1 WADING BIRD (CICONIIFORMES) RESPONSE TO FIRE AND THE EFFECTS OF FIRE IN THE EVERGLADES By LOUISE S. VENNE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREM ENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Louise S. Venne
3 To all those with inquiring minds
4 ACKNOWLEDGMENTS Many people have provided advice and mentorship, help in the field, mathematical and statistic al guidance, translation, technical expertise, critical edits, and support. I thank the following people for their role they played in making this dissertation happen: E. Anderson, R. Borkhataria, M. Brown, B. Burtner, Y. Chen, S. Coates, J. Colee, B. Faus tini, J. Fidorra, E. Fishel, E. Gaiser, L. Garner, T. Glover, C. Hansen, M. Johnston, K. Kerr, J. Kline, W. Loftus, J. Mansuetti, E. Posthumus, M. Schlothan, T. Schrage, J. Seavey, G. Smith, C. Stiegler, N. Vitale, A. Williams, C. Winchester, and K. Yaguch i. A special debt of gratitude goes to J. Simon for his ingenuity, wisdom, and patience. I thank M. Ward and M. Juntunen with Florida Fish and Wildlife Conservation Commission for their willingness to share burn information, opportunities to burn with them and general support of my research. I thank my committee for challenging me: P. Frederick, W. Kitchens, L. Kobziar, T. Osborne, and J. Trexler. I also thank my friends and family for their love and support.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 2 EFFECTS OF FI RE ON WETLAND DEPENDENT WILDLIFE: A REVIEW ........... 16 Introduction ................................ ................................ ................................ ............. 16 Mammals ................................ ................................ ................................ ................ 19 Avians ................................ ................................ ................................ ..................... 21 Direct Mortality Result ing from Fire ................................ ................................ .. 21 Use of Vegetation Post Burn ................................ ................................ ............ 22 Use of Burned Wetlands for Foraging ................................ .............................. 30 Opportunistic for aging during fire ................................ ............................... 30 Foraging after fire ................................ ................................ ....................... 30 Foraging during migration ................................ ................................ .......... 31 Use of Burned Wetlands for Nesting ................................ ................................ 33 Herpetofauna ................................ ................................ ................................ .......... 39 Indirect Effects of Prescribed Burning ................................ .............................. 39 Presence and Abundance ................................ ................................ ................ 39 Breeding ................................ ................................ ................................ ........... 41 Lessons Learned ................................ ................................ ................................ .... 43 3 EFFECTS OF PRESCRIBED FIRE ON FORAGING BY WADING BIRDS (CICONIIFORMES) IN THE EVERGLADES ................................ ........................... 67 Introduction ................................ ................................ ................................ ............. 67 Study Area ................................ ................................ ................................ .............. 69 Methods ................................ ................................ ................................ .................. 70 Prey Item Survey ................................ ................................ .............................. 70 Prey Density ................................ ................................ ................................ ..... 71 Foraging Observations ................................ ................................ ..................... 72 Foraging Habitat Selection ................................ ................................ ............... 72 Statistical Analysis Prey ................................ ................................ ................. 74 Statistical Analysis Foraging Observations ................................ .................... 74 Statistical Analysis Foraging Habitat Selection ................................ .............. 75 Results ................................ ................................ ................................ .................... 75
6 Foraging Habitat Selection ................................ ................................ ............... 75 Foraging Observations ................................ ................................ ..................... 77 Prey Item Survey ................................ ................................ .............................. 79 Prey Density ................................ ................................ ................................ ..... 79 Discussion ................................ ................................ ................................ .............. 80 4 EFFECTS OF FIRE ON PERIPHYTON PRIMARY PRODU CTION AND FISH STANDING STOCK IN AN OLIGOTROPHIC WETLAND ................................ ..... 101 Introduction ................................ ................................ ................................ ........... 101 Methods ................................ ................................ ................................ ................ 104 Results ................................ ................................ ................................ .................. 110 Nutrients ................................ ................................ ................................ ......... 110 Environmental Factors ................................ ................................ .................... 110 Periphyto n ................................ ................................ ................................ ...... 111 Overall Fish Metrics ................................ ................................ ........................ 111 Fish Community Response ................................ ................................ ............ 113 Discussion ................................ ................................ ................................ ............ 114 5 CONCLUSIONS ................................ ................................ ................................ ... 137 APPENDIX: WHITE IBIS ( EUDOCIMUS ALBUS ) AND SNOWY EGRET ( EGRETTA THULA ) CAPTURE EFFICIENCIES AND CAPTURE RATES .............................. 139 LIST OF REFERENCES ................................ ................................ ............................. 141 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 154
7 LIST OF TABLES Table page 2 1 Selected references of fire effects on wetland dependent avian species ........... 53 3 1 Description of prescribed burns conducted by the F WC in WCA 3A used for wading bird foraging observ ations and /or prey studies in 2009 2011 .............. 85 3 2 Great egret habitat selection ratio (Bonferroni adjusted 95% confidence interval ) for 2009 ................................ ................................ ................................ 86 3 3 Great egret habitat selection ratio (Bonferroni adjusted 95 % confidence interval) for 2010 ................................ ................................ ................................ 87 3 4 White ibis habitat selection ratio (Bonferroni adjusted 95 % confidence inte rval) for 2009 ................................ ................................ ................................ 88 3 5 White ibis habitat selection ratio (Bonferroni adjusted 95% confidence interval) for 2010 ................................ ................................ ................................ 89 3 6 Capture r ates and capture efficiencies reported for the great egret ( Ardea alba ) in southern Florida marshes ................................ ................................ ...... 90 3 7 Capture rates and capture efficiencies of great egret ( Ardea alba ) in 2009 and 2010 i n WCA 3A of the Everglades, USA ................................ .................... 91 3 8 Candidate set of models of great egret capture rate using corrected AICc of foraging locations in WCA 3A of the Everglades, USA, 2009 and 2010 ............. 92 3 9 Coefficients of generalized linear models of great egret capture rate and capture efficiency selected using AIC ................................ ................................ 93 3 10 Candidate s et of models of great egret capture efficiency using QAICc of foraging locations in WCA 3A of the Everglades, USA, 2009 and 2010 ............. 94 3 11 Mean of environmental variables and aquatic organisms s ampled with 1 m 2 throw trap and minnow trap in WCA 3AS of the Everglades, USA, in 2011 ....... 95 4 1 Mean of environmen tal variables measured in plots ................................ ......... 120 4 2 Summary of ANCOVAs testing differen ces due to treatment and period ......... 121 4 3 Summary of responses of biotic variables to treatments ................................ .. 122 4 4 Frequency of capture of aquatic organisms in minnow traps by treatment plot and species in the Everglades, 2010 ................................ ................................ 123
8 4 5 Summary of generalized least squares r egression examining response of fish measures to treatment and sampling period ................................ .................... 124 4 6 Mean of fish captured in 1 m 2 throw traps ................................ ........................ 126 4 7 Summary of Analysis of Variances examining response of all and individual fish species captured in throw traps to light and nutrient treatments ................ 127 4 8 Characteristics of fish species caught in at least 80% of plots sampled ........... 128 4 9 Summary of ANOSIM (Analysis of Similarities) results testing differences of relative abundance ................................ ................................ ........................... 129 A 1 Summary of capture rates and capture effic iencies reported for white ibis and snowy egret in south ern Florida marshes ................................ ......................... 139 A 2 Capture rate and capture efficiency of whi te ibis ( Eudocimus albus ) in 2009 and 2010 in WCA 3A of the Everglades, USA ................................ .................. 140
9 LIST OF FIGURES Figure page 1 1 Simplified food web model in the Ev erglades illustrating hypotheses (numbered hypotheses tested in Chapters 3 and 4 ................................ ............ 15 2 1 Number of studies per year of fire ef fects on each group in wetlands ................ 66 3 1 Map of study area including prescribed burns conducted in 2009 2011 used in various components of this study ................................ ................................ .... 96 3 2 Habitat selection ratio for great egre ts ( Ardea alba ) in 2009 in the central Everglades, USA ................................ ................................ ................................ 97 3 3 Habitat selection ratio for great egrets ( Ardea alba ) in 2010 in the central Everglades, USA ................................ ................................ ................................ 98 3 4 Habitat selection ratio for white ibis ( Eudocimus albus ) in 2009 in the central Everglades, USA ................................ ................................ ................................ 99 3 5 Habitat selection ratio for white ibis ( Eudocimus albus ) in 2010 in the central Everglades, USA ................................ ................................ .............................. 100 4 1 Experimental design showing tr eatments ................................ ......................... 130 4 2 Concentrati ons of TP and SRP in water sampled collected pre burn a nd post burn in treatment plots ................................ ................................ .............. 131 4 4 C haracteristics of Flagfish ( Jordanella floridae ) captured in minnow traps post burn in W CA 3A S of the Eve rglades, FL USA ................................ ......... 133 4 5 Characteristics of Sailfin Mollies ( Poecilia latipinna ) captured in minnow traps post burn in WCA 3A S of the Everglades, F L USA ................................ ......... 134 4 6 C haracteristics of Least Killifish ( Heterandria formosa ) captured in minnow traps post burn in W CA 3A S of the Everglades, F L USA ................................ 135 4 7 C haracteristics of Eastern Mosquitofish ( Gambusia holbrooki ) captured in minnow traps post burn in W C A 3A S of the Everglades, F L USA ................... 136
10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy WADING BIRD (CICONIIFORMES) RESPONSE TO FIRE AND THE EFFECTS OF FIRE IN THE EVERGLADES By Louise S. Venne Augu st 2012 Chair: Peter Frederick Major: Wildlife Ecology and Conservation Despite considerable knowledge about fire effects on wildlife in uplands, there is a relative paucity of information about fire effects on wetland dependent wildlife Many wetland co mmunities are pyrogenic, and even those that rarely experience wildfire naturally are often burned with prescribed fires. Fire in wetlands was initially conducted for the purpose of benefiting waterfowl and muskrat. Since then, there is recognition that ot her species such as sparrows, wading birds, and salamanders are affected by fire, at least on a short term basis. Wading birds may benefit from fire through the exposure of prey after vegetation removal or through a trophic response to added nutrients and light resulting from fire. I determined whether wading birds select for and benefit by foraging in burned areas in the central Everglades. Great egrets and white ibis selected for burned ridges and adjacent sloughs and avoided areas of dense, tall, unburn ed sawgrass. Great egrets had higher capture rates in sloughs adjacent to burns than in burns, but were more efficient at capturing prey in burned areas than in the adjacent sloughs. Prescribed fires created short term shallow water habitats with limited s ubmerged and emergent vegetation, apparently making prey more accessible.
11 Fire releases nutrients and increases light via the combustion of vegetation. I manipulated light and nutrients in a 2x2 factorial experiment to determine fire effects on primary pro duction and standing stock of fish in the oligotrophic wetlands of the Everglades. I used prescribed burns (nutrient s ) and mowing with removal of vegetation (no nutrients) to manipulate nutrients. T o manipulate light, I constructed shade houses (no light) to limit light and left other plots open (light). Significantly greater periphyton cover and mass (dry weight) per area was observed in the N utrients + Light treatment than in other treatments. Fish generally did not respond to treatments, but least killif ish ( Heterandria formosa ) had larger individuals while flagfish ( Jordanella floridae ) and sailfin mollies ( Poecilia latipinna ) had smaller individuals in nutrient treatments. Increases in size may equate to increased reproductive output or to differences i n age structure of fish using these areas. Fire apparently augment s primary production, how ever fish response was limited.
