<%BANNER%>

Small Mammal Response to Food, Fire, Herbicide and Predation in Longleaf Pine Ecosystem

Permanent Link: http://ufdc.ufl.edu/UFE0043777/00001

Material Information

Title: Small Mammal Response to Food, Fire, Herbicide and Predation in Longleaf Pine Ecosystem
Physical Description: 1 online resource (63 p.)
Language: english
Creator: Karmacharya, Binab
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: application -- capture-mark-recapture -- combination -- ecosystem -- exclusion -- fire -- glaucomys -- gossypinus -- habitat -- herbicide -- herbicide-fire -- hispidus -- longleaf -- management -- peromyscus -- pine -- predator -- prescribed -- sigmodon -- survival -- volans
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Little is known about demographic parameters of the southern flying squirrel (Glaucomys volans) or factors influencing those parameters. We conducted a capture-mark-recapture study to provide rigorous estimates of survival rates for the southern flying squirrel in a longleaf pine ecosystem. We also experimentally examined the effect of food supplementation, prescribed fire and mammalian predator exclusion on their survival rates. Apparent monthly survival was 0.84 ± SE 0.01, with no evidence for sex-specific or seasonal variation. Survival was higher for adults (0.85 ± 0.01) than for juveniles (0.77 ± 0.05). Prescribed fire positively influenced survival; survival increased for a period up to six months after burns. There was no evidence that food supplementation and exclusion of mammalian predators influenced survival. These results suggest that the southern flying squirrel population in our study site is not food limited and mammalian species are not the major predator for this species. Rapid decline and degradation of longleaf pine ecosystems in southeastern United States are conservation concerns. Prescribed fire is the primary management activity in this fire-dependent ecosystem, but prescribed fire is under increasing scrutiny due to air quality issues. There are concerns that prescribed fire may be removed, or replaced by herbicide, as a forest management tool without adequate understanding of its ecological consequences. We conducted a capture-mark-recapture study from April 1999 to April 2002 to examine experimentally the effect of prescribed fire, herbicide application and herbicide-prescribed fire combination on apparent survival rates of Peromyscus gossypinus and Sigmodon hispidus population. There was no evidence that herbicide application affected survival of cotton mice. Evidence to support effect of prescribed fire and herbicide-fire combination on survival of cotton mice was weak; apparent monthly survival increased after these treatments. There was strong evidence that prescribed fire and herbicide-fire treatments affected survival of cotton rats, but the evidence of a herbicide effect alone was weak; survival rates declined after all three treatments. Fire alone had a stronger effect than an herbicide-fire treatment in both species. A detailed study of ecological consequences of these alternative management practices should be carried out before replacing or augmenting prescribed fire with herbicide.
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 Binab Karmacharya.
Thesis: Thesis (M.S.)--University of Florida, 2011.
Local: Adviser: Oli, Madan K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043777:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043777/00001

Material Information

Title: Small Mammal Response to Food, Fire, Herbicide and Predation in Longleaf Pine Ecosystem
Physical Description: 1 online resource (63 p.)
Language: english
Creator: Karmacharya, Binab
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: application -- capture-mark-recapture -- combination -- ecosystem -- exclusion -- fire -- glaucomys -- gossypinus -- habitat -- herbicide -- herbicide-fire -- hispidus -- longleaf -- management -- peromyscus -- pine -- predator -- prescribed -- sigmodon -- survival -- volans
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Little is known about demographic parameters of the southern flying squirrel (Glaucomys volans) or factors influencing those parameters. We conducted a capture-mark-recapture study to provide rigorous estimates of survival rates for the southern flying squirrel in a longleaf pine ecosystem. We also experimentally examined the effect of food supplementation, prescribed fire and mammalian predator exclusion on their survival rates. Apparent monthly survival was 0.84 ± SE 0.01, with no evidence for sex-specific or seasonal variation. Survival was higher for adults (0.85 ± 0.01) than for juveniles (0.77 ± 0.05). Prescribed fire positively influenced survival; survival increased for a period up to six months after burns. There was no evidence that food supplementation and exclusion of mammalian predators influenced survival. These results suggest that the southern flying squirrel population in our study site is not food limited and mammalian species are not the major predator for this species. Rapid decline and degradation of longleaf pine ecosystems in southeastern United States are conservation concerns. Prescribed fire is the primary management activity in this fire-dependent ecosystem, but prescribed fire is under increasing scrutiny due to air quality issues. There are concerns that prescribed fire may be removed, or replaced by herbicide, as a forest management tool without adequate understanding of its ecological consequences. We conducted a capture-mark-recapture study from April 1999 to April 2002 to examine experimentally the effect of prescribed fire, herbicide application and herbicide-prescribed fire combination on apparent survival rates of Peromyscus gossypinus and Sigmodon hispidus population. There was no evidence that herbicide application affected survival of cotton mice. Evidence to support effect of prescribed fire and herbicide-fire combination on survival of cotton mice was weak; apparent monthly survival increased after these treatments. There was strong evidence that prescribed fire and herbicide-fire treatments affected survival of cotton rats, but the evidence of a herbicide effect alone was weak; survival rates declined after all three treatments. Fire alone had a stronger effect than an herbicide-fire treatment in both species. A detailed study of ecological consequences of these alternative management practices should be carried out before replacing or augmenting prescribed fire with herbicide.
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 Binab Karmacharya.
Thesis: Thesis (M.S.)--University of Florida, 2011.
Local: Adviser: Oli, Madan K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043777:00001


This item has the following downloads:


Full Text

PAGE 1

1 SMALL MAMMAL RESPONSE TO FOOD, FIRE, HERBICIDE AND PREDATION IN LONGLEAF PINE ECOSYSTEM By BINAB KARMACHARYA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Binab Karmacharya

PAGE 3

3 To my parents wife and daughter

PAGE 4

4 ACKNOWLEDGMENTS I thank my advisor Dr. Madan Oli for his continuous support and guidance throughout my graduate study and for his patience, critical suggestions and encouragement through the preparation of this manuscript. Thanks to my committee members Drs. L. Mike Conne r and Michael Moulton, for their help in every aspect of the project design, field work, data analysis and writing the manuscript. I would like to express my sincere gratitude to Michael Simmons for managing database and Jessica Rutledge for arranging fiel dwork logistics completing this work without their help. Gail introduced me to the basics of the mark recapture analysis and guided me through the initial stages of my s truggle with MARK, and Jeff provided me with all important R codes and suggestions through the whole process of data analysis. Sincere thanks to my lab mates Oscar Murillo and Miguel Acevedo for helping me during my fight with R codes. I am grateful to man y people and agencies for supporting my graduate degree and this project. The Fulbright Commission and the Department of Wildlife Ecology and Conservation, University of Florida supported my graduate program, and Jones Ecological Research Center provided f unding, equipment and manpower for the fieldwork. I would also like to thank all the staff members of the Department of Wildlife Ecology and Conservation at the University of Florida, and Jones Center for their continuous support. I am grateful to my pare nts and wife Janabi for their support and encouragements throughout my graduate program.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 ABSTRACT ................................ ................................ ................................ ..................... 8 CHPATER 1 INTRODUCTION ................................ ................................ ................................ .... 10 2 THE INFLUENCE OF MAMMALIAN PREDATOR EXCLUSION, FOOD SUPPLEMENTATION AND PRESCRIBED FIRE ON SURVIVAL OF THE SOUTHERN FLYING SQUIRREL IN A LONGLEAF PINE ECOSYSTEM .............. 12 Background ................................ ................................ ................................ ............. 12 Materials and Methods ................................ ................................ ............................ 14 Study Area ................................ ................................ ................................ ........ 14 Study Species ................................ ................................ ................................ .. 14 Field Methods ................................ ................................ ................................ ... 15 Statistical Method and Capture Mark Recapture (CMR) Analysis ................... 17 Results ................................ ................................ ................................ .................... 19 Discussion ................................ ................................ ................................ .............. 21 3 LONGLEAF PINE MANAGEMENT PRACTICES AND THEIR IMPACT ON SMALL MAMMAL POPULATIONS ................................ ................................ ......... 35 Background ................................ ................................ ................................ ............. 35 Materials and Meth ods ................................ ................................ ............................ 37 Study Area. ................................ ................................ ................................ ....... 37 Study Species ................................ ................................ ................................ .. 37 Field Methods ................................ ................................ ................................ ... 37 Statistical Method and Capture Mark Recapture (CMR) Analysis .................... 38 Results ................................ ................................ ................................ .................... 40 Cotton Mice ................................ ................................ ................................ ...... 40 Cotton Rat ................................ ................................ ................................ ........ 41 Discussion ................................ ................................ ................................ .............. 42 4 CONCLUSIONS ................................ ................................ ................................ ..... 54 LIST OF REFERENCES ................................ ................................ ............................... 56 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 63

PAGE 6

6 LIST OF TABLES Table page 2 1 Model comparison table for capture mark recapture analysis of flying squirrel to invest igate the best base model for capture probability ................................ .. 27 2 2 Model comparison table for capture mark capture analysis of flying squirrel to investigate the best base model for survival. ................................ ...................... 28 2 3 Model comparison table for capture mark recapture anal ysis of flying squirrel to investigate the effect of food, fire and predation on survival ........................... 29 2 4 Model comparison table for captu re mark recapture analysis of flying squirrel to investigate the best fire effect window on survival ................................ .......... 30 2 5 Model comparison table for capture mark recapture analysis of flying squirrel to investigate the additive and interactive effects of food, fire and predation on survival ................................ ................................ ................................ .......... 31 3 1 Model comparison table for capture mark recapture analysis of cotton mice to investigate base model for capture probability survival rate and transition probability ................................ ................................ ................................ ........... 46 3 2 Model comparison table for capture mark recapture analysis of cotton mice to investigate the effect window of fire, herbicide and herbici de fire combination on survival ................................ ................................ ...................... 47 3 3 Model comparison table for capture mark recapture analysis of cotton mice to investigate effect of fire, herbicide and herbicide fire combination on survival ................................ ................................ ................................ ............... 48 3 4 Model comparison table for capture mark recapture analysis of cotton rat to in vestigate base model for capture probability, survival rate and transition probability. ................................ ................................ ................................ .......... 49 3 5 Model comparison table for capture mark recapture analysis of cotton rat to investigate the effect window of fire, herbicide and herbicide fire combination on survival. ................................ ................................ ................................ ......... 50 3 6 Model comparison table for capture mark recapture analysis of cotton rat to investigate effect of fire, herbicide and herbicide fire combination on surviv al. .. 51

PAGE 7

7 LIST OF FIGURES Figure page 2 1 Monthly apparent survival estimates of flying squirrels from 2003 2010.. .......... 32 2 2 Comparison of monthly apparent survival estimates of southern flying squirrels before and after fire. ................................ ................................ ............. 33 2 3 Seasonal variation in transition probability from juvenile to adult state for southern flying squirrels from 2003 2 010. ................................ .......................... 34 3 1 Effect of prescribed fire (a) and herbicide fire combination (b) on monthly apparent survival ( SE) of cotton mice in the year 2000 ................................ ... 52 3 2 Effect of prescribed fire (a), herbicide application (b) and herbicide fire combination (c) on monthly a pparent survival estimates of cotton rat in the year 2000.. ................................ ................................ ................................ .......... 53

