. & MI DGLEY, J. J. (2008) Stem mortality of Acacia nigrescens induced by the synergistic effects of ele phants and fire in Kruger National Park, South Africa. J. Trop. Ecol. 24, 655-662. MOSUGELO, D. K., MOE, S. R., RINGROSE, S. & NELLEMANN, C. (2002) Vegetation changes during a 36-year period in nort hern Chobe National Park, Botswana. Afr. J. Ecol. 40, 232-240. NELLEMANN, C., MOE, S. R. & RUTINA, L. P. (2002) Links between terrain characteristics and forage patterns of elephants (Loxodonata africana) in northern Botswana. J. Trop. Ecol. 18, 835-844. PELLEW, R. A. P. (1983) The impacts of elephant, giraffe and fire upon the Acacia tortilis woodlands of the Serengeti. Afr. J. E col. 21, 41-74. REDFERN, J. V., GRANT, R., BIGGS, H. & G ETZ, W. M. (2003) Surface-water constraints on herbivore foraging in the Kruge r National Park, South Africa. Ecology 84, 2092-2107. SHANNON, G., DRUCE, D. J., PAGE, B. R ., ECKHARDT, H. C., GR ANT, R. & SLOTOW, R. (2008) The utilization of large savanna trees by elephant in southern Kruger National Park. J. Trop. Ecol. 24, 281-289.

PAGE 32

32 SIMPSON, D. C. (1975) A detailed vegeta tion study on the Chobe River in North-East Botswana. Kirkia 10, 185-227. SKARPE, C. (1992) Dynamics of savanna ecosystems. J. Veg. Sci. 3, 293-300. SKARPE, C., AARRESTAD, P. A., ANDREASSE N, H. P., DHILLION, S. S., DIMAKATSO, T., DU TOIT, J. T., DUNCAN, HALLEY, J., HYTTEBORN, H., MAKHABU, S., MARL, M., MAROKANE, W., MASUNGA, G., MODISE, D., MOE, S. R., MOJAPHOKO, R., MOSUGELO, D., SEJOE, T. B., STOKKE, S., SWENSON, J. E., TAOLO, C., VANDEWALLE, M. & WEGGE, P. (2004) The return of the giants: ecological effects of an in creasing elephant population. Ambio 33, 276-282. STUART, C. & STUART, T. (2000) A Field Guide to the Tracks & Signs of Southern and East African Wildlife Struik Publishers, Cape Town. THRASH, I. (1998) Impact of water provision on herbaceous vegetation in Kruger National Park, South Africa. J. Arid Environ. 38, 437-450. THRASH, I. & DERRY, J. F. (1999) The natu re and modelling of piospheres: a review. Koedoe 42, 73-94. THRASH, I., NEL, P. J., THERON, G. K. & BO THMA, J. D. P. (1991) The impact of the provision of water for game on the woody ve getation around a dam in the Kruger National Park. Koedoe 34, 131-148. VALEIX, M., FRITZ, H., DUBOIS, S., KANEN GONI, K., ALLEAUME, S. & SAID, S. (2007) Vegetation structure and ungulate abundance over a period of increasing elephant abundance in Hwange National Park, Zimbabwe. J. Trop. Ecol. 23, 87-93. VAN LANGEVELDE, F., VAN DE VIJVER, C. A. D. M., KUMAR, L., VAN DE KOPPEL, J., DE RIDDER, N., VAN ANDEL, J., SKIDMORE, A. K., HEARNE, J. W., STROOSNIJDER, L., BOND, W. J., PRINS, H. H. T. & RIETKERK, M. (2003) Effects of fire and herbivory on the st ability of savanna ecosystems. Ecology 84, 337-350. WALKER, B. H. (1976) An approach to the monitoring of changes in the composition and utilization of woodlands and savanna vegetation. S. Afr. J. Wildl Res, 6, 1-32. WALLENIUS, T. H., KUULUVAI NEN, T. & VANHA-MAJAMAA, I. (2004) Fire history in relation to site type and ve getation in Vienansalo wilderness in eastern Fennoscandia, Russia. Can. J. For. Res. 34, 1400-1409. YOUNG, T. P., PALMER, T. M. & GADD, M. E. (2005) Competition and compensation among cattle, zebras, and elephants in a se mi-arid savanna in Laikipia, Kenya. Biol. Conserv. 122, 351-359.

PAGE 33

33 BIOGRAPHICAL SKETCH Tim othy Fullman was born and raised in southe rn California. His parents instilled in him a love of animals and the outdoors and a desire to see them protected. In 2007, he graduated with a Bachelor of Science degree in animal biology from the University of CaliforniaDavis. He earned his Master of Science in interdisciplinar y ecology in 2009 from the University of Florida, Gainesville. He currently lives in Gainesville, Florida, with his wife and enjoys hiking and camping as well as learning about wildlife. He is pursuing a doctoral degree from the University of Florida and a career as a professor at a resear ch university, combining field research in Africa with teaching the next generation of conservationists.