12 CHAPTER 1 INTRODUCTION Fire is a natural process in many wetlands that helps maintain the structure, function, and communities of t hese wetlands (DeBano et al. 1998) Fire return intervals of wetlands range from approximately once per year to once every 300+ years, depending on the hydrological cycle, vegetative growth, and other environmental factors. Fire typically resets succession in wetlands, maintai ning species associations typical for the wetland. Our knowledge of the effects of fire on wetland dependent wildlife is limited in scope. In this dissertation, I report three studies aimed to improve our knowledge of fire effects on foraging ecology of wa ding birds. These studies include a literature review that identifies existing knowledge of effects of fires on wildlife in wetlands (Chapter 2), an observational study on wading bird foraging in burned areas that addresses benefits of foraging in burned a reas (Chapter 3), and an experimental manipulation of burned habitats to fire effects on wading bird prey (Chapter 4). Much of the early literature of fire effects on wildlife is observational in nature due to the lack of control treatments and replication (e.g., Lynch 1941, Givens 1962, Zontek 1966) waterfowl production. Since the mid 1990s, a need to understand the effects fire ha s on target and non target species has resulted in many more studies of fire effects on wildlife in wetlands. In spite of this trend, studies on the effects of fire on wildlife are still very limited (Chapter 2). I review the available literature of fire e ffects on wetland dependent wildlife to illustrate how fire in wetlands impacts wildlife in comparison to fire in uplands.
13 Wading bird selection of foraging areas is driven largely by water depth, prey availability, and vegetation density (Bancroft et al. 2002, Gawlik 2002, Lantz et al. 2010, Pierce and Gawlik 2010, Lantz et al. 2011) Changes in any of these affect foraging success of wading birds. Anecdotal observations by fire management specialists and scientists of wading birds (Order Ciconiiformes) foraging in burned areas suggest that these birds may benefit from burns. Fire removes vegetation (exposing additional areas containing prey resources), releases nutrients and increases light, and changes foraging hab itat, potentially attracting wading birds. I generated four hypotheses to test whether wading birds preferred foraging in burns and if they benefited by foraging in these burned areas (Chapter 3). I hypothesized that wading birds would select for burned ar eas more than unburned areas (H 1 ; Fig. 1 1 ). I hypothesized that fires make prey available by injuring or killing prey during the burn (H 2 ). I also hypothesized that prey densities would be greater in burned than unburned sawgrass because of increased prim ary production post burn resulting from light and nutrients (H 3 ). Finally, I hypothesized that wading birds would have a higher capture rate (captures per minute) and capture efficiency (captures per attempt) in burned areas than in unburned areas (H 4 ) Fi re effects on the aquatic community in wetlands are relatively unknown. Increases in nutrients and light stimulate primary production (Mosisch et al. 2001) and p rovide additional food resources to primary consumers. If aquatic consumers are food limited, increased food resources may lead to an increase in their size, nutritional value, or abundance, any of which could benefit predators such as wading birds. In Cha pter 4, I investigate whether the release of nutrients and increase of light to the underlying
14 substrate by fire increases periphyton primary production with a concordant response by the fish community. I hypothesized that an increase in light and nutrient s would result in more periphyton biomass and cover (H 5 ). I also hypothesized that total and individual fish size, condition factor, and relative abundance would increase, assuming that periphyton biomass increased (H 6 ). If light and nutrients post fire do not result in a subsequent increase in periphyton, there is then little evidence to suggest that fire increases primary productivity. Rather, wading birds and other predatory animals may respond to burns because prey are easier to catch or attracted to re cently burned areas for reasons other than an increase in primary productivity.
15 Figure 1 1. Simplified food web model in the Everglades illustrating hypotheses (numbered hypotheses (e.g., H 1 ) correspond to dashed lines to illustrate pathways) tested in Chapters 3 and 4. Lines with arrows indicate direction of influence. Box indicates the realm of hydrologic influence on this food web.
16 CHAPTER 2 EFFECTS OF FIRE ON W ETLAND DEPENDENT WILDLIFE: A REVIEW Introduction Fire is a natural disturbance in many up land systems that affects nutrient cycling, plant species composition, pest and pathogen prevalence, and wildlife use and movements on the landscape (Whelan 1995) Fire is also a natural disturbance in many wetland systems. Occurrence and frequency of fire can be limited by environmental conditions, with fire often starting during periods of drought or a drop in water levels (DeBano et al. 1998) As in terrestrial systems, fire can affect succession in wetlands (e.g., Wharton et al. 1982, Kantrud et al. 1989, Laderman 1989, Gagnon 2009) resulting in a shift in vegetation composition and maintenance of function in the wetland. Effects of fire are also dependent on the timing of the fir e and conditions of the wetland (e.g., water levels). However, our understanding of fire effects on wetland dependent wildlife is limited and while inferences from upland studies may be drawn, sufficient differences of fire effects between uplands and wetl ands exist to warrant further study of fire effects on wetland dependent wildlife. Fire frequency in wetlands is largely dependent on environmental conditions such as hydrology, unlike terrestrial habitats (Mitsch and Gosselink 2007) Fire frequency is imp ortant for sustaining the structure and function in many wetlands. For example, feedbacks between fire and hydrology reinforce the dome shape of isolated cypress domes (Watts et al. 2012) Regular fires maintain wetland structure and dynamics in pyrogenic wetlands such as the Everglades. This wetland experiences a high density of lightning (Orville and Huffines 2001) that ignites wi ldfires just before the onset of the wet season when water depths are typically at their lowest level (Slocum et al. 20 07) The
17 dominant wetland vegetation grows quickly and senesces, a growth form conducive to spreading frequent fires (Wade et al. 1980) Les s frequent fires can also help maintain structure and function of certain wetlands. In northern climates, peatlands and bogs set in forested ecosystems burn on the same infrequent time scale as the surrounding forest (DeBano et al. 1998) helping maintain the wet, anoxic condi tions that perpetuate this type of wetland. If fires are too frequent species such as Atlantic white cedars may be eliminated via additional fires post germination ( Laderman 1989) Prescribed fires in terrestrial systems tend to be conducted outside of the natural fire season, altering the expected effects of fire on the ecosystem (Cox and Widener 2008) This appears true f or wetlands also, since prescribed fires are often conducted in winter when lightning is less prevalent (Orville and Huffines 2001) Prescribed fires are designed to burn fuels only above the surface of the water and avoid ignition of the peat soi ls. Such fires are frequently conducted to create early succession habitat for wildlife such as muskrat and waterfowl. Caution is taken when planning prescribed fires to minimize damage to nesting waterfowl while increasing food for wildlife (Lynch 1941, Hoffpauir 1961) Thus, prescribed fires can be used to achieve goals that may not be met through letting wil dfires burn. In contrast to prescribed fires, wildfires that occ ur when water levels drop or during droughts may result in peat fires. Fire in peatlands impacts the vegetative structure, peat depth, and nutrient availability (DeBano et al. 1998) Peat fires typically are impos sible to control, but change vegetation composition and provide deep water habitat which may be important for certain wildlife species such as diving ducks and turtles. However, deep peat fires can also eliminate desired wetland species and alter
18 vegetatio n composition (e.g., Atlantic white cedar swamp to deciduous hardwoods; Laderman 1989) While fire has long been recommended as a management tool in wetlands for waterfowl habitat enhancement and increased food quality for herbivorous species (e.g., Lynch 1941, Givens 1962, Lugo 1995, Nyman and Chabreck 1995) o nly more recently have studies started to quantify the effects of fire on other avian species and other wetland dependent wildlife (Fig. 2 1). Effects of fire on upland species may partially be used as a guide for what to expect for wetland speci es. Birds often target prey fleeing the flame front (e.g., Tewes 1984) herbivorous species take advantage of nutritious regrowth and granivores of increased seed or mast production (Lyon et al. 2000) and other species target insects that exploit weakened or killed vegetation (e.g., Warren et al. 1987, Cox and Widen er 2008, Hutto 2008) In addition to food resources, changes in habitat structure and cover affect how species utilize wetlands, increasing use for species that prefer open areas or sparse vegetation and decreasing use for species that prefer dense cover. While more work to understand responses to fire by birds and other species in grasslands and forests is still necessary (Warren et al. 1987, Russell et al. 1999, Pilliod et al. 2003, Saab and Powell 2005) our unde rstanding of fire effects on wetland dependent wildlife lags far behind our knowledge in uplands. Much work is still needed if fire is used to manage wildlife habitat and minimize unintended consequences on species of management concern and non target wild life. The literature of fire effects on wetland dependent wildlife is fairly limited, recent, and primarily focused on presence/absence and abundance of avian species. While species presence and abundance post fire is important to determine whether species
19 respond to fire, understanding the mechanism for the response is much more informative in making decisions regarding the use of fire for the purpose of management of wildlife species and wetland ecosystems. However, f ew studies have looked at underlying c auses to the responses to fire by species studied. Kirby et al. (1988) reviewed fire effects in wetlands, creating an annotated bibliography of peer reviewed and gray literature publications. For an extensive, although somewhat dated, bibliography of fire effects on wetla nd systems and effects on wildlife, I recommend readers consult Kirby et al. (1988) Subsequently, Mitchell et al. (2006) thoroughly reviewed the effects of fire and other management strategies in coastal marshes on birds. Russell et al. (1999) and Pilliod et al. (2003) reviewed the herpetofauna literature, illustratin g how scant our knowledge is of fire effects on herps in wetlands. Since these reviews, a number of studies have been published that begin to address some of the gaps in our knowledge. While m any studies of fire in wetlands investigate the effect of fire o n wetland vegetation from which we may be able to draw some conclusions about wildlife response, direct and indirect effects on wildlife are typically not included in these studies, leaving many unanswered questions about wildlife response. In this review, I focus on ecological effects on wildlife of fire. A summary table of studies can be found in Table 2 1. Mammals Only a handful of studies of fire effects on mammals in wetlands exist, despite mention of the ease of trapping some furbearing species post f ire and the use of fire in marshes to enhance forage for cattle (e.g., McAtee et al. 1979) I found 8 studies of fire effects on mammals in wetlands, most published between 1940 and 1970 (Fig. 2 1). Fire effects studies from this era typically did not include information indicating
20 experimental rigor had been applied to observations or management suggestions. F ire typically does not cause mortality of muskrats or deer because they have escape strategies such as taking refuge or fleeing (Lynch 1941) However, fire re moves cover, exposing the mammals inhabiting these marshes. Humans often burned marshes so furbearers such as muskrat could be more easily managed and trapped by concentrating them in the limited remaining cover in marshes (Lynch 1941, Singleton 1951, Givens 1962, Perkins 1968, Ward 1968) Burning vegetative cover in wetlands results in poor habitat for most rodents until vegetation regrows (Tewes 1984) Wetlands in high altitude areas of Kamberg Nature Reserve, South Africa are burned triennially and represent areas of high smal l mammal populations and richness, including the preferred habitat of South African vlei rat ( Otomys irroratus ; Bowland and Perrin 1993) Natal Mastomys ( Mastomys natalensis ) was captured in wetland s only after burning, indicating that changes due to fire either exposed or benefited this species, however Bowland and Perrin (1993) did not give reasons for this response. Other small mammal populations initially dec lined post burn due to reduced cover and food supply. Herbivores such as muskrat and deer supposedly benefit from fire due to increased nutritive content and marsh grass production and an increase in preferred food plants, respectively (Lynch 1941, Loveless and Ligas 1959, Smith et al. 1984) Beavers have also long been thought to benefit from fires through regeneration of woody forage, however, spring fires that burn up to the edge of wetlands appear to cause beaver lodge abandonment by reducing habitat quality in Elk Island National Park, Alberta Canada (Hood and Bayley 2003, Hood et al. 2007) A single fire resulted