PAGE 8

8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Master of Science SMALL MAMMAL RESPONSE TO FOOD, FIRE, HERBICIDE AND PREDATION IN LONGLEAF PINE ECOSYSTEM By Binab Karmacharya December 2011 Chair: Madan K. Oli Major: Wildlife Ecology and Conservation Little is known about demographic parameters of the southern flying squirrel (Glaucomys volans) or factors influencing those parameters. We conducted a capture mark recapture study to provide rigorous estimates of survival rates for the south ern flying squirrel in a longleaf pine ecosystem. We also experimentally examined the effect of food supplementation, prescribed fire and mammalian predator exclusion on their survival rates. Apparent monthly survival was 0.84 SE 0.01, with no evidence f or sex specific or seasonal variation. Survival was higher for adults (0.85 0.01) than for juveniles (0.77 0.05). Prescribed fire positively influenced survival; survival increased for a period up to six months after burns. There was no evidence that f ood supplementation and exclusion of mammalian predators influenced survival. These results suggest that the southern flying squirrel population in our study site is not food limited and mammalian species are not the major predator for this species. Rapid decline and degradation of longleaf pine ecosystem s in southeastern United States are conservation concern s P rescribed fire is the primary management activity in this fire dependent ecosystem but prescribed fire is under increasing scrutiny due to air quality issues. There are concern s that prescribed fire may be removed or replaced by

PAGE 9

9 herbicide, as a forest management tool without adequate understanding of its ecological consequences We conducted a capture mark recapture study from April 1999 to Apri l 2002 to examine experimentally the effect of prescribed fire, herbicide application and herbicide prescribed fire combination on apparent survival rates of Peromyscus gossypinus and Sigmodon hispidus population. There was no evidence that herbicide appli cation affected survival of cotton mice. Evidence to support effect of prescribed fire and herbicide fire combination on survival of cotton mice was weak; apparent monthly survival increased after these treatments. There was strong evidence that prescribed fire and herbicide fire treatments affected survival of cotton rats, but the evidence of a herbicide effect alone was weak; survival rates declined after all three treatments. Fire alone had a stronger effect than an herbicide fire treatment in both spec ies. A detailed study of ecological consequences of these alternative management practices should be carried out before replacing or augmenting prescribed fire with herbicide.

PAGE 10

10 CHAPTER 1 INTRODUCTION Longleaf pine ( Pinus palustris ) ecosystem is a species rich endangered ecosystem of North America (Noss et al. 1995; Alavalapati et al. 2002). Once a dominant ecosystem in southeastern United States covering more than 36 million ha before the arrival of European settlers (Landers et al. 1995), it is now restricted to less than 1.2 million ha (Alavalapati et al. 2002). Over millenni a, anthropogenic and lightening ignited frequent fire has been dominant ecological process maintaining the longleaf pine ecosystem; disruption of these fire regimes, along with various other anthropogenic factors, is thought to be an important cause of rap id decline and degradation of this ecosystem (Van Lear et al. 2005). Longleaf pine ecosystems require some management activity to maintain the favorable grassy understory. Prescribed fire during the growing season has been the primary management activity f or the maintenance or restoration of this fire dependent ecosystem (Frost 1990; Glitzenstein et al. 1995; Streng 1993). However, prescribed fire is under increased scrutiny due to perceived air quality issues, particularly the emission of atmospheric parti cles and their potential impacts on human health (Sandberg et al. 2002) A possible alternative to prescribed burning involves applications of herbicide alone or in combination with prescribed fire. Besides these ecosystem level management tools, population level ma nagement activities like supplemental feeding in different form and predator exclusion or removal are also employed, primarily for the management of game birds. Small mammal population dynamics in longleaf pine forest affected by many natural processes an d anthropogenic activities. Food, fire and predation are arguably three of the most important factors affecting small mammal demography in fire

PAGE 11

11 maintained habitat such as the longleaf pine ecosystem. Several studies have experimentally evaluated additive a nd interactive effects of food, predation and fire on demographic variables of small mammals ( Krebs et al. 1995; McNamara and Houston 1987; Mitchell et al. 1992; Morris et al. 2011b; Morris et al. In press; Sinclair and Pech 1996) However, no study has tested for the population level effects of food, fire and predation on the southern flying squirrels ( Glaucomys volans ). In fact, little is known a bout demography and population dynamics of the southern flying squirrel, the most widely distributed flying squirrel of North America. Similarly, we are not aware of any study that has investigated how herbicide, alone or in combination with prescribed fir e, affects small mammals inhabiting longleaf pine habitats. To fill this gap in knowledge, the first chapter of this thesis investigates the survival rate of southern flying squirrel and tests for the effect of food supplementation, mammalian predator excl usion and prescribed fire on southern flying squirrel survival. The second chapter of this thesis aims to provide data on population level impact of longleaf pine ecosystem management practices (Prescribed fire, herbicide application and herbicide fire com bination) on small mammals inhabiting the longleaf pine ecosystem ecosystem.

PAGE 12

12 CHAPTER 2 THE INFLUENCE OF MAM MALIAN PREDATOR EXCL USION, FOOD SUPPLEMENTATION AND PRESCRIBED FIRE ON S URVIVAL OF THE SOUTH ERN FLYING SQUIRREL IN A LONGLEAF PINE ECOSYS TEM Background Food and predation are thought to be two of the most important factors influencing population dynamics of small mammals; every small mammal requires food to survive and is a potential prey for others. Several studies have evaluated population le vel influence of experimentally supplemented food on small mammal s These studies revealed that food supplementation can lead to a larger body size (Cole and Batzli 1978) longer breeding season (Hansen and Batzli 1979; Taitt and Krebs 1983) greater density (Hubbs and Boonstra 1997) increased survival (Ransome and Sullivan 2004) and larger litter size (Campbell and Slade 1995 ; Cole and Batzli 1978 ) However, the population level effect of food supplementation can vary among species and habitats. For example, recent food supplementation studies (Morris et al. 2011b) in a longleaf pine ( Pinus palustris ) wiregrass ( Aristida beyrichiana ) ecosyst em showed that food supplementation increased survival, abundance and rate of transition from non reproductive to reproductive states in the hispid cotton rat ( Sigmodon hispidus ) O n the same site, Morris et al. ( In press) found that supplemental food increased abundance but not su rvival in cotton mice ( Peromyscus gossypinus ) and oldfield mice ( P. polionotus ). Likewise, Ransome and Sullivan ( 1997) found that supplemental food increased population size but had no measurable influence on proportion of adult s breeding, or survival or recruitment rates in northern flying squirrel ( Glaucomys sabrinus ) and red squirrels ( Tamiasciurus hudsonicus ). However, another study with the same two species (Ransome and Sullivan 2004) revealed that survival of northern flying

PAGE 13

13 squirrels was sig nificantly higher on grids with food supplementation whereas the red squirrel population did not respond to supplemental food. Predation plays a vital role in population dynamics of any prey species Increase in mortality rate is the most obvious effect of predation and in small mammals the amount of activity leading to increased visibility, audibility or smell raises the predation risk (Norrdahl and Kporpimaki 1995). P redator exclusion or removal experiments l ed to increased densities in cotton rat s and me adow voles (Microtus pennsylvanicus) (Schnell 1968) and survival in degu ( Octodon degus ) (Meserve et al. 1993) In southern Georgia, experimental exclusion of mammalian predators had no effect on demographic parameters of hispid cotton rat and cotton mice, but it positively influenced survival and abundance of oldfield mice (Morris et al. 2011b; Morris et al. In press) In small mammal species inhabiting fire maintained ecosystems, such as the longleaf pine ecosystem, frequency and scale of fire can also influence populations (Pyne et al. 1996) Fire reduced survival, abundance and/or transition from non reproductive to reproductive state s in cotton rat, cotton mice and oldfield mice (Morris et al. 2011b; Morris et al. In press) ; similar results were reported for northern flying squirrels in ponderosa pine (Pinus ponderosa ) and mixed conifer forests (Lehmkuhl et al. 2006) Tree squirrels appeared to benefit indirectly from low intensity ground fires which stimulated cone and fungi production and created open stands that facilitate foraging (Weigl et al. 1989) Several studies have experimentally evaluated additive and i nteractive effects of food, predation and fire on demographic variables of small mammals ( Krebs et al. 1995 ; McNamara and Houston 1987; Mitchell et al. 1992; Morris et al. 2011b; Morris et

PAGE 14

14 al. In press; Sinclair and Pech 1996) However, we know of no study that has tested for the population level effects of food, fire and p redation on the southern flying squirrel ( Glaucomys volans ). In fact little is known about demography and population dynamics of the southern flying squirrel, the most widely distributed flying squirrel of North America. Our specific objectives were to: ( 1) provide rigorous estimates of survival rates, and to test for sex and age specific and seasonal variation in survival, and (2) experimentally test for additive and interactive effects of food supplementation, mammalian predator exclusion, and prescrib ed fire on survival of the southern flying squirrel. Materials and Methods Study Area The study was conducted at the Joseph W. Jones Ecological Research Center at Ichauway, a 12,000 ha site managed for research, education and conservation, located in Baker County, Georgia. The site is primarily a longleaf pine and wiregrass ecosystem with associated hardwood tree species, wetlands, slash pine ( Pinus elliottii ) and a variety of other habitats (Atkinson et al. 1996) Long leaf pine grassland ecosystem is shaped frequent fire, induced by either humans or lightening ( e.g., Van Lear et al. 2005) Our study site was ma naged by prescribed fire (Glitzenstein et al. 1995) and the Center employed dormant and growing seaso n burns in an approximate two year burn rotation 4,000 6,000 ha annually. Study Species The southern flying squirrel is one of two species of the genus Glaucomys ; the other is the northern flying squirrel Its distribution ranges from temperate to sub temperate pine hardwood forests from southernmost Quebec southward through the

PAGE 15

15 eastern half of the United States to Honduras (Dolan and Carter 1977) Flying squirrels are relatively long lived (~10 years), and use tree cavities for nut storage, defecatoria, den sites and breeding sites; individuals may use several cavities (Brady et al. 2000) They compete with other cavity nesting species, such as th e red cockaded woodpecker (Picoides borealis), for available nest sites (Stabb et al. 1989) It is generally accepted that populations of southern flying squirrels have increased and that they are a threat for conservation of endangered re d cockaded woodpeckers ( Borgo et al. 2010; Werner and Peacor 2003) but little is known about population ecology of this species. Field Methods There were four mammalian predator exclusion plots in the Jones Center; each was surrounded by 4 ft. tall woven wire fencing with three parallel e lectric wires at top, middle and top to prevent climbing and digging by potential predators. The size of the fence weave allowed movement of small mammals and snakes in and out of the predator exclosures. Significantly fewer mamma lian predators in exclosur es tha n in control plots was evident from regular track count s and thermal camera survey s (Conner et al. 2010) When mammalian predators were detected inside exclosures, efforts were made to trap and remove them. A nearby site with same size and similar habitat was s elected as a control plot for each exclosure Size of these control and exclosure plots ranged from 36 to 49 ha. Because of their geographical proximity and habitat similarity, pair s of control and exclosure plots were considered as one site Supplemental food was provided by placing cans with 113 g of rabbit chow at every other station in two exclosure and two control grids from June 2007 to August 2009. These grids were randomly selected and f eeding stations were checked and refilled every other week Su pplemental feeding was continued on two randomly