21 in abandonment of lodges for multiple years post burn and a dditional fires resulted in further lodge abandonment. Interaction of fire with high levels of herbivory and drought further exacerbated reductions in habitat quality (Hood and Bayley 2003, Hood et al. 2007) Thus, frequent fires could significantly reduce beaver populations in this habitat by reducing forage already limited by herbivory, rather than benefiting this species with increases in woody plant regrowth. Avians Avians are the longest and most studied group i n regards to response to fire in wetlands (Fig. 2 1). I found 3 3 published studies of fire effects on avians in wetlands, largely on sparrows and waterfowl. While gaps in our knowledge of fire effects on upland species exist (Saab and Powell 2005) we know much less about the effect of fire in wetlands on avians inhabiting these ecosystems. The first reports on the use of fire to manage for wetland species started with waterfowl. Thereafter, effects on other wetland dependent species were reported. Most studies focus on the effects of changes in vegetation affect presence and abundance of avian species post burn, yet effects on foraging and nesting are being incorporated into studies Direct Mortality Resulting from Fire Instances of direct mortality due to fire appear to be rare. Typically, wildlife can avoid mortality via fire whether by fleeing or taking refuge in burrows, underwater, or densely vegetated moist areas. However, mortality due to fire does occur. During a wildfire, approximately 50 adult white ibis ( Eudocimus albus ) with fire charred feathers were found dead in desiccated and brown cattail (Epanchin et al. 2002) Epanchin et al. (2002) suggest scenarios for the death of these birds, including debilitation due to smoke inhalation, taking refuge from the fire in the cattail stand before it burned, or
22 foraging close to the fire line as smoke and flames c orralled and drove prey items. While birds taking refuge in vegetation near a burn may seem counterintuitive when sloughs with open water would be a safer refuge, other birds have similarly been reported to take refuge in a wet area of marsh, resulting in mortality when fire burnt the refugia (Legare et al. 1998) Use of Vegetation Post Burn Waterfowl (Family Anatidae) were some of the first species suggested to benefit from the use of fire to manage vegetation in and around wetlands (Lynch 1941, Givens 1962, Schlichtemeier 1967, Perkins 1968, Ward 1968) Lynch (1941) reported that experimental fall burns in the Chenier Plains along the coast in southwestern Loui siana increased the abundance of snow geese ( Chen caerulescens ) and also attracted Canada ( Branta canadensis ) and white fronted geese ( Anser albifrons ). This supposedly was the first time snow geese had been seen at Lacassine National Wildlife Refuge in so uthwestern Louisiana. While Lynch (1941) gave no indication of presence or abundance of these species in unburned areas, abundance of ducks, particularly mallards ( Anas platyrhynchos ) and northern pintails ( A. acuta ), were noted as being in the t housands in the burned areas. Hochbaum et al. (1985) found no difference in waterfowl use of wetlands with burned and unburned edges (within 10 m of wet meadow edge) after spring and fall fi res in southern Manitoba and southeastern Saskatchewan, Canada. Waterfowl benefit from changes in vegetative structure if burns are timed appropriately, as discussed in a later section. Sparrows typically inhabit dense, grassy areas so changes in vegetativ e cover can be expected to affect sparrow presence and abundance post burn. After a winter prescribed fire in the Chenier Plains in coastal southwestern Louisiana, most sparrows
23 (i.e., seaside ( Ammodramus maritimus tailed ( A. nelsoni ), and swamp ( Melospiza georgiana )) did not return to burned areas for at least a year or more (Gabrey et al. 1999) Gabrey and Afton (2004) showed that sparrows are typically at lower abundances during the first breeding season post burn on plots burned in winter than prior to the burn, but by the second year have returned to pre burn abundances. When vegetative structure returned to pre fire levels, sparrows utilize d recently burned habitat and in some cases, were found in higher abundances in the second year post burn than in unburned plots (Gabrey and Afton 2000) In the third y ear, sparrows per plot remained at abundances similar to year two post burn, corroborating the importance of dead vegetation structure for sparrows (Gabrey et al. 2001) Similarly, Vogl (1973) found swamp and song ( M. melodia ) sparrows in greater abundance within one year in the unburned part of a pond shoreline in the panhandle of Florida compared to the burned portion. However, no further observations were provided and no knowledge of habitat, foraging resources, long term use, or behavior were provided. In shrub/scrub wetlands in northern Minnesota, Hanowski et al. (1999) counted more clay colored ( Spizella pallida ), savannah ( Pas serculus sandwichensis A. leconteii ) sparrows in managed (i.e., sheared and/or burned) than unmanaged burned treatments (treatment blocks grouped into 0 3 yea rs post treatment, and 3+ years post treatment) may benefit this species, but not significantly more than other treatments (p=0.053). Based on the yearly changes of sparrow responses in the Chenier Plains, it is likely that categorizing treatments into 3 y ear time blocks obscured responses by species.
24 Wintering sparrows that tend to breed in upland habitats responded differently to fires in wetlands than wetland dependent sparrows. White throated sparrows ( Zonotrichia albicollis ) were observed in greater ab undance on the burned than unburned shoreline of a pond (Vogl 1973) A. henslowii ) wintered at higher densities in Gulf Coast pitcher plant bogs the first year post burn than in subsequent years post burn (Tucker and Robinson 2003) Fires during the growing season appeared to attract sparrows for more years, suggesti ng that growing season burns may benefit sparrows longer. However, forb density and availability of seed stalks Fire and season of burn can be used to manipulate availability of food resources. Wrens in wetlands tended to avoid recent burns while there was little or no vegetation. In coastal marshes in the Chenier Plains, sedge wrens ( Cistothorus platensis ) were not found during the winter in areas burned the previous month, b ut were present the following winter in approximately an equal number of burned and unburned plots (Gabrey et al. 1999) During the breeding season, Gabrey et al. (2001) found no difference in number of wrens (i.e., marsh ( Cistothorus palustris ) and sedge wrens) per survey among years or treatments in the three years after winter prescribed burns in the Chenier Plains. However, sedge wrens were absent from burned plots the first summer, but were in these plots in subsequent years. Marsh wrens were encounte red in burned plots in the first month and again, one year later, after a winter burn, but were more frequently observed in unburned plots in both years. Four months after a late winter burn, sedge wrens (named grass wren in this study) were detected in bu rned Juncus marshes in the Mar Chiquita Biosphere Reserve, Argentina, and after six
25 months abundances were similar between burned and unburned treatments in Juncus marshes (Isacch et al. 2004) Sedge wrens were not detected in burned Spartin a marshes. Height of Juncus in burned plots was not significantly different from unburned plots by summer, corresponding to the recovery of abundances of sedge wren, whereas height of Spartina remained significantly shorter in burned than unburned plots fo r the duration of the study. In shrub/scrub wetlands in northern Minnesota, sedge wrens were more abundant on managed sites (managed sites were treated with combinations of shearing and burning) than unmanaged sites (Hanowski et al. 1999) Sites that had been sheared, but not burned were not significantly different than sites with a combination of burning and shearing within the previous 0 3 years. The time scale of this study was very coarse (i.e., categories of 3 years) so seasonal or yearly changes in wren abundance in these shrub /scrub wetlands could not be discerned, but provides a longer term outlook on wren response to habitat changes. Many icterids nest in tall, thick vegetation or trees, but forage in more open areas, accounting for the increase in abundance in burns immediat ely post burn, and greater abundances in unburned areas during the breeding season. B oat tailed grackles ( Quiscalus major ) in the Chenier Plains were observed in burned plots immediately post burn, but not the following winter (Gabrey et al. 1999) Red winged blackbirds ( Agelaius phoenicius ) were twice as abundant in burned as unburned plots immediately post burn but were still present in burned areas the fol lowing winter. During the spring breeding and nesting season in these same study plots, icterids tended to be more numerous during the second spring after the burn than the first and third spring, but the differences were not significant (Gabrey et al. 2001) Vogl (1973) did not find any
26 difference in abundances of red winged blackbirds after part of the shoreline of a po nd was burned in the winter. Yellow winged blackbirds ( Agelaius thilius ) in salt marshes in the Mar Chiquita Biosphere Reserve, Argentina were only detected in unburned Spartina but did not differ in burn treatments in Juncus marshes (Isacch et al. 2004) Secretive marsh birds typically require dense marsh vegetation for breeding and foraging. Flooding in river deltas is a natural disturbance that typically removes decadent vegetation and resets succession of vegetation. Changes in flood reg imes of wetlands have negatively impacted many species adapted to these disturbances, including the US endangered Yuma clapper rail ( Rallus longirostris yumanensis ). Prescribed burns were conducted as a surrogate management tool for flooding during late wi nter or early spring in the Colorado River Delta over the course of six years and monitored for secretive marsh bird use (Conway et al. 2010) Detection probabilities were calculated to determine whether changes in vegetation structure a ffected detectability on control and burn plots and impacted interpretation of results. Yuma clapper and Virginia ( Rallus limnicola ) rails were more numerous in burned areas post burn than pre burn, an increase that only occurred in burned areas. As time since burn increased, the difference i n abundance pre and post burn diminished. Virginia rails declined to pre burn abundances faster than Yuma clapper rails. Other secretive marsh birds (California black rail ( Laterallus jamaicensis coturniculus ; listed as endangered in Mexico and threatened in California), sora ( Porzana carolina ), and least bittern ( Ixobrychus exilis )) did not differ between burned and unburned areas. Conway et al. (2010) suggest that dense vegetation reduces foraging efficiency and prey availability. Also, rails are likely mo re susceptible to predation if they have to walk on top of matted
27 dead vegetation to reach foraging and nesting locations Fire or flooding removes these thick mats of dead vegetation. Prior to this study, managers of the delta have not been allowed to bur n habitat critical to and inhabited by an endangered species. Conway et al. (2010) demonstrated that prescribed fire in fact increases Yuma clapper and Virginia rail abundance without negatively impacting other secretive marsh birds. Gallinag o delicata ) typically avoid wetlands with dense, tall vegetation (Mueller 1999) suggesting that burns should be beneficial to snipe until by (Lynch 1941) though no direct comparison was made with unburned areas. Snipe were more numerous in burned than in unburned areas along the shore of a pond (Vogl 1973) and in isolated wetlands in the R ainwater Basin in Nebraska (Brennan et al. 2005) In Rainwater Basin wetlands, snipe were also obse rved in more burned wetlands post burn than pre burn (Brennan et al. 2005) corroborating evidence that snipe burned areas provide open areas for foraging. A number of other species were observed during comparison of bird use of burned and unburned wetlands and can be grouped into categories of species that use open areas and species that prefer dense v egetation. Species that use open areas were more abundant in burned than unburned wetlands. Vogl (1973) reported higher abundances of common crow ( Corvus brachyrhynchos ), mourning dove ( Zenaida macroura ), northern cardinal ( Cardinalis cardinalis ) great blue h eron ( Ardea herodias ), little blue heron ( Egretta cerulea ), snowy egret ( Egretta thula ), tricolored heron ( Egretta tricolor ), and other species on burned than unburned shorelines of a pond. Similarly, southern lapwings ( Vanellus chilensis ) used recently bu rned salt marshes in Mar
28 Chiquita Biosphere Reserve, Argentina, but were present only briefly in Juncus marshes and only until vegetation in Spartina marshes started sprouting (Isacch et al. 2004) Correndera pipits ( Anthus correndera ) were only detected in burned plots. Pipits were briefly observed in Juncus marshes during the first two months of surveys whereas in Spartina marshes they persisted for the duration of the study. Since burned Spartina did not regrow as quickly as Juncus (Isacch et al. 2004) species such as pipits which are typically found in grasslands may have benefited from the shorter vegetation, accounting for their persistence in burned Spartina While it is likely that the reduction of vegetation provided a dditional areas in which these birds could forage, other aspects of the fire may have attracted these species to use burned wetlands. Many other species typically found in wetlands with thick vegetation were not commonly observed in recently burned wetland s. Species typically associated with shrubs were more abundant on unmanaged than managed (i.e., shearing and burning) schrub/scrub wetlands whereas species that are associated with emergent vegetation wetlands were more abundant in managed wetlands (Hanowski et al. 1999) Common yel lowthroat ( Geothlypis trichas ), which typically are described as skulking through marsh vegetation, were not seen in burns in the Chenier Plains along coastal southwestern Louisiana until the winter following early winter prescribed burns (Gabrey et al. 1999) The red capped wren spinetail ( Spartonoica maluroides ), a wetland associated species, and freckle breasted thornbirds ( Phacellodomus striaticollis ) app eared in Juncus marshes in Mar Chiquita Biosphere Reserve, Argentina three months post burn (Isacch et al. 2004) Shortly thereafter, thornbirds were at similar abundances and by one year post burn wren spinetails was equally abundant in bur ned