PAGE 16

16 selected exclosures from July 2010 to November 2010 with 300 g of rolled oat This time plastic feeders were placed 2 m high up in the tree in 4*4 grids with approximately 60 m between the stations. Feeding stations were checked and refilled every week. Motion sensitive trail cameras revealed the regular use of feeding stations by flying squirrels and other species of small mammals Prescribed fire s were executed in Februaries of 2005, 2007 and 2009 according to As part of an ongoing study of small mammals, a 1212 small mammal trapping grid was established in all 4 control and 4 exclosure sites, with about 15 m between stations. S herman live traps (H.B Sherman Traps, Tallahassee, Florida, USA) were placed on the ground. Additionally, 24 traps (4 traps in each alternate row) were set at about 2m high on the nearest suitable tree. Grids in each plot were trapped once per season from May 2003 through June 2007, and twice per season from July 2007 through August/September 2009. Trapping was continued twice per season in one exclosure and one control from September 2009 to May 2010. The remaining 6 plots were not trapped during that time In May 2010, 6 plots (two controls, two predator exclosure s and two predator exclosure s plus food supplementation) were randomly selected for a flying squirrel study. Traps, mounted on wooden platforms, were set 2 m high in the nearest suitable tree in t he same 12 12 small mammal trapping grid Trapping was done twice per season through November 2010. During each trapping session, traps were set for 4 consecutive days ( Monday Thursday ). The traps were baited with rolled oat s and mixed bird feed, and ch ecked every morning. Fire ant invasion of baited trap for small mammals is a major problem in

PAGE 17

17 the southeastern United States (Mitchell et al. 1996), so we sprinkled granular insecticide (Talstar) around each trap to prevent fire ant infestation. Newly caug ht individuals were marked with metal ear tags in both ear and released at the capture site. For all captured individuals we recorded: date, location, identity (if previously captured), sex, mass state (adult or juvenile based on mass ), reproductive condi tion and hind foot length Individuals weighing more tha n 50 g were classified as adult (Laves and Loeb 2006). In June 2010, i n the six plots selected for the flying squirrel study 16 wooden nest boxes were placed in a suitable tree at about 2m high in a 44 grid with 60 m between the nest boxes. The internal dimensions of the nest boxes were 131336 cm with 3 cm entrance hole on the side ( Brady et al. 2000) Nest boxes were checked every week for the presence of flying squirrels. Squirrels in the nest boxes were captured and newly captured individuals were marked and data were collected as described previously. If a litter was found in t he nest box, number, sex and mass of newborns were recorded. After recording the relevant data, squirrels were released back into the nest box. All field methods were approved by the University of Florida Institute of Food and Agricultural Sciences Animal Research Committee and were in accordance with the r ecommendations of the American Society of Mammalogists (Gannon et al. 2007) Statistical Method and Capture Mark Rec apture (CMR) A nalysis There were 39 4 day capture occasions from May 2003 to November 2010. Each year was divided into four seasons; spring (March 20 June 20), summer (June21 September 22), fall (September 23 December 20) and winter (December 21 March 19). Two differen t capture protocols (tree traps only, and ground traps with intermittent tree traps) used in this study were taken into account to investigate the effect of capture

PAGE 18

18 method on capture probability. Individuals not caught in the traps but found during weekly nest box monitoring were added to the trapping dataset for that week. Because control and exclosure plots within each site share similar habitat characteristics, we also tested for a site effect on capture probability and apparent survival rate. M odels we re built for program MARK version 6.1 (White and Burnham 1999) by analyzing CMR data with R version 2.12.2 (R Development Core Team 2011 ) package RMark 2.0.1 (Laake and Rexstad 2008) Monthly apparent survival ( ) and capture probability ( p ) were estimated and modeled using Cormack Jolly Seber (CJS) models. We performed a goodness of fit test using the median approach in program MARK (White and Burnham 1999) and found no evidence of over dispersion or lack of fit ( = model comparison and statistical inference. We conducted preliminary analyses to identify the best base model for p and such that treatment effects co uld be evaluated in subsequent analyses. The best base model for p was investigated with time dependent taking into account the potential influence of trapping session, year, sex, capture method, season, site, and their additive and interactive effects on capture probability. The best supported model for p was then used to investigate the best base model for taking into account the potential influence of trapping session, year, sex, season, site, and their additive and interactive effects on survival. Effects of food supplementation, predator exclusion and prescribed fire on were then sequentially added to the best base model (i.e., p (sex + capture type) (site + year)).

PAGE 19

19 The effect of prescribed fire on small mammals is mainly through the changes in habitat brought about by the fire event (Smith 2000). Duration of this effect varies for different species ( Morris et al. 2011b; Morris et al. In press ) We tested for the different effect window of prescribed fire for southern flying squirrel by fixing the survival for all sessions except those following fire and allowing survival to vary for different length of time (3, 6 and 9 months) after the fire event. The best supported fire effect window on survival was then added to the base model to test for t he effect of fire on survival. We also tested for additive and interactive effect s of all three treatments (food supplementation, predator exclusion and prescribed fire; 2 way interactions only) on survival. Finally, we used multistate CMR models to estima te and model state specific (adult or juvenile) apparent survival ( S ) and state transition probability ( ) (Nichols et al. 1991) There was no evidence that capture probability varied between juvenile and adult squirrels; thus, we fixed p to p (sex + captu re type), and tested for seasonal, annual, sex strata and site specific variation in S and Probability of transition from adult to juvenile was fixed to zero. Data limitations did not allow us to test for treatment effects using the multistate CMR mo del. Results From July 2003 to November 2010, we captured 433 individual flying squirrels 1258 times. Of these, 556 were captured from May 2003 to May 2010 (ground traps with intermittent tree traps) and 702 were captured from June 2010 to November 2010 ( all tree traps). The sex ratio of captured individuals was 49.5% females and 50.5% males. Weekly nest box monitoring from June to October 2010 resulted in 59 (40

PAGE 20

20 females and 19 males) individuals being captured a total of 105 times. Only two females with l itters were found; with 4 and 3 offspring Initial investigations revealed that capture probability was best described by an additive effect of sex and capture type (ground traps vs. tree traps; Table 2 1). Capture probability was substantially higher for tree traps (0.80 SE 0.039) than for ground traps with intermittent tree traps (0.27 0.028); it was higher for females (0.59 0.0359) than for males (0.33 0.032). Overall monthly apparent survival was 0.84 0.011. There was no evidence for sex spe cific or seasonal variation in survival; instead, the model that included an additive effect of site and year received overwhelming support (Table 2 2). Survival varied over years in all four sites (Fig ure 2 1); with the higher survival at site 2 (0.88 0 .012) and lowest at site 4 (0.73 0.04) and highest in year 2004 (0.95 0.033) and lowest in year 2010 (0.72 0.024). The most parsimonious model testing for the food effect included an additive effect of food supplementation (Table 2 3a ). However, the effect of food supplementation was negative, with monthly apparent survival being lower during the period of food supplementation (food: 0.67 0.05; no food: 0.76 0.03). Furthermore, the model with an additive effect of supplemental food differed from 1.12, suggesting no strong evidence for the food supplementation effect on survival. Analyses to quantify the fire effect window indicated that fire affected survival for up to 6 months following the prescribed burn (Table 2 4 ). There was some evidence 2 3b ). Apparent monthly survival increased following burn s ( after fire: 0.91 0.05; before fire: 0.74 0.047)

PAGE 21

21 (Fi g ure 2 2 ). The base model (which did not include the effect of mammalian predator exclusion) had a lower AICc value than that for the model includ ing the effect of mammalian predator exclusion, suggesting that exclusion of mammalian predator s did not influence survival (Table 2 4 c ). Next, we evaluated the additive and interactive effects of food supplementation, predator exclusion and prescribed fire. The most parsimonious model included the additive effect of fire; adding additive or interactive effect of food supplementation and predator exclusion to the base model did not improve the model parsimony (Table 2 5). However, the most parsimonious model carried only 23% model weight, suggesting substantial model selection uncertainty. Based on multi state CMR model ing the apparent monthly survival was higher for adult s (0.85 0.01 ) than for juveniles (0.77 0.05). The t ransition probability (probability of juveniles reaching an adult weight) varied seasonally, and was highest in fall (0.92 0.06) and lowest in summer (0.23 0.13) (Fig ure 2 3). D iscussion Despite their wide geographic distribution (Dolan and Carter 1977) little is known about the population ecology of the southern flying squirrel. In fact, we were unable to find any study that reported survival or population growth rate for this species; this is surprising given its reputation as an abundant species th at is considered a pest in some areas (Borgo et al. 2010 ; Laves and Loeb 1999 ) Using a CMR approach, this study was undertaken to fill the gap in knowledge, and to provide sex and state specific (based on age class adult or juvenile) estimates of survival rates. Furthermore, we experim entally tested for the effect of food, fire and mammalian predators, arguably 3 of

PAGE 22

22 the most important factors affecting small mammal demography, on survival of southern flying squirrels. The overall capture probability in our study was 0.47, but it was su bstantially higher for tree traps compared to ground traps. Lehmkuhl et al (2006) in a study using both tree and ground traps reported a capture probability of 0.14 for northern flying squirrel s which is substantially lower than our estimates for southe rn flying squirrels. We found that females were more likely to be captured than males, an observation also reported for the Siberian flying squirrel ( Pteromys volans ) (Lampila et al. 2009). High energy requirement s of females during pregnancy and lactation could make food more attractive to females, resulting in increased capture probabilities in baited traps. The overall monthly apparent survival was 0.84, but it varied both among sites and years. These variations may reflect the influence of microhabitat differences among sites and/or environmental variation. However, apparent annual survival for a close relative, the northern flying squirrel, has been reported to be 0.47 0.51 in eastern Washington Casca de Range (Lehmkuhl et al. 2006) and 0.32 0.68 in the northern coast al range of Oregon (Gomez et al. 2005) Likewise, average annual adult survival for Siberian flying squirrels has been reported to be 0.43 0.53 (Lampila et al. 2009) The overall annual survival (calculated from the monthly survival) for our study population was 0.12, which is substantially lower than those r eported for other species of flying squirrels. The causes of the rather substantial difference s in survival among flying squirrel species (particularly between northern and southern flying squirrels) remain unclear. However, it should be noted that these a pparent survival estimates confound mortality and permanent immigration, and none of these studies addresses

PAGE 23

23 the issue of movement. Indeed, the relatively low survival rate for the southern flying squirrels in our study site is surprising, given the widely held belief that the southern flying squirrel populations are increasing in numbers in most of its range. Among the three factors considered in this study (fire, food and mammalian predator exclusion), only fire was found to substantially influence surviv al of southern flying squirrels. Apparent monthly survival was higher for up to 6 months after the burn in all four sites during all three burn years. The positive impact of fire on arboreal flying squirrels might be due to a decreased structural complexit y of understory vegetation after fire event s which facilitates locomotion and increases the ability of flying squirrels to detect and avoid predators ( Bendel and Gates 1987; Metcalfe 1984; Schooley e t al. 1996) ; an increase in abundance of food due to fire triggered growth of fungi (Fujimura et al. 2005) may also help improve survival post fire. Dense understory vegetation due to fire suppression may force squirrels to travel long distances for foraging, thereby increasing predation risk (Pasch and Koprowski 2011) The improvement in survival for up to 6 months suggests that the benefits to survival following fire disappear after about 6 months, likely due to a progressive increase in understory vegetation. However, we do not know if the observed increase in survival due to prescribed fire in our study is specific to our study site or occurs more widely, because no other study has evaluated the response of southern flying squirrels to prescribed fire. Survival of ground dwelling small mammals on the same study sites either showed no response (Cotton mice and oldfield mice, Morris et al. In press) or was negatively affected by prescribed fire (Cotton rat, Morris et al. 2011) Th ese differences may have occurred because of differences in resource needs between the southern flying squirrel, an arboreal species,

PAGE 24

24 and the ground dwelling small mammals. However, analysis of annual variation in apparent monthly survival showed that surv ival was highest in 2004, which was a non burn year, and lowest in 2007, which was a burn year. This probably implies that annual variation in environmental variables have greater influence than prescribed fire on the survival of the southern flying squirr el. There was limited evidence that food supplementation affected survival; furthermore, the effect was in the opposite direction than expected in that food supplementation reduced survival. Other small mammals (cotton rats, cotton mice and oldfield mice) on the same study sites had increased survival, increased rates of transition to reproductive states and greater abundance in response to food supplementation (Morris et al. 2011; Morris et al. In press) The lack of respons e to supplemental food by the flying squirrel s suggests that that food was not limiting southern flying squirrels during our study. Other species of squirrels have shown either no effect or only marginal effect of supplemental feeding on population paramet ers (Havera and Nixon 1980; Klenner and Krebs 1991 ; Sullivan 1990) but we know of no study reporting the negative effect of supplemental food. The decrease in flying squirrel survival could just be a coin cidence, with factors other than supplemental food influencing survival during the time of supplemental feeding, or it could be due to a higher prey activity attracting predator towards the feeding stations (Boutin 1990; Turner et al. 2008) Mammalian predator exclusion had no effect on survival of southern flying squirrel s Although mammalia n mesopredators were excluded from predator exclusion sites, access to these sites by avian predators and snakes was not affected; indeed,

PAGE 25

25 raptors and snakes, rather than mammalian mesocarniovores, are perhaps more important predators of these cavity dwell ing arboreal mammals ( Carey et al. 1992; Laves and Loeb 1999; Mitchell et al. 1999 ; Rudolph et al. 1990 ) Scat analysis of bobcat ( Lynx rufus ) at the present study site reported < 10% of the bobcat scats contained southern flying squirrel remains (Godbois 2003) ; o ther studies have reported flying squirrel as the most common prey species for spotted owl (Strix occidentalis) ( Carey et al. 1992 ; Forsman et al. 1984 ) Spotted owls are not present in this study site but there is plenty of barred owl ( Strix varia ) w hich would have similar impact like that of spotted owl. Studies have shown high use of cavity nests (Holloway and Malcolm 2007) and nest predation of various avian species by southern flying squirrel (Goertz et al. 1975; Kilham 1968; Stabb et al. 1989). T o avoid larger predators (such as raccoons and bobcats) southern flying squirrels apparently prefer cavities which are not yet enlarged by other cavity users ( Loeb 1993 ; Rudolph et al. 1990b ) and this makes them especially vulnerable to snake predation as the small entrance hole restricts their escape route once the cavi ty is intruded by snakes. It is intriguing that the southern flying squirrel had a lower apparent survival rate than its threatened counterpart, the northern flying squirrel considering that the former is often considered a pest species. It is well known that species characterized by early maturity and high reproductive rates have potential for rapid population growth in favorable circumstances ( Oli and Dobson 1999, 2003 200 5 ; Pianka 1970 ) The ability of southern flying squirrel population s to increase rapidly may, therefore, be facilitated by high reproductive rates and/or early maturity. In fact, flying squirrels can produce more than one litter a year (Dolan and Carter 1977). A detailed study of their population

PAGE 26

26 ecology is needed to understand the demographic basis and environmental effects of southern flying squirrel population dynamics

PAGE 27

27 Table 2 1 Model comparison table of top ten models for Cormack Jolly Seber capture mark recapture analysis to investigate the best base model for capture probability ( p ) for southern flying squirrels at Jones Ecological Research Center, Newton Georgia from year 2003 2010. For these analyses, the apparent survival rate ( ) was constrained to be time specific. Table includes AICc) and model weights (relative likelihood of models in the set). Model no Model K AICc Model weight 1 p (sex + capture type a ) 41 1311.545 0.000 0.633 2 p (sex capture type) 42 1313.636 2.091 0.223 3 p (sex + year) 47 1314.505 2.960 0.144 4 p (sex year) 54 1329.731 18.186 0.000 5 p (sex + time) 77 1333.625 22.080 0.000 6 p (capture type) 40 1336.894 25.349 0.000 7 p (year) 46 1339.033 27.488 0.000 8 p (time) 76 1362.750 51.206 0.000 9 p (sex + season b ) 43 1390.749 79.204 0.000 10 p (site c season) 54 1395.881 84.336 0.000 a capture type: Ground traps and tree traps b season: Winter (December 21 March 19), Spring (March 20 June 20), Summer (June 21 September 22) and Fall (September 23 December 20) c site: The mammalian predator exclosure plot, paired with a nearby control plots with similar habitat characteristics. Thus, there were 4 sites, each with 1 control and 1 exclosure plot.

PAGE 28

28 Table 2 2. Model comparison table of top 15 model for Cormack Jolly Se ber capture mark capture analysis of southern flying squirrels to investigate the best base model for survival ( ) in the Jones Ecological Research Center, Newton, Georgia from 2003 2010. For this analysis capture probability ( p ) was modeled as p (sex + capture type). See Table 1 for column definitions, and description of capture type, site, and season. Model no Model K AICc Model weight 1 (site + year) 14 1268.581 0.000 0.980 2 (site + season) 10 1276.850 8.268 0.016 3 (site season) 19 1279.896 11.314 0.003 4 (year) 11 1285.582 17.001 0.000 5 sex + year) 12 1287.650 19.069 0.000 6 (season) 7 1288.356 19.775 0.000 7 (season sex) 11 1290.001 21.419 0.000 8 (season + sex) 8 1290.402 21.821 0.000 9 (sex year) 19 1290.440 21.858 0.000 10 (site year) 35 1294.149 25.568 0.000 11 (site) 7 1307.689 39.108 0.000 12 (sex + site) 8 1309.232 40.651 0.000 13 (site sex) 11 1310.264 41.682 0.000 14 (time) 41 1311.545 42.963 0.000 15 (sex + time) 42 1313.820 45.239 0.000

PAGE 29

29 Table 2 3. Model comparison table of top four models for Cormack Jolly Seber capture mark recapture analysis of southern flying squirrels to investigate the effect of food supplementation, prescribed fire and mammalian predator exclusion on survival ( ) in the Jones Ecological Research Center, Newton, Georgia from 2003 2010. The best base model for was an additive effect of site and year. For these analysis capture probability ( p ) was modeled as p (sex + capture type). See Tabl e 1 for column definitions and description of site and capture type. Model no Model K AICc Model weight (a) Effect of food supplementation 1 (site + year + food ) 15 1267.460 0.000 0.606 2 (site + year) 14 1268.581 1.121 0.345 3 (site food + year) 18 1272.571 5.111 0.047 4 (site + year food ) 22 1280.387 12.93 0.001 (b) Effect of prescribed fire 1 (site + year + fire ) 15 1265.370 0.000 0.600 2 (site fire + year) 18 1267.720 2.355 0.185 3 (site + year ) 14 1268.581 3.216 0.120 4 (site + year fire ) 18 1269.060 3.691 0.095 (c) Effect of mammalian predator exclusion 1 (site + years) 14 1268.581 0.000 0.691 2 (site + year + predator exclusion ) 15 1270.578 1.997 0.254 3 (site + year predator exclusion ) 22 1274.686 6.105 0.033 4 (site predator exclusion + year) 18 1275.439 6.857 0.022

PAGE 30

30 Table 2 4. Model comparison table for Cormack Jolly Seber capture mark recapture analysis of southern flying squirrels to investigate the best fire effect window on survival ( ) in the Jones Ecological Research Center, Newton, Georgia in the 2005, 2007 and 2009. For these analysis capture probability ( p ) was modeled as p (sex + capture type). See Table 1 for column definitions. Model no Model K A ICc Model weight 1 ( fire effect over the period of 6 months) 15 1265.366 0.000 0.594 2 ( fire effect over the period of 3 months) 15 1266.420 1.054 0.351 3 ( fire effect over period of 9 months) 15 1270.113 4.748 0.055

PAGE 31

31 Table 2 5. Model comparison table of the top ten models for Cormack Jolly Seber capture mark recapture analysis of southern flying squirrels to investigate the additive and interactive effects of food supplementation, mammalian predator exclusion and prescribed fire on sur vival ( ) in the Jones Ecological Research Center, Newton, Georgia from 2003 2010. The best base model for was an a dditive effect of site and year For these analysis capture probability ( p ) was modeled as p (sex + capture type). See Table 1 for column definitions and description of site. Model no Model K AICc Model weight 1 (site + year + fire ) 15 1265.370 0.000 0.230 2 (site + year + fire + food ) 16 1265.780 0.412 0.187 3 (site + year + predator exclusion + fire ) 16 1267.400 2.033 0.083 4 (site + year + food ) 15 1267.460 2.094 0.081 5 (site + year + predator exclusion fire ) 17 1267.480 2.115 0.080 6 (site + year + fire food ) 17 1267.660 2.296 0.073 7 (site + year + predator exclusion + food + fire ) 17 1267.860 2.495 0.066 8 (site + year + predator exclusion fire + food ) 18 1268.58 0 3.212 0.046 9 (site + year) 14 1268.58 0 3.216 0.046 10 (site + year + predator exclusion + food ) 16 1269.54 0 4.172 0.029

PAGE 32

32 Figure 2 1. Monthly apparent survival estimates ( SE) of southern flying squirrels in the Jones Ecological Research Center, Newton, Georgia at four sites from 2003 2010. Survival was modeled with additive effects of site and year. The mammalian predator exclosure plot, paired with a nearby control plot with similar habitat characteristics constitutes a site.

PAGE 33

33 Figure 2 2. Comparison of monthly apparent survival estimates ( SE) of southern flying squirrels before and 6 months following prescribed fire at four sites in the Jones Ecological Research Center, Newton, Georgia during three years of prescribed fire (2005, 2007 and 2009). The effect of fire is modeled as an additive effect to the base model for survival (site + year). The mammalian predator exclosure plot, paired with a nearby control plot with similar habitat characteristics constitutes a site.