29 and unburned Juncus In Spartina marshes, wren spinetails appeared four months post burn, but remained at a lower abundance in burned than unburned Spartina for the year these plots were monitored. Crakes were only observed in unburned Spartina marshes in the year following spring burns. While dot winged crake ( Porzana spiloptera ) were also observed in unburned Juncus marshes, speckled crake ( Coturnicops notata ) were never seen in Juncus marshes. Burns were only observed for one year so it is impossible to know whether crakes used recent burns sometime after the study finished. Had this study continued more than a year, differences in abundances of many species in Spartina plots might not have been seen. At the end of the study, Juncus marshes had recove red their vegetative structure, but Spartina marshes had not, likely resulting in differences of species presence and abundances noted between marsh types. Patches of vegetation during and after a fire serve as refugia and potentially as a source for recol tailed (in winter) and swamp sparrows (in summer) were only detected in burned plots that contained some unburned vegetation (Gabrey et al. 2001) suggesting that a complete burn that does not leave patches of vegetation may make the entire burned area unsuitable for these species. Similarly, dusky seaside sparrows ( A. mirab ilis nigrescens ) were displaced for approximately six months after wildfires burned part of St. Johns NWR marsh in winter, but then defended territories within the burned area shortly thereafter (Baker 1974) Unburned patches appear to have provided cover for these species between the time when fires occurred and breeding. Black rails used unburned patches of vegetation within a prescribed burn conducted during the su mmer on the St. Johns National
30 Wildlife Refuge in Florida (Legare et al. 1998) A more complete winter prescribed burn (~90% of area burned) resulted in mortality of black rails that had taken refuge in wetter, vegetated portions of the marsh left by a previous burn (Legare et al. 1998) Use of Burned Wetlands for Foraging Opportunistic foraging during fire Burning exposes prey resources for many species, making prey more vulnerable to predation than they are in dense vegetative cover. Blackbirds (Family Icteridae), swallows (Family Hirundinidae), gulls (Family Laridae), and raptors (Family Accipitridae) have been observed flying through smoke of a spring prescribed f ire in coastal wetlands to catch prey such as insects and small mammals (Lynch 1941, Stevenson and Meitzen 1946, Tewes 1984) This is similar to observations during forest and grass fires where many avian species fo rage on prey fleeing the flames (Komarek 1969) This phenome non likely occurs in most wetlands during fires and may serve as a beneficial resource pulse to species responding to the disturbance. To the best of my knowledge, the benefits of this behavior have not been quantified. Foraging after fire Burning of wetla nds can enhance food resources for species, however, timing and availability of food resources is an important consideration. In coastal marshes in southeastern US wildlife refuges, Perkins (1968) and Givens (1962) found that fall to early winter burns result ed in succulent browse and heavy use of recently burned areas by geese. In the first study with a reported control of waterfowl response to burning, Gabrey et al. (1999) conducted five aerial surveys of burned and unburned plots for white geese (i.e., readily visible from the air) over 2 months beginning immediately after winter prescribed burns in the Chenier Plains. Ten flocks of white geese (lesser snow
31 C. rossii )) were observed in burned sites (eight flocks) and in grit sites (2 flocks) adjacent to a burned area, ranging from 300 to 17,500 birds per flock, while no flocks were observed in unburned areas. Summer or early fa ll burns (in the Gulf Coast wintering grounds) often attract thousands of wintering snow geese that may forage on plant roots (Hoffpauir 1961) Gabrey et al. (1999) suggest that by removing above ground v egetation through burning plant roots were accessible to foraging geese. Consumption of roots will prevent plant regrowth if root stocks are reduced substantially (Hoffpauir 1961) and can shift species composition (Gauthier et al. 2004) Almost no studies of foraging benefits post fire to birds other than waterfowl have been conducted. Marsh harriers ( Circus spilonotus ) did not hunt as much over burned sections of the Watarase Marsh, J apan during the first year post burn as they did over unburned marsh area (Hirano et al. 2003) In the second year post burn when reeds had regrown, harriers foraged equally over burned and unburned areas, indicating that their prey had returned. Foraging during migration Most research on the use of burned wetlands during migration has focused on waterfowl species. Many waterfowl species use recently burned areas in the season after the fire. Waterfowl used burned wetlands in the Sandhills of Nebra ska during spring migration after prescribed burning in winter (Schlichtemeier 1967) Waterfowl use of the burn continued into summer and fall because of the open water and edge effect created by burning. In Phragmites stands in the Delta Marsh, a late summer prescribed burn substantially reduced stem den sity resulting in large congregations of waterfowl in the fall and heavy use by ducks for nesting for multiple years following the burn (Ward 1968) Brennan et al. (2005) explored the effect of spring prescribed burns near or up to
32 the edge of isolated wetlands in the Rainwater Basin in Nebraska. Pairs of burned and unburned wetlands were surveyed within seven days prior t o and within seven days after the burn. Ducks and other waterfowl (excluding geese) were detected in burned and unburned wetlands pre and post burn, but did not appear to respond to burn treatments. Timing of wetland fires appears less important for attr acting migrating geese than other waterfowl due to differences in forage preferences. Prescribed b urns in the Rainwater Basin were conducted after peak snow goose migration, yet geese were nearly twice as abundant in wetlands during post burn than pre burn surveys (Brennan et al. 2005) Migrating snow geese were observed in fewer burned wetlands post bu rn than pre burn, but were observed in an equal number of burned and unburned wetlands post burn. Given that wetlands were surveyed within seven days post burn, it is unlikely that geese were attracted by succulent new growth, but rather were foraging on r eadily accessible roots. Similar abundances of geese in burned and unburned wetlands post not food limited in unburned wetlands. Alternatively, geese may have rested in bur ns because the burns are open, providing a clear line of sight to watch for predators. Wet meadows between the Tule and Little Tule rivers in northern California were burned in fall and surveyed for geese use the following spring (McWilliams et al. 2007) Geese made little use of the experimental blocks (2.3 ha) or the peninsula on which the experimental blocks were situated. When geese were present in the experimental blocks, Pacific greater white fronted geese ( Ans er albifrons frontalis ) foraged exclusively in the burned portions during evening feeding periods. The results of
33 Brennan et al. (2005) and McWilliams et al. (2007) indicate geese forage in burned areas due to reduction in vegetation, regardless of timing of the burn. Unburned vegetation likely im pedes access to roots and tubers a preferred food of many geese (Lynch et al. 1947) However, for other waterfowl species, timing of fire to produce beneficial changes in vegetation structure will impact wetland use by these species. During migration, many waterfowl forage on seeds and nutritious regrowth (e.g., van der Graaf et al. 2006) which needs time to grow and develop post burn. Use of Burned Wetlands for Nesting Vegetative structure plays a big role in determining use of burned areas for nesting by many species. Male Louisiana seaside sparrows ( Ammodramus ma ritimus fisheri ) declined in abundance during the breeding season in the first year after a winter burn, but were more numerous in burned than unburned plots the second year (Gabrey and Afton 2000) Oth er sparrow species in this same habitat also used recently burned areas limitedly for nesting in the summer until the second year (Gabrey et al. 2001) Similarly, Cape Sable seaside sparrows ( A. m. mirabilis ) avoided burned areas after a fire burned nesting habitat in the Everglades in southern Florida (Taylor 1983, Curnutt et al 1998, Walters et al. 2000, La Puma et al. 2007) Two years post burn, Cape Sable seaside sparrows were detected in burned areas. Sparrows that were detected in the burned area during the first two years post burn had territories on the edge of the burn (La Puma et al. 2007) It is likely that sparrows may have been exploiting food resources that were available in the burn in the unburned grass edge. Taylor (1983) observed Cape Sable seaside sparrows within one year of prescribed fire in unburned areas of the transect traversing the burn edge, suggesting that the fire caused the birds
34 to clump along this edge. Conversely, seaside sparrows in coastal Maryland had higher densities of territories and nests within one year after a wint er prescribed burn than after longer durations since fire (Kern et al. 2012) Additionally, after more than five years without a burn, nest and territory density was 50% less than densities within the first year post burn. Once vegetative structure returned to pre fire levels, sparrows nested in burned areas. In the Chenier Plains, dead vegetation coverage percentages were similar between burned and unburned plots by the second year and Louisiana seaside sparrows were using recently burned habitat (Gabrey and Afton 2000) In the second year, average nesting activity indicators (i. e., an index of sparrow productivity including adults with nesting materials or food, copulation, nests, flightless juveniles ) per plot were higher in burned than unburned plots, suggesting that the renewal of vegetation benefited this species once vegetation structure recovered. By the third year in the Everglades, re growth provided appropriate vegetation structure in the burned are a and Cape Sable seaside sparrows were found in densities similar to adjacent unburned areas, formed territories, and nested in the burned area (La Puma et al. 2007) For most species with large populations, a distu rbance event impacting 1 2 breeding seasons in a small area is not considered much of a concern. However, for the endangered Cape Sable seaside sparrows fire occurring frequently or over a large area is viewed as a threat to the longevity of this species. Knowledge of sparrow movements as a fire burns Cape Sable seaside sparrow breeding habitat would resolve the question of whether fire creates a displacement or mortality event for this species.
35 Timing of burns can be detrimental to nesting species if nest s are destroyed or fledglings are unable to flee from the fire. Burning too late in the spring can harm nesting ducks (Cartwright 1942) A spring wildfire in the Delta Marsh in Manitoba caused nest failure and mortality of hens and ducklings (Ward 1968) Thus, Ward (1968) suggested burning in summer just after completion of nesting. Summer burns may damage nests of late nesting mottled ducks ( Anas fulvigula ; Hoffpauir 1961) but can be very beneficial if conducted late enough. Ward (1968) conducted a late summer prescribed burn to open Phragmites stands in the Delta Marsh. In multiple subsequent years, ducks heavily used the area for nesting. Winter burns occurring five months prior to the nesting season do not appear to res ult in higher depredation rates (Gabrey et al. 2002) Consumption by fire of nesting substrate may impact nesting locations of many species. Fires occurring prior to the nesting season remove readily available nesting substrate for wad ing birds (Family Ciconiiformes; Giles and Marshall 1954, Bray 1984) At the Bear River Migrato ry Bird Refuge in Utah, Bray (1984) compared densities of vegetation during the summer 4 and 16 months post burn to a wading bird colony under the same water control management scheme. Fire eliminated all dead stems of Scirpus acutus for at least the first four months. Great blue herons, snowy egrets and black crowned night herons ( Nycticorax nycticorax ) nested only in stands of Scirpus acutus with a combination of dead and live stems. While live stem density 4 and 16 months post burn was similar to stem density in colonies, dead stem density 16 mont hs post burn was still less than in colonial nesting areas. The authors suggest that nesting material and nest site availability might be limited. Wading birds did not nest at the
36 burned site (pre or post burn) so conclusions regarding live vs. dead stem density cannot be drawn. American coot ( Fulica americana ) nest success was reduced post burn because fire burned vegetation used for nest material and water levels rose (Austin and Buhl 2011) Nest success declined post burn in the year following the burn, approaching a similar rate of nest success that resulted from grazing. While nest success in the burn treatment was al ready declining pre burn, fire did not benefit the coots in any way. Nest site selection can impact the likelihood of nests being damaged or eliminated by fire. Red crowned cranes ( Grus japonensis ) show a preference for nesting in areas with tall reeds (Wu and Zou 2011) While cranes nested in more diverse habitats after a wildfire, they typically avoided burned areas because of the change in vegetati on, with many birds nesting more than 1.5 km from burned areas (Zou et al. 2003) Cranes occasionally nested in burns, but Wu and Zou (2011) suggest cranes also avoid burned areas because these white birds are more visible in a blackened habitat. No mention was made of how s oon cranes return to these burned areas. Rather than nesting in Scirpus stands, wading birds in the Everglades nest in colonies in tree islands covered with woody vegetation composed primarily of willow ( Salix caroliniana ), buttonbush ( Cephalanthus occiden talis ), and/or cypress ( Taxodium spp.). Two separate wildfire events burned around a large wading bird colony during breeding season, corresponding with dry down near the end of the dry season (Epanchin et al. 2002) Both fires burned the sawgrass ( Cladium jamaicense ) and cattail ( Typha spp.) surrounding the colony, but did not burn the buttonbush and willows on the island. The shallow wate r present within the colony likely protected it from burning. Epanchin et al.