PAGE 34

34 Figure 2 3. Seasonal variation in transition probability ( SE) from juvenile to adult state for southern flying squirrels in the Jones Ecological Research Center, Newton, Georgia from 2003 2010. Individuals weighing less than 50 g were classified as juveniles

PAGE 35

35 CHAPTER 3 LONGLEAF PINE MANAGE MENT PRACTICES AND T HEIR IMPACT ON SMALL MAMMAL POPULATIONS Background The longleaf pine ( Pinus palustris ) ecosystem once covered more than 36 million ha bef ore the arrival of European settlers (Landers et al. 1995) This ecosystem has now been restricted to less than 1.2 million ha and is listed as critically endangered (Noss et al. 1995) Over millennia, anthropogenic an d lightning ignited frequent fire has been a dominant ecological process maintaining the longleaf pine ecosystem; disruption of natural fire regimes, along with other anthropogenic factors, is thought to have been an important cause of the rapid loss and d egradation of the longleaf pine ecosystem (Van Lear et al. 2005) Rich biologica l diversity of longleaf pine ecosystem supports hundreds of floral and faunal species (Alavalapati et al. 2002) ; more than 30 of these, including some notable species like gopher tortoise ( Gopherus polyphemus ) and red cockaded woodpecker ( Picoides borealis ) are now listed as either threatened or endangered (Landers et al. 1995) Abo ut 14% of mammal species inhabiting the longleaf pine ecosystem are identified as species of conservation concern ( Engstrom et al. 2001 ). Loss and degradation of this ecosystem is a serious threat to the long term persistence of many obligate species (Brockway et al. 2005) Prescribed fire during the growing season has been the primary management activity for the maintenance or restoration of this f ire dependent ecosystem (F rost 1990; Glitzenstein et al. 1995; Streng 1993) Every year 10 6 ha of forestlands are subjected to prescribed fire in the southeastern United States to maintain native ground cover vegetation and to control encroachment of hardwood (Cain et al. 1998; Richter et al. 1982 ) However, prescribed fire is under increased scrutiny due to perceived air quality

PAGE 36

36 issues, particularly the emission of atmospheric particles and their potential impacts on human health (Sandberg et al. 2002) A possibl e alternative to prescribed burning involves applications of herbicide alone or in combination with prescribed fire. Herbicide application, followed by prescribed fire, has been suggested to be more effective in restoring the longleaf pine ecosystem than the use of prescribed fire alone (Brockway and Outcalt 2000) The use of herbicide for modifying wildlife habitats and for inv asive plant control has increased in commercial pine plantation of southern United States (Miller and Miller 2004) There are concerns that prescribed fire may be removed, or replaced by herbicide, as a forest management tool without adequate understanding of ecological consequences of rep lacing (or augmenting) prescribed fire by herbicide. The population level response of small mammals to prescribed fire is fairly well understood (Arata 1959; Conner et al. 2011; Landers 1987; Morris et al. In press; Morris et al. 2011) but I am not aware of any study that has investigated the effect of herbicide, alone or in combination with prescribed fire, on small mamma ls inhabiting longleaf pine habitats. To fill this gap in knowledge, my goal was to provide data on population level impact of aforementioned longleaf pine ecosystem management practices on small mammals inhabiting the longleaf pine ecosystem. Specifically I aimed to test experimentally whether and to what extent survival of cotton rat ( Sigmodon hispidus ) and cotton mice ( Peromyscus gossypinus ) was affected by herbicide, prescribed fire and the combination of the two.

PAGE 37

37 Materials a nd Methods Study Area Th is study was conducted at the Joseph W. Jones Ecological Research Center in Baker county, Georgia. This 12,000 ha site is located on the upper coastal plain and is managed for research, education and conservation. This is primarily a longleaf pine and wire grass ( Arista stricta ) ecosystem with associated hardwood tree species, wetlands, slash pine and a variety of other habitats (Atkinson et al. 1996) Study S pecies Cotton mice are one of the most common and abundant semi arboreal small mammals of pine forests in southeastern United States (Clark and Durden 2002) The semi arboreal nature of this species makes sparse mat of grass, abundant woody materials and logs, and short over story an important microhabitat component (Mengak and Guynn 2003) Cotto n rats are also abundant across the southeastern United States. These solitary, crepuscular rodents occur in many habitats, but require thick a cover of dense grasses, herbs and shrubs for food and cover (Goertz 1964) Field Methods This study was conducted at four different sites within the Jones Ecological Research Center from April 1999 to April 2002. Each site, known as burn units, was divided into 3 plots with each plot receiving 1 of 3 treatments; 1) prescribed fir e alone, 2) herbicide application, followed by prescribed fire, and 3) herbicide application alone. A total of 12 treatment plots (3 treatment 4 replicates) was established. Size of each e plots were delineated using existing roads and firebreaks.

PAGE 38

38 Small mammal trap stations were placed along 3 systematically located transects within each plot. Ten trap stations were placed on each transect with 15 m between stations, and at each station t wo Sherman live traps (H.B. Sherman Traps, Tallahassee, Florida, USA) were placed approximately 1 m apart. During each trapping session, traps were set for four consecutive days. The traps were baited with rolled oats. Each captured of small mammal was ide ntified to species, weighed, sexed and released at the capture site. Cotton rats were marked individually with ear tags on both ears, and cotton mice were marked by toe clipping. Grids in each site were trapped every two months from April 1999 through Nove mber 2001, and after an interval of five month in April 2002. Herbicide was applied in February 2000 when vegetation was just beginning to leaf out. Hexazinone herbicide (Velpar R L, DuPont) was applied as spot treatment using backpack sprayers and metered spot gun applicators to achieve a rate of 2.5 to 3 quarts per acre. Application of prescribed fire was delayed due to draught condition in spring but burns were successfully carried out in June 2000. All field methods followed recommendations of the Americ an Society of Mammalogists (Animal Care and Use Committee 1998) Statistical Method and Capture Mark Recapture (CMR) A nalysis There were 17 4 day capture occasions from April 1999 to April 2002. Each year was divided into four calendar seasons; spring (March 20 June 2 0), summer (June 21 September 22), fall (September 23 December 20) and winter (December 21 March 19). Monthly apparent survival ( S ), capture probability ( p ) and transition probability ( ) from juvenile to adult were estimated and modeled using multis tate CMR models

PAGE 39

39 (Nichols et al. 1992; Williams et al. 2002) Cotton rats weighing > 50 g (Bergstrom and Rose 2004) and cotton mice weighing > 19 g (Bigler and Jenkins 1975) were classified as adults. Models were built for program MARK version 6.1 (White and Burnham 1999) with R package R Mark 2.13.0 (R Development Core Team 2011; Laake and Rexstad 2008). Goodness of fit tests using the median approach in program MARK provided no evidence for lack of fit for either species (cotton mice: = 0.997; cotton rat: = for model comparison and statistical inferences (Burnham and Anderson 2002; Williams et al. 2002) Prelimi nary analyses were conducted to identify appropriate base models for S p and such that treatment effect could then be added in the subsequent analyses. For all these analyses, transition from adult to juvenile was fixed to zero because such transitions are not biologically plausible. The base model for p was investigated with time dependent S and taking into account the potential influence of trapping session, year, sex, seasons, site, state and their additive and interactive effects on p The result ing best supported model for p along with time dependent was then used to investigate the base model for S taking into account the potential influence of factors that could potentially affect survival rate; trapping session, year, sex, season, site, s tate, and additive and interactive effects of site and year. Finally, the base model for p and S were used to investigate the appropriate base model for taking into account the effect of trapping session, year, sex, season and site. Changes in cover and food resources brought about by prescribed fire, herbicide application, and their combination, can last for months; thus, effects of these treatments

PAGE 40

40 could influence cotton mice and cotton rat survival for varied amounts of time (Morris et al. 2011a; Morris et al. 2011b ) Thus, we first investigated the best effect window for each of these treatments on survival of both species, using the base models identified in preceding analyses. For each treatment, we investigated three different effect windows (2, 4 and 6 months). As each of our study plots had either one of the three treatments applied, we used three categories of capture intervals: control (periods of time and sites when none of the treatments were in effect); treatment (sites where the given treatment was in effect for the treatment window in question); and other (periods of t ime and sites when the other two treatments could have been in effect, set with a broad window: June 2000 were allowed to vary for 2, 4 and 6 months of time following treatm ent in that particular treatment plot. The best supported effect window for three different treatments were compared with base models that contained no treatment effect. R esults Cotton M ice From April 1999 to April 2002, we captured 1083 individual cotton mice 3206 times. The sex ratio of captured individuals was slightly biased towards males (55.7%) than females (44.3%). Initial investigations revealed that cotton mice capture probability was best described when modeled with an additive effect of sex and c apture occasion; monthly apparent survival best described when modeled with an interactive effect of site and year; and probability of transition from juvenile to adult when modeled with annual variation in this parameter (Table 3 1). Overall monthly app arent survival was 0.78 SE 0.01. There was no evidence for sex specific or state specific variation in survival. However, apparent monthly survival of

PAGE 41

41 male (0.79 0.01) was slightly higher than of female (0.77 0.01), and survival of adult (0.79 0.01 ) was slightly higher than that of juveniles (0.73 0.03). Monthly apparent probability of transition from juvenile to adult was 0.83 0.05. Analysis to investigate best effect window for different treatments did not reveal strong support for any partic ular effect window (Table 3 2). We selected the top ranked model for further analyses: survival effect of fire lasting up to 2 months after fire event; that of herbicide lasting up to 2 months following herbicide application; and the survival effect of her bicide application, followed by fire lasting up to 6 months (Table 3 2). There was no evidence that herbicide application affected cotton mice survival. Fire, and a combination of herbicide and fire increased apparent monthly survival of cotton mice (Fig ur e 3 1), but the evidence for this effect was weak (Table 3 3). Cotton R at From April 1999 to April 2002, we captured 969 individual cotton rats 2218 times. The sex ratio of captured individuals was even with 51% female and 49% males. Initial analysis reve aled that cotton rat capture probability was best described when modeled with seasonal variation in this parameter; apparent monthly survival best described when modeled with time specific variation in this parameter; and transition probability from adult to juvenile best described when modeled with sex specific variation in this parameter (Table 3 4). Time dependent survival indicates that survival varies for each capture occasion, and this model was not helpful for testing for treatment effects. Also, her bicide and fire treatments were applied at two different seasons, and season and treatment effects could have been confounded. Thus we selected the top ranked model that did not include season and time effects (i.e., S (site + years)) as base model for tes ting treatment effects.

PAGE 42

42 Overall apparent monthly survival was 0.67 0.01. There was no evidence for sex specific or state specific variation in survival. Probability of transition from juvenile to adult was 0.91 0.04. Analysis to investigate best effec t window for different treatments did not reveal strong support for any particular effect window. We selected the top ranked model for further analysis: survival effect of fire lasting up to 6 months after fire event; that of herbicide lasting up to 6 mon ths following herbicide application; and survival effect of herbicide application followed by fire lasting up to 6 months (Table 3 5). There was strong evidence that prescribed fire and herbicide application followed by fire affected cotton rat survival bu t the evidence for effect of herbicide application on survival was weak (Table 3 6). The apparent monthly survival rate declined after the prescribed fire, herbicide application and herbicide fire combination (Fig ure 3 2). The effect of fire alone was stro nger than fire executed after herbicide spray. D iscussion Longleaf pine ecosystems are among the most biologically diverse habitats and harbor many species of conservation concern (Alavalapati et al. 2002; Landers et al. 1995) This ecologically important ecosystem is being lost to various anthropogenic activities at a rapid rate, and much of the remaining longleaf pine forests are seriously de graded due primarily to suppression of fire that historically maintained understory vegetation and kept hardwood encroachment at bay. Prescribed fire is a primary tool for the restoration of longleaf pine ecosystem but various factors associated with expan ding human settlement and economic constraint is increasingly limiting its application (Van Lear at al. 2005). Herbicide application, alone or in combination with fire, has been suggested as an alternative to prescribed fire. Studies investigating the ecol ogical effect

PAGE 43

43 of herbicide application in pine forest of southeastern United States are lacking and I sought to quantify effects of fire and herbicide treatments, alone and in combination, on survival of two widely distributed rodents in southeastern Unite d State. Although the effect of prescribed fire and herbicide fire combination on survival of cotton mice population was weak, these treatments were associated with an increase in cotton mice survival. Increase in local availability of food resources broug ht about by fire in the southeastern United States could be beneficial for cotton mice (Sharp et al. 2009) and this could ha ve increased the survival rate after fire events. Since the support for effect of these two treatments on survival of cotton mice was so weak, observed increase in survival rate could just be a coincidence, with other factors influencing survival. A study in longleaf pine stands in Georgia (Morris et al. In press) stream terrace hardwood stands in Louisiana (Blanchard 1991; Goatcher 1990) and Gulf Coastal plain o f south central Alabama (Sharp et al. 2009) revealed that prescribed fire did not have significant influence on the survival of cotton mice population. Greater versatility in refuge selection and preference for underground sites like stump holes and gopher tortoise ( Gopherus polyphemus ) burrows as refuge (Frank and Layne 1992) makes cotton mice less sus ceptible to direct effects of fire as well as the potential increase in mortality due to reduced vegetative cover after fire (Derrick et al. 2010) In contrast, there was evidence that fire, herbicide and herbicide followed by fire substantially reduced monthly survival of cotton rats. This was expected because cotton rats are highly sus ceptible to predation by different species of raptors, mammalian predators and snakes (Conner et al. 2011), and require heavy vegetative cover to avoid predation. Other studies have also shown that survival rate of cotton rats in longleaf

PAGE 44

44 pine ecosystem de clined after prescribed fire, primarily due to increased predation (Bock and Bock 1978; Conner et al. 2011; Morris et al. 2011) but we could not find any study reporting the survival effect of herbicide and herbicide fire combination in cotton rat population. Positive effect of increased local availability of food on survival rate as observed in cotton mi ce was missing in cotton rat. However, Rehmeier et al (2005) reported that, in the long run, fire application could be beneficial due to increase in local availability of food This suggests that for cotton rats, changes in habitat brought by fire event would be dertimental due to increased predation risk in the sho rt run but in the long run increased abundance in local availabilty food resources may be beneficial. Though all three treatments affected survival in the same direction, their strength differed, with prescribed fire having the strongest effect and herbici de application the weakest. The reasons behind these differences are unclear. I hypothesize that underlying differences in ecological processes brought about by these treatments had varied levels of influence upon the survival rate of cotton rats. A clear understanding of whether or not the herbicide and herbicide fire combination would have same beneficial effect in the long run as been reported with fire will require further study. Two species of small mammals with different natural history responded di fferently to different longleaf pine management practices. Prescribed fire had opposite effects in these two species although for both species the strength of the fire effect declined in response to prior herbicide application. There was some evidence for effect of herbicide application on survival of cotton rat but not on cotton mice. It is possible that even though herbicide alone may mimic some of the ecological effects of fire, it fails to reproduce the same ecological processes as the fire and the appl ication of herbicide

PAGE 45

45 prior to fire. Herbicide reduces understory vegetation but unlike fire cannot replace other ecological functions such as stimulating flowering and enhancing germination of many fire dependent species that characterize the longleaf ecos ystem (Brennan et al. 1998). Restoration of the longleaf pine ecosystem is not only about restoring the structure and function of the ecosystem but also the ecological processes. Various management practices for the restoration of longleaf pine ecosystem b ring about different changes in ecological processes such that they have varied influences on the population dynamics of species inhabiting it. Management activities to restore structure, function and diversity of longleaf pine come with the potential risk for some species of special interest (Van Lear et al. 2005). Long term studies are needed to evaluate the effects of management practices on other species of mammals, birds and reptiles to better understand the broader ecological impacts of these managem ent practices. Without clear understanding of these, no other management practice should be employed as an alternative to prescribed fire.

PAGE 46

46 T able 3 1. Model comparison table for multi state capture mark recapture analysis to investigate base model for capture probability (p), survival rate (S) and Research Center, Newton, Georgia from year 1999 to 2002. Table includes terion corrected for (relative likelihood of models in the set). Only top 5 models are shown for each variable. Model No Model K AICc Model weight Capture probability ( p ) 1 p (sex + time) 49 2857.685 0.000 1.000 2 p (season a ) 36 2877.717 20.032 0.000 3 p (season sex) 40 2884.881 27.196 0.000 4 p (.) 33 2899.125 41.440 0.000 5 p (year) 35 2900.794 43.109 0.000 Survival ( S ) 1 S (site b year) 45 2843.865 0.000 0.923 2 S (site + year) 39 2849.857 5.991 0.046 3 S (year) 36 2850.708 6.843 0.030 4 S (season) 37 2858.708 14.842 0.001 5 S (site) 37 2867.724 23.858 0.000 Transition ( ) 1 (year) 32 2811.428 0.000 0.798 2 (site) 33 2814.782 3.354 0.149 3 (.) 30 2818.078 6.650 0.029 4 (sex) 31 2819.790 8.362 0.012 5 (season) 33 2819.822 8.394 0.012 a season : Winter (December 21 March 19), Spring (March 20 June 20), Summer (June 21 September 22) and Fall (September 23 December 20). b site: Four different burn units within the Jones center. Note: Individuals weighing more than 19 g were classified as adults.

PAGE 47

47 T able 3 2. Model comparison table of multistate capture mark recapture analysis of cotton mice to investigate the effect win dow of prescribed fire, herbicide application and herbicide fire combination on survival (S) in Jones Ecological Research Center, Newton, Georgia in the year 2000. Effect windows of fire, herbicide and combined treatment was added to the base model for sur vival S (site year), capture probability was modeled as p (sex + time) and for column definitions. Model No Model K AICc AICc Model weight Fire effect window 1 S (fire effect up to 2 months) 34 2810.666 0.000 0.499 2 S (fire effect up to 4 months) 34 2811.198 0.531 0.381 3 S (fire effect up to 6 months) 34 2813.513 2.847 0.120 Herbicide effect window 1 S (herbicide effect up to 2 months) 34 2810.155 0.000 0.524 2 S (herbicide effect up to 6 months) 34 2811.732 1.577 0.238 3 S (herbicide effect up to 4 months) 34 2811.737 1.582 0.238 Combined effect window 1 S(effect up to 6 months) 34 2810.914 0.000 0.623 2 S (effect up to 2 months) 34 2812.713 1.780 0.253 3 S (effect up to 4 months) 34 2814.163 3.250 0.123

PAGE 48

48 T able 3 3. Model comparison table for multistate capture mark recapture analysis of cotton mice to investigate effect of prescribed fire, herbicide application and herbicide fire combination on survival (S) in Jones Ecological Research Center, Newton, Georgia in year 2000. The base model for S was interactive effect of site and year. For these analysis capture probability was modeled as p (s ex + time) and transition (juvenile (year). See table 1 for column definitions and description of site. Model No Model K AICc Model weight Effect of prescribed fire 1 S ((site year) + fire) 34 2810.666 0.000 0.719 2 S ((site year) + no treatment) 33 2812.544 1.877 0.281 Effect of herbicide 1 S ((site year) + no treatment) 33 2809.658 0.000 0.562 2 S ((site year) + herbicide) 34 2810.155 0.497 0.438 Effect of combined treatment 1 S ((site year) + combined treatment) 34 2810.914 0.000 0.691 2 S ((site year) + no treatment) 33 2812.526 1.612 0.309

PAGE 49

49 T able 3 4. Model comparison table for multi state capture mark recapture analysis to investigate base model for capture probability (p), survival rate (S) and transition (juvenile to adult) probability ( Research Center, Newton, Georgia from year 1999 to 2002. See table 1 for column definitions, description of site, season, and classific ation of adult and juvenile. Only top 5 models are shown for each variable. Model No Model K AICc Model Weight Capture probability ( p ) 1 p (season) 36 1665.190 0.000 0.972 2 p (.) 33 1673.116 7.926 0.018 3 p (sex) 34 1674.893 9.703 0.008 4 p (site) 37 1678.178 12.988 0.001 Survival ( S ) 1 S (time) 36 1665.190 0.000 1.000 2 S (season) 24 1701.667 36.477 0.000 3 S (site + year) 26 1717.285 52.095 0.000 4 S (year ) 23 1719.939 54.749 0.000 5 S (site ) 24 1721.549 56.360 0.000 Transition ( ) 1 (sex) 22 1638.486 0.000 0.654 2 (year + sex) 21 1640.942 2.456 0.192 3 (season + sex) 23 1641.950 3.464 0.116 4 (.) 24 1645.314 6.829 0.022 5 (year) 24 1645.782 7.297 0.017 Note: Only four models of capture probability were able to estimate all parameters of interest.

PAGE 50

50 T able 3 5. Model comparison table of multistate capture mark recapture analysis of cotton rat to investigate the effect window of prescribed fire, herbicide application and herbicide fire combination on survival (S) in Jones Ecological Research Center, Newton, Georgia in the year 2000. Effect windows of fire, herbicide and combined treatment was added to the base model for survival S (site + year), capture probability was modeled as p (season) and transit ion (juvenile to adult) probability as modeled as definitions. Model No Model K AICc Model weight Fire effect window 1 S (fire effect up to 6 month) 14 1673.031 0.000 0.971 2 S (fire effect up to 4 month) 14 1680.144 7.113 0.028 3 S (fire effect up to 2 month) 14 1686.941 13.910 0.001 Herbicide effect window 1 S (herbicide effect up to 6 month) 14 1691.231 0.000 0.551 2 S (herbicide effect up to 2 month) 14 1693.031 1.800 0.224 3 S (herbicide effect up to 4 month) 14 1693.031 1.800 0.224 Combined effect window 1 S (effect up to 6 month) 14 1676.264 0.000 0.551 2 S (effect up to 4 month) 14 1676.743 0.479 0.434 3 S (effect up to 2 month) 14 1683.476 7.211 0.015

PAGE 51

51 T able 3 6 Model comparison table of multistate capture mark recapture analysis of cotton rat to investigate effect of prescribed fire, herbicide spray and herbicide fire combination on survival (S) in Jones Ecological Research Center, Newton, Georgia in year 2 000. The base model for S was additive effect of site and time. For these analysis capture probability was modeled as p (seasons) and transition (juvenile See table 1 for column definitions and description of site Model No Model K AICc Model weight Effect of prescribed fire 1 S (site + year + fire) 14 1673.031 0.000 0.749 2 S (site + time + no treatment) 13 1690.837 17.806 0.000 Effect of herbicide 1 S (site + time + herbicide) 14 1691.231 0.000 0.303 2 S (site + year + no treatment) 13 1692.683 1.452 0.147 Effect of combined treatment 1 S (site + time + combined treatment) 14 1676.264 0.000 0.500 2 S (site + time + no treatment) 13 1684.116 7.851 0.010

PAGE 52

52 Figure 3 1. Effect of prescribed fire (a) and herbicide fire combination (b) on monthly apparent survival ( SE ) of cotton mice at Jones Ecological Research Center, Newton, Georgia in the year 2000. Sites are four areas within the center selected for the study.