37 (2002) reported no abandonment of nests or loss of chicks after either of the wildfires that burned around the colony. Fire occurring in nesting areas may be detrimental to nest success for species requiring vegetation structure to conceal their nests from predators and reduce exposure of the eggs to the environment. In coastal Louisiana marshes, Gabrey et al. (2002) studied depredation rates on artificial seaside sparrow nests containing Coturnix quail eggs and artificial mottled duck nests in burned and unburne d plots five months after winter prescribed burns. Depredation rates of artificial sparrow nests were high, but not significantly different between burned and unburned plots and not different pre vs. post burn (Gabrey et al. 2002) De predation rates of duck nests did not differ between burn treatments. Vegetation structure did not differ between burn treatments five months post burn, indicating that if burns are conducted so vegetation has time to recover sufficiently, nesting waterfow l likely will not be negatively impacted by predators taking advantage of recent burns. Almario et al. (2009) compared depredation rates of ar tificial and natural seaside sparrow nests in burned and unburned areas in tidal salt marshes in Blackwater National Wildlife Refuge, MD after winter prescribed burns of annually burned areas. In the first year, depredation rates were higher for natural an d artificial nests in the incubation stage in burned than unburned areas, and artificial nests were depredated more than natural nests. Depredation rates of artificial nests in burned areas were similar between studies (Gabrey et al. 2002, Almario et al. 2009) In the following year, more precipitation resulted in greater biological productivity, more standing dead vegetative cover, and a difference in nest placement (Almario et al. 2009) These differences are likely why depredation rates did not differ between burn
38 treatments the second year. Kern et al. (2012) report ed that nest survival was highest in years with high Spartina cover. However, fledging density was similar across all years post burn, indicating that more predation may have impacted recently burned areas that otherwise had higher densities of territories nests, and eggs. Vergeichik and Kozulin (2006) speculate that removal of dead vegetation and fresh green vegetation which normally camouflages nests of the aquatic warbler ( Acrocephalus paludicola ) resulted in increased egg mortality due to predators, especially the abundant shrews in these Polessye lowland mires, Belarus. Besides vegetation structure, food resources and water quality conditions after a burn can influence use of wetlands for breeding purposes. Haszard and Clark (2007) conducted surveys and sampled peatland bogs and fens to determine how a wildfire influenced white winged scoter ( Melanitta fusca ) breeding and brood success. Three years post fire, Haszard and Clark (2007) aerially surveyed for s coter pairs and broods in peatland bogs and fens embedded in burned and unburned upland areas in and adjacent to the Mackenzie River Delta, Northwest Territories, Canada. About 2 weeks after brood surveys, Haszard and Clark (2007) collected water samples from a subset of the surveyed wetlands for analysis for nutrient and dissolved oxy gen concentration and water color and measured conductivity and amphipod abundance in each sampled wetland. While scoter pair density was correlated with higher amphipod abundance, no correlation of scoter pair density or brood occurrence with water chemis try or burn status of the upland was reported. Amphipod abundance was also not related to burned or unburned forest surrounding the wetland. Phosphorus limitation in this area may account for the lack of a response three years post fire. Available nutrient s likely were
39 utilized immediately post fire and thus were distributed and unavailable three years later to increase productivity. Herpetofauna Herpetofauna are impacted by fires occurring in terrestrial and wetland ecosystems (R ussell et al. 1999, Pilliod et al. 2003) While many studies have illustrated the effect of fire in upland habitats, studies of fire effects on amphibians in wetlands are sparse and on reptiles are essentially non existent. Herpetofauna use wetlands for b reeding, development, and refuge. I found 2 reptile and 11 amphibian studies that describe fire effects on herps, including 3 articles that are responses to the initial article. Most studies of fire effects on amphibians go beyond presence and abundance po st fire to examine habitat selection by these species. Indirect Effects of Prescribed Burning Plow lines have been built around wetlands embedded in pine flatwoods to avoid peat fires and negatively impacting the wetland (Russell et al. 1999, Bishop and Haas 2005) When water levels drop, plow lines that were previously submerged within the wetland can become a trap for developing larvae. Bishop and Haas (2005) found >500 desiccated tadpoles in a previously submerged plow line ringing a wetland. Additionally, t hese plow lines may provide a false signal for terrestrially ovipositing flatwoods salamanders seeking a d epression close to the wetland (Russell et al. 1999) Presence and Abundance Herpetofauna use wetlands for foraging, cover, and hibernation, but there is almost no information on effects of fire on species using wetlands. Babbitt and Babbitt (1951) found nearly three dozen injured or dead Florida box turtles ( Terrapene carolina bauri ) and 10 eastern diamondback rattlesnakes ( Crotalus adamanteus ) in Dade
40 County, Florida on a limestone ridge in an area with peat and thick understory vegetation that had burned. It is unclear from the note if the burn occurred in or adjacent to a wetland, but given the proximity of the Everglades, it is likely that this area was at least a short hydroperiod wetland. In response to low intensity prescribed burns in bottomland hardwood forests in Georgia, reptile species richness did not differ between burned and unburned stands (Moseley et al. 2003) However, reptile abundance and diversity was greater in burned than unburned areas, likely because reptiles had more thermoregulatory options as a result of decreased ground cover. It is apparent that more research o f reptile response to fires in wetlands is needed. Amphibian richness, abundance, or diversity within the first year post burn typically does not increase p ositively. Moseley et al. (2003) sampled the amphibian community in bottomland hardwood stands in Georgia 6 10 months after low intensity winter prescribed burns. Amphibian richness, abundance, and diversity did not differ between burned and unburned treatments, likely becaus e volumes of coarse woody debris providing cover remained post fire. Occupancy of isolated wetlands in Montana by long toed salamanders ( A mbystoma macrodactylum ) and Columbia spotted frogs ( Rana luteiventris ) did not change after summer wildfires burned th eir wetlands (Hossack and Corn 2007) Salamanders may have increased occupancy of b urned wetlands post burn, but support was weak for these models. Boreal toads ( Bufo boreas ) colonized wetlands post burn, but were not breeding in these wetlands before the fire. Conversely, Schurbon and Fauth (2003) indicate that fire has immediate and short term negative impacts on amphibian abundance and diversity in ponds embedded in southeastern pine flatwoods in Francis Marion National Forest, South Carolina. Many species of
41 amphibians were not detecte d the first year after a fire, however, the interpretation of these results is limited given that this study was conducted for only one season post burn Using historic fire data, Schurbon and Fauth (2003, 2004) showed th at richness increased with time since the wetland had burned, resulting in the authors recommending that fire frequency should be decreased and burn season switched from winter to summer. This interpretation has been questioned on the grounds of short stud y period, fire history of the study sites, and hydroperiod lengths (Means et al. 2004, Rober tson and Ostertag 2004) Breeding Amphibian species have a variety of requirements when selecting breeding sites that impact whether burning wetlands may be beneficial or not. Boreal toads did not breed in isolated, montane wetlands before they were burne d, but used burned wetlands for breeding the year after the burn (Hossack and Corn 2 007) After the first year post burn, boreal toad abundance declined in these wetlands over the next two years to zero. No boreal toad larvae were found in unburned wetlands, suggesting that boreal toads benefit from fire burning wetlands. In a follow up study, Hossack and Corn (2008) investigated how water temperature post burn influenced breeding site selection by boreal toads in years two and three after a wildfire in Glacier National Park. However, unburned wetlands (which were unoccupi ed by toads due to selection of sampling sites) were warmer than burned wetlands despite open canopy over all sites. T oads did not appear to select against warmer wetlands. Hossack and Corn (2008) note that they have never found boreal toad s breeding in well shaded wetlands, suggesting that sunlight or an open canopy is more important for this species than temperature. Boreal toads appear to prefer open canopy as long as they also have ground cover to
42 escape predators and maintain water bala nce (Guscio et al. 2008) Like boreal toads, Florida bog frogs ( Lithobates okaloosae ) require periodic fires in their breeding wetlands to ma intain vegetation conditions conducive to calling (Gorman and Haas 2011) Calling sites for this species had more submergent vegetation and lower water depths than what the sympatric bronze frog ( L. clamitans clamitans ) selected, and may be a result of o ccasional fire in the wetland (Enge 2005, Gorman and Haas 2011) Interestingly, the bronze frog selects for sites with appropriate water depth and flow conducive to oviposition and egg development, quite different from the habitat selected by Florida bog frogs. Periodic fire may be important for maintaining habitat necessary for larva l development of amphibians. Periodic fires burn bogs embedded in longleaf pine forests in the southeastern US, removing woody vegetation and supposedly maintaining hydroperiod length by reducing evapotranspirative losses (Means and Moler 1979) These fires reset succession of shrub bogs to herbaceous bogs, thereby benefiting larval Pine Barrens treefrogs ( Hyla andersonii ) typ ically found in herbaceous bogs. While adult treefrogs use bogs with woody vegetation, the additional woody vegetation is not good larval habitat. Similarly, larval flatwoods salamanders ( Ambystoma cingulatum ) may benefit from fire in wetlands by reducing woody vegetation in the canopy (Russell et al. 1999, Pilliod et al. 2003, Bishop and Haas 2005) Bishop and Haas (2005) provide indirect evidence tha t burning wetlands during the summer to open the canopy may be beneficial to developing larval flatwoods salamanders, potentially due to some combination of warmer water, a change in predation risk, more food, and
43 higher dissolved oxygen concentrations und er open canopies. However, there is still no direct evidence that flatwoods salamanders benefit from fire burning their wetlands. Lessons Learned Fire is an important process in wetlands that resets habitat conditions for wildlife either via elimination of forbs and woody vegetation (short term changes) or by burning peat and creating deep, open water habitats (long term changes). Direct mortality events in wetlands appear to be rare, just like in upland habitats (Whelan 1995) Indirect effects of fire relate to removal of vegetation, change in food resources, and habitat changes that affect breeding. Specifically, fire initially reduces abundances of species that prefer dense vegetation for cover and nes ting. Conversely, fire increases abundances of species that utilize open habitats for breeding and foraging. However, while we can predict responses of many species according to their life histories and specific requirements for cover, food, and habitat fo r breeding and development, management is best conducted using data rather than inferences. Appropriate vegetative structure is important for habitat occupancy by wildlife (e.g., Gabrey et al. 2001, Gabrey and Afton 2004) and impacts nest depredation rates (Gabrey et al. 2002, Almario et al. 2009) Vegetative cover serves multiple purposes including nesting materials, supports, and cove r, protection from predators and the elements, and food resources (i.e., seeds, fresh growth, and habitat for insects and small mammals). Fire temporarily alters these conditions until vegetative cover regrows. Species such as seaside sparrows and wrens th at typically are found in habitats with dense vegetative structure are not found in a recently burned area until vegetative structure returns to pre burn level. Evidence of preference for vegetative cover was seen in seaside sparrow use of burn edges and o f patches of vegetation within burn
44 units (Taylor 1983, Gabrey et al. 1999, La Puma et al. 2007) and black rail use of unburned patches of marsh (Legare et al. 1998) Co nversely, species that prefer open areas with limited vegetative structure for at least part of their daily activities such as waterfowl and boreal toads used burned areas extensively until vegetative structure becomes too dense. However, even for many spe cies that prefer dense vegetation, vegetation can become too dense and impede movements. For example, Conway et al. (2010) noted that rails walk on top of thick mats of dead vegetation where they are more exposed to predators than walking through recently di sturbed habitat. Food resources such as fresh regrowth, mast, or arthropods may explain use or disuse immediately post burn, yet the response of wildlife to burns in wetlands to changes in food resources has not been well studied. While it is apparent from these studies that geese and birds of prey such as harriers respond to fire due to food resources, it is likely that many other species (e.g., icterids, sparrows) use burns for food resources. Combustion of grass and forbs immediately exposes roots to for aging by many herbivorous species such as geese and continues as fresh shoots emerge, attracting many other herbivores such as muskrats and deer. Fresh forage is often more digestible and of higher nutritional quality (Smith et al. 1984) providing a valuable source of nutrition for many species. Mast production may take much lo nger to produce (Lyon et al. 2000) although seed s that have already dispersed may be immediately available once the overstory is burned. Invertebrate response to fire varies in magnitude and in timing (e.g., de Szalay and Resh 1997, Benson et al. 2007, Hochkirch a nd Adorf 2007, Munro et al. 2009) but increased abundance of many invertebrate species