PAGE 53

53 Figure 3 2. Effect of prescribed fire (a), herbicide application (b) and herbicide fire combination (c) on monthly apparent survival estimates ( SE) of cotton rat at Jones Ecological Research Center, Newton, Georgia in the year 2000. Sites are four are as within the center selected for the study.

PAGE 54

54 CHAPTER 4 CONCLUSIONS Long leaf pine ecosystem is regarded as one of the most treasured forest ecosystem of southeastern United States which supports hundreds of floral and faunal species (Noss et al. 1995; Alavalapati et al. 2002) Over millennia, anthropogenic and lightning ignited frequent fire has been a dominant ecological process maintaining the longleaf pine ecosystem. Various factors have contributed to the decline and degradation of this endangered e cosystem since European colonization. This rapid decline and degradation of longleaf pine ecosystem is a conservation concern. Prescribed fire during the growing season has been the primary management activity for the maintenance or restoration of this fir e dependent ecosystem (Frost 1990; Glitzenstein et al. 1995; Streng 1993) But this practice is under increased scrutiny due to perceived air quality issues. There are concerns that prescribed fire may be removed, or replaced by herbicide, as a forest mana gement tool without adequate understanding of ecological consequences of replacing (or augmenting) prescribed fire by herbicide. Beside these ecosystem levels management tools, population level management tools like food and exclusion of predator are also commonly employed for game bird management. The objective of this study was to evaluate the effect of these management practices on the population dynamics of small mammals inhabiting longleaf habitat. Prescribed fire which at present is a primary manageme nt activity for the restoration of longleaf ecosystem also was the most important factor influencing the population dynamics of small mammal species inhabiting longleaf pine ecosystem However, different species of small mammals responded differently to fi re event.

PAGE 55

55 P rescribed fire positively influenced survival rate of cotton mice and southern flying squirrel the survival rates of cotton rats was negative The changes in vegetation brought about by fire event were ben eficial for the flying squirrel and cotton mice but detrimental to cotton rat. There was no effect of herbicide alone on survival rate of cotton mice but herbicide fire combination increased the survival rate. Both herbicide alone and herbicide fire combi nation decreased the survival of cotton rat. We observed that ecological effect of herbicide and herbicide fire combination was weaker than that of fire alone. The populations of flying squirrel in our study system were not limited by food resources and m ammalian species are not a major predator for this arboreal flying squirrel i n our study system. Various management practices for the restoration of longleaf pine ecosystem bring about different changes in ecological processes such that they have varied influences on the population dynamics of species inhabiting it. This study suggests that herbicide could not mimic all the ecological function of fire and we do not suggest replacing prescribed fire with other management tools without clear understanding o f their ecological consequences.

PAGE 56

56 LIST OF REFERENCES ALAVALAPATI, J. R. R., G. A. STAINBACK, AND D. R. CARTER 2002 Restoration of the longleaf pine ecosystem on private lands in the US South: an ecological economic analysis Ecological Economics 40 : 411 419 ANIMAL CARE AND USE COMMITTEE 1998. Guidelines for the capture, handling, and care of mammals as approved by the American Society of Mammalogists. Journal of Mammalogy 79:1416 1431. ARATA, A. A 1959. Effects of burning on vegetation and rodent po pulation in longleaf pine turkey oak association in north central Florida. Quarterly Journal of the Florida Academy of Sciences 22 : 94 104. ATKINSON, J. B., B. JEAN, AND S. ROBERT 1996. Operational longleaf pine management at Ichauway. Pp 43 45 in Longleaf pine:a regional perspective of challenges and opportunities (J. S. Kush, ed.). The Longleaf pine Alliance Annual Meeting, Mobile, Alabama. BERGSTROM, B. J., AND R. K. ROSE 2004. Comparative life histories of Georgia and Virginia cotto n rats. Journal of Mammalogy 85 : 1077 1086. BENDEL, P. R., AND J. E. GATES 1987. Home range and microhabitat partitioning of the southern flying squirrel ( Glaucomys volans ). Journal of Mammalogy 68 : 243 255. BIGLER, W. J., AND J H. JENKINS. 1975. Populatio n characteristics of Peromyscus gossypinus and Sigmodon hispidus in tropical hammocks of south Florida. Journal of Mammalogy 56 : 633 644. BLANCHARD, V. J. 1991. Stream terrace hardwood forest as refuge for small mammals when adjacent pine forest is burned. M.S. thesis, Louisiana State University, Baton Rouge, LA. BOCK, C. E., AND J. H. BOCK. 1978. Response of birds, small mammals, and vegetation to burning sacaton grasslands in southeastern Arizona. Journal of Range Management 31 : 296 300. BORGO, J. S., M. R. CONOVER, AND L. M. CONNER 2010. Flying Squirrel removal does not reduce their use of simulated red cockaded woodpecker nest clusters. Southeastern Natu ralist 9 : 813 820. BOUTIN, S 1990. Food supplementation experiments with terrestrial vertebrates patterns, problems, and the future. Canadian Journal of Zoology 68 : 203 220. BRADY, M. J., T. S. RISCH, AND F. S. DOBSON 2000. Availability of nest sites does not limit population size of southern flying squirrels. Canadian Journal of Zoology 78 : 1144 1149.

PAGE 57

57 BRENNAN, L. A., R. T ENGSTROM, W. E. PA LMER, S. M. HERMANN, G. A. HURST, L. W. BURGER, AND C. L. HARDY 1998. Whiter wildlife without fire? Transaction of North American Wildlife and Natural Resource Conference 63:402 414. BROCKWAY, D. G., AND K. W. OUTCALT. 2000. Restoring longleaf pine wiregrass ecosystems: Hexazinone application enhances effects of prescribed fire. Forest Ecology and Management 137 : 1 21 138. BROCKWAY, D. G., K. W. OUTCALT, D. J. TO MCZAK, AND E. E. JOH NSON. 2005. Restoration of longleaf pine ecosystems.General Technical Report Southern Research Station, USDA Forest Service, Southern Research Station Asheville, NC. BURNHAM, K. P., AND D. R. ANDERSON. 2002. Model selection and multimodel inference: a practical information theoretic approach. 2nd ed. Springer Verlag, New York. CAIN, M. D., T. B. WIGLEY, AND D. J. REED. 1998. Prescribed fire effects on structure in uneven aged stands of lobl olly and shortleaf pines. Wildlife Society Bulletin 26 : 209 218. CAMPBELL, M. T., AND N. A. SLADE 1995. The effect of maternal mass on litter size and offspring survival in the hispid cotton rat ( Sigmodon hispidus ). Canadian Journal of Zoology 73 : 133 140. CAREY, A. B., S. P. HORTON, AND B. L. BISWELL 1992. Northern spotted owls: Influence of prey base and landscape character. Ecological Monographs 62 : 223 250. COLE, F. R., AND G. O. BATZLI 1978. Influence of supplemental feeding on a vole population. Journ al of Mammalogy 59 : 809 819. CLARK, K. L., AND L. A. DURDEN. 2002. Parasitic arthropods of small mammals in Mississippi. Journal of Mammalogy 83 : 1039 1048. CONNER, L. M., J. C. RUTLEDGE, AND L. L. SMITH 2010. Effects of mesopredators on nest survival of shrub nesting songbirds. Journal of Wildlife Management 74 : 73 80. CONNER, L. M., S. B. CASTLEBERRY, AND A. M. DERRICK. 2011. Effects of mesopredators and prescribed fire on hispid cotton rat survival and cause specific mortality. Journal of Wildlife Management 75 : 938 944. DERRICK, A. M., L. M CONNER, AND S. B. CASTLEBERRY. 2010. Effects of prescribed fire and predator ex clusion on refuge selection by Peromyscus gossypinus Le Conte (Cotton Mouse). Southeastern Naturalist 9 : 773 780. DOLAN, P. G., AND D. C. CARTER 1977. Glaucomys volans Mammalian Species : 1 6.

PAGE 58

58 ENGSTROM, R.T., L. K KIRKMAN, AND R. J. MITCHELL. 2001. Natural history: longleaf pine wiregrass ecosystem. In: Wilson, J.R. (Ed.). The Fire Forest: Longleaf Pine Wiregrass Ecosystem. Georgia Wildlife, Natural Georgia Series 8 (no. 2). Georgia Wildlife Federation, pp. 5 18. FORSMAN, E. D., E. C. MESLOW AND H M W IGH T 1984. Distribution and biology of the spotted owl in Oregon. Wildlife Monographs 87: 3 64. FRANK, P. A., AND J. N. LAYNE. 1992. Nests and daytime refugia of cotton mice ( Peromyscus gossypinus ) and golden mice ( Ochrotomys nuttalli ) in south central Flori da. American Midland Naturalist 127 : 21 30. FROST, C. C. 1990. Natural diversity and status of longleaf pine communities, Pp. 26 35 in Forestry in the 1990s a changing environment. Proceedings of the Regional Technical Conference (G. Youngblood, Frederick, D.L, ed.). Society of American Foresters, Pinehurst, NC. FUJIMURA, K. F., J. E. SMITH, T. R. HORTON, N. S. WEBER, AND J. W. SPATAFORA 2005. Pezizalean mycorrhizas and sporocarps in ponderosa pine (Pinus ponderosa ) after prescribed fires in eastern Oregon, USA. Mycorrhiza 15 : 79 86. GANNON, W. L., R. S. SIKES AND THE ANIMAL CARE AND USE COMMITTEE OF THE AMERICAN SOCIETY OF MAMMALOGISTS 2007. Guidelines of the American Society of Mammalogists for the Use of Wild Mammals in Research. Journal of Mammalogy 88 : 809 823. GLITZENSTEIN, J. S., W. J. PLATT, AND D. R. STRENG 1995. Effects of fire regime and habitat on tree dynamics in north Florida longleaf pine savannas. Ecological Monographs 65 : 441 476. GOATCHER, B. L. 1990.A preliminary investigation of the importance of stream terrace hardwood forests as refuge for small mammals to escape fires. M.S. theisis, Louisiana State University, Baton Rouge, LA. GODBOIS, I. A., L.M.CONNER AND R.J.WARREN 2003. Bobcat diet on an area managed for northern bobwhite. Proceedings of the annual conference Southeastern Association of Fish and Wildlife Agencies 57 : 222 227. GOERTZ, J. W. 1964. The influence of habitat quality upon density of cotton rat populations. Ecological Monographs 34 : 359 381. GOERTZ, J. W., R. M. DAWSON, AND E. E. MOWBRAY 1975. Response to Nest Boxes and Reprod uction by Glaucomys volans in Northern Louisiana. Journal of Mammalogy 56 : 933 939. GOMEZ, D. M., R. G. ANTHONY, AND J. P. HAYES 2005. Influence of thinning of Douglas fir forests on population parameters and diet of northern flying squirrels. Journal of Wildlife Management 69 : 1670 1682.