45 bogs responded strongly to frequency of grass seed stalks and forb density as p redictors of bog occupancy (Tucker and Robinson 2003) The wint er diet of this species is not well known so seeds or arthropod density (using forb density as an indicator of arthropod density) may be driving bog occupancy. Specific studies should be undertaken to describe wildlife response to changing food resources d epending on season of burn, wetland conditions, and frequency and severity of fires. When fire in wetlands removes vegetation, breeding efforts of amphibians typically benefit while birds may or may not benefit. Fire reduced canopy cover over many wetlands benefiting larvae of many amphibian species. Canopy cover impacts temperature which is important to developing larvae (Niehaus et al. 2006) typically increasing the rate of metamorphosis until temperatures stress larvae. For birds, environmental effects do not appear to impact nests like they do for amphibians, However, depredation of nests post burn may result. Recent burns do not appear to impact predation rates of bird nests (Almario et al. 2009) but differences in predators may affect nest success (Vergeichik and Kozulin 2006, Kern et al. 2012) Vergeichik & Kozulin (2006) note that shrews forage by smell, sound, and touch, making them more efficient at finding nestlings rather than eggs. Similar predators may have reduced fledgling densities to be equal between burned and unburned plots despite higher egg densities in burned areas in their study (Kern et al. 2012) To my knowledge there are no studies describing the effects of fire on nest predators such as small mammals or mesopredators. In upland habitats, patches and mo saics are increasingly recognized as important aspects to include in management of ecosystems. Many wetland dependent species
46 benefit from burns that leave unburned patches of habitat as refugia (e.g., Legare et al. 1998, Gabrey et al. 1999, 2001) Patchy burns can provide these refugia while also controlling vegetation density and cover. Waterfowl need a combination of shallow and deep open water areas for feeding, rearing broods, and avoiding predators, but also se ek dense vegetation during nesting as a means of avoiding predators. Ward (1968) recommended burning sections of the marsh in order to leave sufficient nesting habitat available for ducks. Diving ducks primarily use open water areas in marshes while geese, using wetlands duri ng migration or for winter, seek food resources such as succulent new regrowth. For species that prefer recently burned areas, patchy burns meet these habitat requirements and include cover for avoiding predators, particularly if woody debris for herps is left after a fire. At the same time, patchy burns also meet the needs of other species when unburned patches are left standing. By creating refugia, fewer birds are displaced by the burn. There has been no work in wetlands to quantify the size, number, or distribution of refugia post burn to benefit species. Purposely creating refuges within a burn will likely require extensive effort by managers. To overcome this, research is needed to determine burn conditions conducive to naturally creating refuges via t he burn or extensive effort during fire operations. The effects of wildfires and prescribed burns are often different due to season, severity, and other conditions. Most of the studies I reviewed were conducted after prescribed burns outside of the natural fire season. Thus, our understanding of fire effects on wildlife is relevant to our current management scheme, but does not reflect historical ecosystem effects. Historical fire effects likely provided crucial habitat for species, whether through canopy r emoval benefiting larval amphibians (Bishop and
47 Haas 2005) or open water habitat for waterfowl (Ward 1968) Peat fires, which sometimes occur with wildfires, can significantly change a wetland and may be important in managing certain species (Ward 1968, Norton and De Lange 2003) However, peat burns are avoided during prescribed burns because they are impossible to control. Managing wetlands to allow natural peat burns to occur is necessary to maintain many wetland communities (Reardon et al. 2007) Pre burn data are typically unavailable for wildfires and control sites may not be comparable to burned sites despite being adjacent to the burned area, making interpretation of fire effects challengi ng. However, more studies on wildfires are needed to address questions of effects on wildlife post burn. Season of burn was frequently discussed in the early fire literature when burns were commonly conducted to benefit waterfowl. Effects of fire due to f requency, severity, and time since burn are equally important. A recent review of fire dependent upland avian species in longleaf pine forests highlighted the differences in responses to fire as a result of season of burn (Cox and Widener 2008) Historically, prescribed burns in wetlands were conducted sometime from early fall to early spring to avoid biological (e.g., burning nests) and environmental (e.g., burning peat) impacts. However, most natural fires occur during late spring and summer when lightning is most prevalent (e.g., Gunderson and Snyder 1994, Slocum et al. 2007) At this time of year, wetlands often contain much water, reducing the potential for peat burns. If prescribed burns are conducted in mid spring or summer, nesting, breeding, and requirements of eggs or juveniles inhabiting the wetland must be considered in order to avoid eliminating a sensitive cohort while achieving the goal of benefiting target spe cies. Continuance of
48 prescribed burning in other seasons also must consider species movements and resource and habitat needs. Bishop and Haas (2005) suggested that winter prescribed fires in upland areas around wetlands may negatively impact the migration of flatwoods salamanders to wetlands at this time of year f or initiation of breeding. Most studies I reviewed were conducted within 2 years post burn although a few exceptions carried studie s 4 6 years (Taylor 1983, La Puma et al. 2007, Conway et al. 2010) While vegetation in some wetlands recovers rapidly (e.g., 1 2 years; Loveless 1959, Gabrey et al. 1999) l ong term studies are necessary to determine the trends species show in relation to vegetation recovery (e.g., Gabrey et al. 2001, La Puma et al. 2007, Conway et al. 2010) The fire return interval is typically longe r than complete vegetation recovery. By monitoring species from pre burn through vegetation recovery post burn until at least the next burn, fire frequency can be adjusted if the current fire regime negatively impacts species. Studies presenting 1 year of data are informative, however they may result in misleading conclusions and recommendations and should qualify results as preliminary until more research over a longer time period is conducted. In order to understand the effects of fire on wetland dependen t wildlife, studies need to include controls whether side by side comparisons (i.e., burn vs. control) or temporal (i.e., pre vs. post burn) comparisons. Early studies advocating use of fire in marshes for the enhancement of waterfowl foraging and nesting areas were typically incomplete in their description of burn methods, description of environmental and confounding variables, quantification of response, and management techniques. However, some, like Ward (1968) made an effort to provide their methods and means of comparis on by describing their ignition conditions and process so others could
49 replicate the burn prescription. Yet pre burn data and replication were not included in the report of the study results. Before after control impact (BACI) designs are well suited to s tudies of fire effects because they incorporate spatial and temporal comparisons (Underwood 1994) Conway et al. (2010) used this type of design to determine trends of rail abundances in plots prior to and after conducting prescribed fires. Furthermore, the importance of designing studies th at incorporate knowledge of fire history and regime, life history requirements, and appropriately quantified response to fire by species (Means et al. 2004, Robertson and Ostertag 2004, Schurbon and Fauth 2004) cann ot be overemphasized, as this is critical when providing recommendations to managers. While long term studies such as the study by Conway et al. (2010) may be difficult to conduct when time and money are limited, an understanding of bird response as vegetati ve cover and insect populations recover post fire is necessary to appropriately implement fire as a management tool in wetlands. Many of the studies of Cape Sable seaside sparrows were terminated shortly after sparrows in burned and unburned areas became s imilar in abundance. Whether the trend in abundance continued to increase or peaked was not determined, but is important for implementing or adjusting a prescribed fire regime, particularly if target species are a species of concern. Basic presence and abu ndance information is important to initially focus research on important questions, but more extensive, complex studies are needed. Most studies I reviewed report solely on presence/absence, abundance, and duration of use of species post burn. Reptiles and mammals were essentially unrepresented in the wetland fire effects literature. Only a handful of studies looked at amphibian or avian breeding or
50 nest site selection, larval development, or depredation rates of nests. A series of experiments by Gabrey and coauthors (Gabrey et al. 1999, Gabrey and Afton 2000, 2001, Gabrey et al. 2001, Gabrey et al. 2002, Gabrey and Afton 2004) on Louisiana seaside sparrow response to winter prescribed fires in the Chenier Plains prov ide a great illustration of the type of studies needed to understand effects of fire on species. (Hossack and Corn 2007, Guscio et al. 2008, Hossack and Corn 2008, Hossack et al. 2009) few follow up studies of species responses to fire in wetlands have been conducted once presence and abundance data were collected. Saab and Powell (2005) called for a move towards increasing our understanding of fire effects on reproductive success, nest survival, and changes in population. I echo this call for fire effects research on all wetland dependent wildlife. While we now know that certain groups of species are attracted to or avoid burns in wetlands one to two years post burn, we still cannot confidently point to the causal mechanism for many of these s pecies, whether mortality, vegetative cover requirements, food resources, or other factors. A number of overarching questions exist regarding fire effects on wetland dependent wildlife. I highlight some of them here. How do mammals respond to fire in wetla nds? A number of small mammal species inhabit wetlands, but fire effects on most of these species have not been studied. Given changes in vegetation density and structure, I would expect many of these species to respond similarly to wrens and sparrows and avoid recently burned areas until vegetation recovers. However, some species (e.g., Florida salt marsh vole ( Microtus pennsylvanicus dukecampbelli ) are adapted to daily disturbances (i.e., tides) and may
51 have a different strategy. Other mammals should bene fit from fresh vegetative growth with higher nutrient content. I found almost no studies of non domesticated herbivorous mammals in wetlands. This could be a factor of the type of wetlands, primarily salt marsh, that research of fire effects on wildlife ha s been conducted in. An understanding of the mammal community is important for the sake of managing for mammals, to understand the impact of mammal populations on bird nest predation rates, and as a food source for species. How do reptiles respond to fire in wetlands? Reptiles use wetlands for a variety of purposes, like other species. However, I only found one study specifically investigating fire effects on reptiles in wetlands, which indicated snakes may benefit due to increased thermoregulatory opportun ities. Additionally, turtle mortality near a wetland was reported (Babbitt and Babbitt 1951) indicative that some species may be negatively impacted by fire if refugia are not available. Deep peat burns may be important for turtles to maintain deep open water areas, however use of fire s to maintain conditions in wetlands appropriate for use by turtles has not been studied. How does fire control of food resources affect species response to fire? While much of the emphasis throughout this review was placed on vegetative density and struct ure, it is not entirely clear which environmental factors dictate species response to fire. Vegetative cover provides protection from predators, serves as breeding habitat, and is a source of food resources for many species, whether directly or indirectly. Fire temporarily alters these conditions promoting regrowth and a change in habitat structure and composition. Invertebrate response to changes in vegetation, microclimate, and debris, impacts their availability to species preying on them. An understandin g of these
52 mechanisms will help inform management decisions regarding burn season, frequency, and severity. How do peat fires impact wildlife use of wetlands post burn? Peat fires typically occur during drought conditions in the natural fire season, but ar e actively avoided by limiting implementation of prescribed burns to times when water levels protect peat. However, peat burns maintain wetland communities by resetting succession and creating open water habitats favorable to some species. Suppression of p eat burns may negatively impact wetlands and the species utilizing them. Additionally, peat fires are important in maintaining wetland characteristics and vegetative communities (i.e., Atlantic white cedar) and suppression of peat fires alters these wetlan ds and affects the wildlife that use these fire maintained habitats. How do fires impact individual movements and long term wildlife population trends? Individual animal movements and population trends related to fire in wetlands remains unknown. Much disc ussion regarding mortality vs. dispersal of species due to fire and the impacts on the population has occupied reviews of species such as the Cable Sable seaside sparrow. Yet, to my knowledge no telemetry of individuals has been conducted to confirm or ref ute direct negative fire impacts that would also negatively impact a small population. Similar studies should be done on other species to better understand how species respond to fire. Studies of responses to fire by many species need to be conducted so we understand how prescribed burns and wildfires impact target and non target species.