PAGE 59

59 HANSEN, L. P., AND G. O. BATZLI 1979. Influence of supplemental food on local populations of Peromyscus leucopus Journal of Mammalogy 60 : 335 342. HAVERA, S. P., AND C. M. NIXON 1980. Winter feeding of fox and gray squir rel populations. Journal of Wildlife Management 44 : 41 55. HOLLOWAY, G. L., AND J. R. MALCOLM 2007. Nest tree use by northern and southern flying squirrels in central Oontario. Journal of Mammalogy 88 : 226 233. HUBBS, A. H., AND R. BOONSTRA 1997. Populatio n limitation in Arctic ground squirrels: Effects of food and predation. Journal of Animal Ecology 66 : 527 541. KILHAM, L 1968. Reproductive behavior of hairy woodpeckers II. nesting and habitat. The Wilson Bulletin80 : 286 305. KLENNER, W., AND C. J. KREBS 1991. Red squirrel population dynamics. I. The effect of supplemental food on demography. Journal of Animal Ecology 60 : 961 978. KREBS, C. J., ET AL 1995. Impact of food and predation on the snowshoe hare cycle. Science 269 : 1112 1115. LAAKE, J., AND E. REXTAD 2009. RMark an alternative approach to building linear models in MARK. Pp. C1 C115 in Program MARK: a gentle introduction (E. Cooch and G. White, eds.). http:// www.phidot.org/software/mark/docs/book/. Accessed 15 January 2011. LAMPILA, S., R. WIST BACKA, A. MAKELA, AND M. ORELL 2009. Survival and population growth rate of the threatened Siberian flying squirrel ( Pteromys volans ) in a fragmented forest landscape. Ecoscience 16 : 66 74. LANDERS, J. L. 1987. Prescribed burning for managing wildlife in southeastern pine forests. Pp. 19 27 in Managing southern forests for wildlife and fish a proceedings (J. G. Dickson and 0. E. Maughan,eds.). United States Department of Agriculture, Forest Service, General Technical Report, SF 50 65:I1 85 LANDERS, J. L., D. H. VANLEAR, A ND W. D. BOYER. 1995. The longleaf pine forests of the southeast requiem or renaissance. Journal of Forestry 93 : 39 44. LAVES, K. S., AND S. C. LOEB 1999. Effects of southern flying squirrels Glaucomys volans on red cockaded woodpecker Picoides borealis reproductive success. Animal Conservation 2 : 295 303. LEHMKUHL, J. F., K. D. KISTLER, J. S. BEGLEY, AND J. BOULANGER 2006. Demography of northern flying squirrels informs ecosystem management of western interior forests. Ecological Appli cations 16 : 584 600. LOEB, S. C. 1993. Use and selection of red cockaded woodpecker cavities by southern flying squirrel. Journal of Wildlife Management 57 : 329 335.

PAGE 60

60 MCNAMARA, J. M., AND A. I. HOUSTON. 1987. Starvation and predation as factors limiting population size. Ecology 68 : 1515 1519. MENGAK, M. T., AND D C. GUYNN. 2003. Small mammal microhabitat use on young loblolly pine regeneration areas. Forest Ecology and Management 173 : 309 317. MESERVE, P. L., J. R. GUTIERREZ, AND F. M. JAKSIC 1993. Effects of vertebrate predation on a caviomorph rodent, the degu ( Octodon degus ), in a semiarid thorn scrub community in Chile. Oecologia 94 : 153 158. METCALFE, N. B. 1984. The effects of habitat on the vigilanc e of shorebirds is visibility important. Animal Behaviour 32 : 981 985. MILLER, K. V., AND J H. MILLER. 2004. Forestry herbicide influences on biodiversity and wildlife habitat in southern forests. Wildlife Society Bulletin 32 : 1049 1060. MITCHELL, L. R., L. D. CARLILE, AND C. R. CHANDLER 1999. Effects of southern flying squirrels on nest success of red cockaded woodpeckers. Journal of Wildlife Management 63 : 538 545. MITCHELL, P., W. ARTHUR, AND M. FARROW 1992. An investigation of population limitation us ing factorial experiments. Journal of Animal Ecology 61 : 591 598. MORRIS, G., L. M. CONNER, AND M. K. OLI 2011a. Effects of mammalian predator exclusion and supplemental feeding on space use by hispid cotton rats. Journal of Mammalogy 92 : 583 589. MORRIS, G ., J. HOSTETLER, L. MIKE CONNER, AND M. OLI 2011b. Effects of prescribed fire, supplemental feeding, and mammalian predator exclusion on hispid cotton rat populations. Oecologia : 1 12. MORRIS, G., J. HOSTETLER, L. MIKE CONNER, AND M. OLI In press. Effects of predation, fire, and supplemental feeding on populations of two species of Peromyscus mice. Journal of Mammalogy. NICHOLS, J. D., J. R. SAUER, K. H. POLLOCK, AND J. B. HESTBECK 1992. Estimating transition probabilities for stage based population projection matrices using capture recapture data. Ecology 73 : 306 312. NORRDAHL, K., AND E. KORPIMKI 1995. Does predation risk constrain maturation in cyclic vole populations? Oikos 72 : 263 272. NOSS, R. F., E. T. L AROE III, AND J. M. SCOTT. 1995. Endanger ed ecosystems of the United States: a preliminary assessment of loss and degradation. National Biological Report, Washington, D.C. OLI, M. K., AND F. S. DOBSON 1999. Population cycles in small mammals: the role of age at sexual maturity. Oikos 86 : 557 565.

PAGE 61

61 OLI, M. K., AND F. S. DOBSON 2003. The relative importance of life history variables to population growth rate in mammals: Cole's prediction revisited. American Naturalist 161 : 422 440. OLI, M. K., AND F. S. DOBSON 2005. Generation time, elasticity patte rns, and mammalian life histories: A reply to Gaillard et al. American Naturalist 166 : 124 128. PASCH, B., AND J. L. KOPROWSKI 2011. Impacts of fire suppression on space use by Mexican fox squirrels. Journal of Mammalogy 92 : 227 234. PIANKA, E. R 1970. On r and K Selection. The American Naturalist 104 : 592 597. PYNE, S. J., P. L. ANDREWS, AND R. D. LAVEN 1996. Introduction to wildland fire. John Wiley & Sons, Inc., New York. R DEVELOPMENT CORE TEAM 2009. R: a language and environment for stati stical computing in R Foundation for Statistical Computing, Vienna, Austria. http://www.r project.org. Accessed 15 January 2011. RANSOME, D. B., AND T. P. SULLIVAN 1997. Food limitation and habitat preference of Glaucomys sabrinus and Tamiasciurus hudsoni cus Journal of Mammalogy 78 : 538 549. RANSOME, D. B., AND T. P. SULLIVAN. 2004. Effects of food and den site supplementation on populations of Glaucomys sabrinus and Tamiasciurus douglasii Journal of Mammalogy 85 : 206 215. REHMEIER, R. L., G. A. KAUFMAN, D W. KAUFMAN, AND B. R. MCMILLAN. 2005. Long term study of abundance of the hispid cotton rat in native tallgrass prairie. Journal of Mammalogy 86 : 670 676. RICHTER, D. D., C. W. RALSTON, AND W. R. HARMS. 1982. Prescribed fire effects on water quality and forest nutrient cycling. Science 215 : 661 663. RUDOLPH, D. C., H. KYLE, AND R. N. CONNER 1990. Red cockaded woodpeckers vs rat snakes the effectiveness of resin barrier. Wilson Bulletin 102 : 14 22. SANDBERG, D. V., R. D. OTTMAR, J. L. PET ERSON, AND J. CORE. 2002. Wildland fire in ecosystems: Effects of fire on air. General Technical Report RMRS GTR 42 vol. 5. Ogden, UT: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. p. 79 SCHNELL, J. H 1968. Limiting effects of natural predation on experimental cotton rat populations. Journal of Wildlife Management 32 : 698 711.

PAGE 62

62 SCHOOLEY, R. L., P. B. SHARPE, AND B. VANHORNE 1996. Can shrub cover increase predation risk for a desert rodent? Canadian Journal of Zoology 74 : 157 163. SHARP, N. W., M. S. MITCHEL L, AND J. B. GRAND. 2009. Sources, sinks and spatial ecology of cotton mice in longleaf pine stands undergoing restoration. Journal of Mammalogy 90 : 1440 1448. SINCLAIR, A. R. E., AND R. P. PECH 1996. Density dependence, stochasticity, compensation and predator regulation. Oikos 75:164 173. SMITH, J. K 2000. Wildland fire in ecosystems. Effects of fire on fauna, General Technical Report Rocky Mountain Research Station, USDA Forest Service : v + 83 STABB, M. A., M. E. GARTSHORE, AND P. L. AIRD 1989. Interactions of southern flying squirrels, Glaucomys volans and cavity nesting birds. Canadian Field Naturalist 103 : 401 403. STRENG, D. R., J. S. GLITZENSTEIN, AND W. J. PLATT 1993. Evaluating sea son of burn in longleaf pine forests: a critical literature review and some results from an ongoing longterm study, Pp. 227 263 in Proceedings of the Tall Timbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee, FL. SULLIVAN, T. P. 1990. Responses of red squirrel ( Tamiasciurus hudsonicus ) populations to supplemental food. Journal of Mammalogy 71 : 579 590. TAITT, M. J., AND C. J. KREBS 1983. Predation, cover, and food manipulations during a spring decline of Microtus townsendii Jour nal of Animal Ecology 52 : 837 848. TURNER, A. S., L. M. CONNER, AND R. J. COOPER 2008. Supplemental feeding of northern bobwhite affects red tailed hawk spatial distribution. Journal of Wildlife Management 72 : 428 432. VAN LEAR, D. H., W. D. CARROLL, P. R. KAPELUCK, AND R. JOHNSON 2005. History and restoration.of the longleaf pine grassland ecosystem: Implications for species at risk. Forest Ecology and Management 211 : 150 165. WEIGL, P. D., M. A. STEELE, L. J. SHERMAN, AND J. C. HA 1989. The ecology of the fox squirrel (Sciurus niger) in North Carolina:implications for survival in the southeast. Bulletin of the Tall Timbers Research Station 24:1 93. WILLIAMS, B. K., J. D. NICHOLS, AND M. J. CONROY 2002. Analysis and management of animal populations. Aca demic Press. San Diego, California. WHITE, G. C., AND K. P. BURNHAM 1999. Program MARK: survival estimation from populations of marked animals. Bird Study 46, supplement:120 138.

PAGE 63

63 BIOGRAPHICAL SKETCH Binab Karmacharya is Fulbright scholar from Nepal. He went to veterinary school at Tribhuvan University, Nepal and got his bachelor degree in v eterinary s cience in 2000. After getting his veterinary degree, Binab worked as a veterinary officer at the Central Zoo, Kathmandu Nepal for 6 years. He did a gra duate certificate course in conservation and started his master s in w ildlife e cology and c o nservation at University of Florida in fall 2009