53 Table 2 1. Selected references of fire effects on wetland dependent avian species. Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Respons e to Burns Comments Sparrows Cape Sable seaside sparrow ( Ammodramus maritimus mirabilis ) Everglades; sawgrass ( Cladium jamaicense ) 4 yrs post fire W late dry season (May) breeding primarily avoided for first 2 yrs; 3 yrs post fire densities & te rritories similar to unburned; nesting start 3 yrs post fire appropriate veg structure returned 3 yrs post burn; suggested need for refugia & more time between burns (i.e.,10+ yr) a Cape Sable seaside sparrow Everglades; Muhlenbergia & sawgrass ( Cladium j amaicense ) 4 yrs post fire W winter Rx & June lightning fire breeding deeper soil: return in 2 nd yr & maybe peak in yr 4; shallow soil: returning in yr 4 b Dusky Seaside Sparrow St. Johns NWR, FL; salt marsh ~1 yr W winter breeding returned to burned are a 6 mo. post burn to set up & defend territories 3 birds banded in burn were found in unburned habitat 900 m from banding location c Louisiana seaside sparrow Chenier Plains, LA; brackish & salt marsh 3 breeding seasons Exp winter (mid Jan.) breeding 1 st yr: male abundance increased in season; 2 nd yr: more males in burn; nesting lower in burn in 1 st yr, but 2 nd yr higher in burn dead veg cover recovered in 2nd yr likely why nesting so much better second yr d
54 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments seaside sparrow Chenier Plains, LA; brackish & salt marsh 2 winters Exp winter (Dec., early Jan.) cover not found in burns until 2 nd yr e seaside sparrow Chenier Plai ns, LA; brackish & salt marsh 3 breeding, 1 pre & 2 post Exp Winter (Dec., early Jan.) nesting abundance dropped in burned plots & then increased 2 nd yr post burn positively correlated with dead veg & S. patens f seaside sparrow Chenier Plains, LA; bracki sh & salt marsh 2 breeding, 1 pre & 1 post burn Exp winter (mid Jan) nesting (artificial nests) high depredation, but no diff between yrs or trmts veg cover 5 mo. post burn similar to pre burn so likely reason for no difference g seaside sparrow Blackwate r NWR, MD; tidal marsh 2 breeding seasons Rx winter nesting nest depredation high during incubation, total depredation did not differ between trmts, next yr showed no differences artificial nests were depredated at much higher rate in burn than unburned h
55 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments seaside sparrow Blackwater NWR, MD; tidal marsh 5+ yrs Rx winter nesting <1 yr post burn, highest territory and nest density; 50% lower nest and territory density 5+ yr than <1 yr post burn; egg density higher <1 yr than 3 4 yr post burn; no fledging dens ity difference percent Spartina cover and year explained nest success; predation may have caused depression of fledging density in recent burns ag Nelson's sharp tailed sparrow Chenier Plains, LA; brackish & salt marsh 2 winters Exp winter (Dec., early Ja n.) cover found in patches of unburned veg in one burn station e sparrows Chenier Plains, LA; brackish & salt marsh 3 breeding seasons Exp winter (Dec., early Jan.) foraging nesting 2 nd yr post burn 2x more than 1 st yr, but no diff with 3 rd yr i swamp sparrow Chenier Plains, LA; brackish & salt marsh 2 winters Exp winter (Dec., early Jan.) cover found only in stations with bunch of unburned veg e sparrow AL & FL; Gulf Coast pitcher plant bogs 2 winters NA growing, dormant wintering higher ab undance 1 st yr post fire; densities post growing season higher thru more yrs y
56 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments sparrows (transients) Tall Timbers Research Station Gannet Pond 4 mos. Exp winter song & swamp sparrow had more in unburned shoreline j grassland yellow finch Pampas, Arg entina; salt marsh ~1 yr NA spring NA only in unburned Spartina ; 1 mo. Post Juncus burn k great pampa finch Pampas, Argentina; salt marsh ~1 yr NA spring NA only in burned Spartina ; 2 mo. Post Juncus burn k Wrens marsh wren Chenier P lains, LA; brackish & salt marsh 2 winters Exp winter (Dec., early Jan.) cover found more in unburned immediately post fire although this increased in 2 nd yr other birds detected on <5% of surveys e sedge wren ( Cistothorus platensis ) Chenier Plains, LA; b rackish & salt marsh 2 winters Exp winter (Dec., early Jan.) cover not found in burns until 2 nd yr, but still primarily in unburned e sedge wren Northeast MN; scrub/shrub 1 yr, but 0 3+ yr fires Rx NA breeding highest abundance on burned sites time scale very coarse in this study l grass wren ( Cistothorus platensis ) Pampas, Argentina; salt marsh ~1 yr NA spring NA only in unburned Spartina ; 4 mo. Post Juncus burn, 6 mo. Similar abundance k
57 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments wrens Chenier Plains, LA; brackish & salt marsh 3 breeding seas ons Exp winter (Dec., early Jan.) NA no diff i Wetland Associated spp. common yellowthroat Chenier Plains, LA; brackish & salt marsh 2 winters Exp winter (Dec., early Jan.) cover not found in burns until 2 nd yr unburned patches provide c over for small birds e red capped wren spinetail Pampas, Argentina; salt marsh ~1 yr NA spring NA appeared 4 mo. Post burn, but lower abundance than unburned Spartina ; 3 mo. Post Juncus burned; similar abundance btwn habitats 1 yr post burn Juncus recover ed structure within 1 yr, but not Spartina k emergent wetland spp. Northeast MN; scrub/shrub 1 yr, but 0 3+yr fires Rx NA breeding more abundant on managed sites, includes sheared sites l shrub/forest spp. Northeast MN; scrub/shrub 1 yr, but 0 3+yr fire s Rx NA breeding more abundant on unmanaged sites l
58 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments southern lapwing Pampas, Argentina; salt marsh ~1 yr NA spring NA in burn only until Spartina sprouted; first in Juncus but brief k songbirds, but some others Lake Victoria, Uganda; papyrus swamps 1 yr NA NA foraging other generalist spp. use burns more, but papyrus reliant spp. not present z aquatic warbler ( Acrocephalus paludicola ) Belarussian Polessye; fen marshland 1 yr obs. NA spring nesting suggest that lack of dead veg and green grass that eg g mortality increased due to predation fire occurred during one year of study and was not part of study design af red crowned crane ( Grus japonensis ) Zhalong Nature Reserve, China; reed swamp NA W fall, spring nesting foraging avoid blackened burn, were f arther from burned area with dense reeds nearby for concealment aa red crowned crane Zhalong Nature Reserve, China; reed swamp NA H fall, spring nesting prefer tall reeds, may nest in burned areas ad songbirds, dove Tall Timbers Research Station Gannet Pond 4 mos. Exp winter all had more on burned shoreline j
59 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments "residents", crow, cardinal Tall Timbers Research Station Gannet Pond 4 mos. Exp winter all had more on burned shoreline j correndera pipit Pampas, Argentina; salt marsh ~1 yr NA spring NA in burn only, first in Juncus & then persisted k Hudsons canastero Pampas, Argentina; salt marsh ~1 yr NA spring NA seen first months post burn & then absent k freckle breasted thornbird Pampas, Argentina; salt marsh ~1 yr NA spring NA appear 3 mo. post Juncus burn; similar abun btwn habitats 1 yr post burn k gulls, swallows Chenier Plains, LA; salt marsh obs. Exp Fall foraging catching insects in smoke of fire m marsh harrier Watarse Marsh, Japan; reed marsh 2 winters W winter wint ering breeding flew less over burned area 1 st yr post burn; same use 2 nd yr post burn as unburned marshes reed beds regrew by 2 nd yr; suggest that mid March Rx of reeds inhibits breeding ae Icterids boat tailed grackle Chenier Plains, L A; brackish & salt marsh 2 winters Exp winter (Dec., early Jan.) foraging found immediately post burn, but not following yr e
60 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments red winged blackbird Chenier Plains, LA; brackish & salt marsh 2 winters Exp winter (Dec., early Jan.) foraging cover found 2x m ore in burn following fire, 2 nd yr still lots of birds in burn e icterids Chenier Plains, LA; brackish & salt marsh 3 breeding seasons Exp winter (Dec., early Jan.) nesting NS, but 1.5 yr (2 nd yr) post burn, more than 1 st or 3 rd yr l red winged blackbi rd Chenier Plains, LA; brackish & salt marsh 3 breeding, 1 pre, 2 post Exp winter (Dec., early Jan.) nesting abundance increased in burned plots 1 st yr & then decrease 2 nd yr toward pre burn negatively correlated with % cover of dead veg & S. patens f boa t tailed grackle Chenier Plains, LA; brackish & salt marsh 3 breeding, 1 pre, 2 post Exp winter (Dec., early Jan.) nesting abundance increased in burned plots 1 st yr & then decrease 2 nd yr toward pre burn negatively correlated with % cover of dead veg & S patens f yellow winged blackbird Pampas, Argentina; salt marsh ~1 yr NA spring NA only in unburned Spartina similar abundance at end of study in Juncus k icterids (residents) Tall Timbers Research Station Gannet Pond 4 mos. Exp winter NA more on bu rned shoreline except Red wing blackbird j
61 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments blackbirds Chenier Plains, LA; salt marsh obs. Exp Fall foraging catching insects in smoke of fire m Marsh Birds "transients": snipe Tall Timbers Research Station Gannet Pond 4 mos. Exp wi nter all had more on burned shoreline j Wilson's snipe Rainwater Basin, NE; isolated wetland 2 wks pre post of burn, 3 yrs of burns Rx spring increased frequency & abundance in burn burned adjacent to area surveyed n jacksnipes & shorebirds Chenier Pl ains, LA; salt marsh obs. Exp Fall NA seen in burn m black rail ( Laterallus jamaicensis ) St. Johns NWR, FL; ( Spartina bakeri ) marsh obs. Rx NA use unburned patches for refuge mortality occurred in patches that subsequently burned o Yuma clapper rail Co lorado River delta, CA & Mexico; cattail, reed, bulrush 1 6 yrs pre & 2 5 post burn Rx,H late winter early spring breeding foraging more post than pre in burned areas, diminished over time burns conducted shouldn't be large in spatial extent so birds hav e refugia p
62 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments other rails & least bittern Colorado River delta, CA & Mexico; cattail, reed, bulrush 1 6 yrs pre & 2 5 post burn Rx,H late winter early spring breeding foraging no pre post difference p Virginia rails Colorado River delta, CA & Mexico; catt ail, reed, bulrush 1 6 yrs pre & 2 5 post burn Rx,H late winter early spring breeding foraging more post than pre burn p dot winged crake Pampas, Argentina; salt marsh ~1 yr NA spring NA only in unburned Spartina; not in burned Juncus k speckled crake Pampas, Argentina; salt marsh ~1 yr NA spring NA only in unburned Spartina k Waterfowl Geese: snow, Canada, white fronted Chenier Plains, LA; salt marsh obs. Exp Fall foraging cover increased abundance, 1939 burn: ~500k note importanc e of spotty burns no specific details provided m ducks Chenier Plains, LA; salt marsh obs. Exp Fall thousands of ducks m
63 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments geese SE US wildlife refuges freshwater tidal to coastal salt marsh obs. Rx early fall, late winter foraging geese forage on gr een, succulent browse q blue geese (snow geese) LA coastal marshes obs. NA late Sept. to Jan. foraging heavy use s waterfowl Sandhills, NE; isolated wetlands obs., 2 yrs Rx winter used during spring migration & for activities during the su r ducks, other waterfowl Delta Marsh, Manitoba; open sloughs & bays along lake obs. Rx summer: after nesting; after end of July nesting migration burn more heavily used for nesting & autumn gatherings t ducks Delta Marsh, Manitoba; open sl oughs & bays along lake obs. W,H April thru May nesting mortality of females, ducklings, destruction of nests t white geese Chenier Plains, LA; brackish & salt marsh 2 winters Exp winter (Dec., early Jan.) foraging 8 of 10 flocks in burn, other two elsew here e
64 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments snow geese Rainwater Basin, NE; isolated wetlands 2 wks pre post of burn, 3 yrs of burns Rx spring foraging more abundant although in fewer burned wetlands n white winged scoters Mackenzie Delta, Canada; bogs & fens 1 summer (3 yrs post fire) W NA, but likely summer foraging nesting scoter density higher with amphipod abundance amphipods not related to fire; no nutrient effects 3 yrs post fire u cackling & Pacific greater white fronted geese Fall River Valley, CA; wet meadow, could be riparian a rea ~2.5 mo. in spring post burn, maybe 1 yr later on some plots Rx winter foraging resting on a couple days WF geese foraged exclusively in burned areas & were more numerous on these days cackling geese did not visit experimental plots v ducks south cent ral Canada; small isolated wetlands 2 breeding Rx spring & fall breeding occupancy not different between wetlands with burnt or unburnt edges ab coots Grays Lake NWR, ID; fields on perimeter of montane wetland 2 yr pre & 2 yr post Rx fall nesting nest su ccess declined, but was already declining pre burn nest success was already declining pre burn ac
65 Table 2 1. Continued Species Wetland Type Length of Study Fire Type Season of Fire Use of Wetland Response to Burns Comments mottled duck Chenier Plains, LA; brackish & salt marsh 2 breeding, 1 pre, 1 post Exp winter (mid Jan.) nesting (artificial nest) high depredation, but no dif ference between yrs or trmts veg cover 5 mo. post burn similar to pre burn, likely reason for no difference g Wading Birds black crowned night heron, snowy egret, great blue heron Bear River Migratory Bird Refuge, UT; impoundment 16 mo. Rx March nesting burn removes nesting substrate & nest materials burned area compared to colonies, colonies not burned w wading birds Everglades sawgrass ( Cladium jamaicense ) obs. W late dry season foraging nesting no effect on colony, mortality of whi te ibis in slough x wading birds (resident) Tall Timbers Research Station Gannet Pond 4 mo. Exp winter all had more on burned shoreline j a La Puma et al. 2007; b Taylor 1983; c Baker 1974; d Gabrey & Afton 2000; e Gabrey et al. 1999; f Gabrey & Afton 200 4; g Gabrey et al. 2002; h Almario et al. 2009; i Gabrey et al. 2001; j Vogl 1973; k Isacch et al. 2004; l Hanowski et al. 1999; m Lynch 1941; n Brennan et al. 2005; o Legare et al. 1998; p Conway et al. 2010; q Givens 1962; r Schlichtemeier 1967; s Perkins 1968; t Ward 1968; u Haszard & Clark 2007; v McWilliams et al. 2007; w Bray 1984; x Epanchin et al. 2002; y Tucker & Robinson 2003; z Maclean et al. 2003; aa Wu & Zou 2011; ab Hochbaum et al. 1985; ac Austin & Buhl 2011; ad Zou et al. 2003; ae Hirano et al. 2003; af Vergeichik & Kozulin 2006; ag Kern et al. 2012 *Type of fire: Exp=experimental, H=human caused, Rx=prescribed burn, W=wildfire NA = not available obs. = observational study
66 Figure 2 1. Number of studies per year of fire effects on each group in wetlands.
67 CHAPTER 3 EFFECTS OF PRESCRIBE D FIRE ON FORAGING B Y WADING BIRDS (CICONIIFORMES) IN T HE EVERGLADES Introduction Disturbance of upland areas via farm machinery, fire, and other physical disturbances often attract birds to forage or scavenge for displaced, injured, or recently killed prey (e.g., Komarek 1969, Smallwood et al. 1982, Tewes 1984, Toland 1987) The removal of vegetation can increase the availability of prey (Vickery et al. 2001) even when abundance or density of prey does not increase (Vickery et al. 2001, Munro et al. 2009) Thu s, i ntake efficiency of birds may be increased because of improved availability of prey, but not due to abundance or diversity of prey (Devereux et al 2006) Although changes in prey availability due to disturbance are known from upland habitats, it is unclear whether similar kinds of disturbance would result in the same effects on aquatic prey animals The absence of appropriate habitat and vegetativ e structure appears to strongly affect species specific responses to fire (Gabrey et al. 1999, Baldwin et al. 2007) Studies of responses of wetland birds to burns tend to record presence/absence of species after a burn, with many species not returning for a year or more (Venne, Chapter 2). For example, savannah sparrows tend to be found in areas with sparse vegetatio n within one year post burn while sedge wrens prefer dense vegetation which has not burned in the previous two years (Baldwin et al. 2007) Fire may also positively affect forag ing conditions for many wetland dependent birds. Recent burns appear to provide enhanced access to belowground plant parts for wintering geese (Gabrey et al. 1999) if burns are conducted during the appropriate time frame to make these resources available (Brennan et al. 2005, McWilliams et al. 2007) Fires also remove
68 dead plant litter and thus release nutrients. This process can increase nutritive value of vegetation (Smith et al. 1984) ac cessibility of food resources (Gabrey et al. 1999) and abundance of food resources such as invertebrates (de Szalay and Resh 1997) Foraging s uccess of long legged wading birds (Ciconiiformes) depends largely on prey availability (Bancroft et al. 2002, Gawlik 2002) Water depth is a primary determinant of prey availability since wading birds are limited to foraging in water no deeper than their leg length (Powell 1987, Ga wlik 2002) Emergent vegetation density also plays an integral role in prey availability in two important ways. Dense vegetation can impede access to prey, but may also increase prey density by improving cover to hide from predators. Thus, sparse vegetati on may be preferred by wading birds compared to no or dense, emergent vegetation (Lantz et al. 2011) and edges may be preferred over open water (Stolen 2006) Fire may affect foraging opportunities for wading birds through several mechanisms, includin g direct mortality of prey, alteration of habitat that prey depend on, increased primary production benefiting primary consumers through the release of nutrients and increased light, or changing accessibility of prey Thus, burned areas should be attractiv e to wading birds for the duration of the effect resulting from fire. I tested the hypothesis that wading birds select for burned habitats over unburned habitat because burned areas in my study area (i.e., sawgrass ridges) are shallower than the surroundin g marsh, vegetation is relatively sparse (making prey more accessible) compared to thick Eleocharis marshes, and prey densities are potentially greater than in similar unburned areas. I also predicted that wading birds in the Everglades would have a higher capture rate (captures per minute) and capture efficiency (captures per
69 attempt) in burned areas than in unburned areas, accounting for differences in water depth, flock size, and time since burn Number of birds within a foraging flock affects foraging s uccess of wading birds (Krebs 1974) I controlled for this variable because it could confound the anal ysis of fire effects on foraging success. Changes in vegetation and other factors associated with time since burn may also affect foraging success of wading birds, so I included time since burn as a covariate to describe the effect this variable has on for aging success of wading birds. I also tested the hypothesis that fires make prey available by injuring or killing prey during the burn, predicting that d ead or injured prey would be more abundant in burned than unburned areas. Finally, I assumed that prima ry production post burn would be elevated and predicted that prey densities would be greater in burned than unburned sawgrass Study Area The Everglades is a large oligotrophic wetland in southern Florida, USA, where primary production is phosphorus limite d (Noe et al. 2001) Sawgrass ( Cladium jamaicense ridges) surrounded by open water sloughs and wet prairies dominated by sedg es ( Eleocharis spp.; Gunderson 1994) Rainfall occurs seasonally, primarily during the wet season from May to October resulting in strong ly fluctuating water levels. Fire is a natural component of this landscape (Wade et al. 1980) occurring most frequently at the onset of the wet season (May June) when lightning is common (Gunderson and Snyder 1994, Slocum et al. 2007) More acreage is burned during the transition from dry to wet season starting in May (when thunderstorms are prevalent and water depths are low) than at any other time of year and the greatest number of fires occurs at the peak of the seasonal thunderstorm pattern in July (Gu nderson and Snyder 1994) Sawgrass
70 is a fire adapted plant, growing quickly and recovering within 2 years post burn (Wade et al. 1980) Furt hermore, as sawgrass grows, the leaves spread away from the culm and senesce which helps promote fire, resulting in a wetland system that burns frequently (Gu nderson and Snyder 1994) State and federal agencies conduct prescribed burns to mimic fire return intervals, although often not during the same season as natural fires. Moreover, prescribed burns are conducted to manage habitat for a variety of wildlife species and protect ecological features (e.g., tree islands) on the landscape from catastrophic fires. Prescribed fires generally are conducted in winter and spring when at least 10 cm of surface water protects the underlying peat layer. In contrast, wildf ires that occur in dry years often burn the peat. Methods In 2009 and 2010, the Florida Fish and Wildlife Conservation Commission (FWC) conducted 3 burns each year in Water Conservation Area 3A of the Everglades that I used as treatments for effects of fir e on foraging success of wading birds (Fig. 1). These burns ranged from 548 to 1039 ha and were composed of approximately 70 85% sawgrass (Table 1). All prescribed burns occurred within a six week period from 15 February until 01 April. At the time of the burns, minimum estimated water depths ranged from 10 to 30.5 cm on the sawgrass ridges. Prey Item Survey During 2010, I surveyed 25 randomly selected locations in sawgrass in each burn unit for injured or dead biota that could serve as prey items for wadin g birds. I surveyed locations once pre burn and one day and one week post burn A set of random points (i.e., 50 100) was generated in ArcGIS within each burn border Points landing in slough habitat were not used since slough s do not burn. P oints falling more than 200 meters
71 into the burn from the edge of the slough were also not used for safety concerns. At each location, I recorded water depth and maximum vegetation height, and two people searched within a 0.5 m radius for potential prey items pre burn and injured or dead prey items post burn. I searched in the water and among burned sawgrass culms, but did not count live fish since my presence disturbed the se species Prey Density I used two trapping techniques to sample aquatic prey within an approxima tely 884 ha prescribed burn and an area of similar size immediately adjacent to the prescribed burn. The prescribed burn was conducted on 02 March 2011 in Water Conservation Area 3AS (Fig. 1). Minimum water depths were 15 20 cm at the time of burn. I gener ated random points within the sawgrass area of the burn unit and in an adjacent unburned area. The unburned sawgrass was east and adjacent to the prescribed burn. Starting one day post burn, I measured small fish and macroinvertebrate density and environme ntal characteristics at the random points in burned and unburned sawgrass ridges with 1 m 2 aluminum sided throw traps and Gee minnow traps (23 x 45 cm, 3.2 mm mesh, Memphis Net & Twine Co., Inc., Memphis, TN) I threw three throw traps in each sampling loc ation and removed vegetation to facilitate clearing of traps I cleared all traps with bar seine and dip net following methods of Jordan et al. (1997a) and Within each throw trap, I measured water depth, vegetation height, estimated percent periphyton cover to the nearest 5%, and counted sawgrass stems. At each sampling location with adequate water d epths ( 10 cm) I set 3 Gee minnow traps (23 x 45 cm, 3.2 mm mesh) for 2 hours. After 2 hours, I collected and preserved all aquatic organisms captured. O rganisms that were too large for the collection vials were measured in the field and
72 released. All org anisms were identified to species and measured (standard length (SL) for fish and snout:vent length (SVL) for amphibians) to the nearest mm Foraging Observations I observed foraging great egrets ( Ardea alba ) in burned sawgrass and sloughs adjoining burned tower mounted on an airboat. I selected individuals for observations that were foraging either singly or in groups and would be visible (i.e., not readily obstructed by vegetation or ot her wading birds) for much of the observation period. I usually observed individual birds for 5 min, though some observations were as long as 15 min. I accepted observations of less than 5 min duration if my view of the bird was obstructed by vegetation or the bird flew away. I counted the number of attempts to capture prey and number of successful attempts each bird made. I recorded number of individuals of each species in a flock. When observing a flock, I observed as many birds in each flock as possible until the birds flew away or I could no longer ensure I was observing an identifiable new bird. After observations were completed, I recorded water depth and coordinates at the foraging location. Foraging Habitat Selection In 2009 and 2010, I set up aerial survey transects (8 transects in 2009 and 5 in 2010) to cover 100% of the three burn units in each year and an equivalent adjacent area at the same latitude that would remain unburned (Fig. 1). Transects were oriented east west and separated by 1.33 km. A reas were surveyed weekly for 8 10 weeks until the sloughs dried or the wading birds dispersed. I flew with a second observer looking out the opposite side of a fixed wing Cessna 182 on transects at 244 m (800 ft) at 185 222 km/h (100 120 knots). We record ed species, number of individuals, and habitat in
73 which white foraging wading birds were observed on our side of the plane. Habitat categories were sawgrass, burned sawgrass, slough, sloughs adjacent to burns, and track (i.e., trails created by airboat s ). When groups of birds were >6, the observer would take one or more photos of the group to be counted later by two observers. Birds were categorized to species when possible. If they could not be identified to species, I categorized them as white wader or sm all blue heron. Groups were individually numbered so that if they overlapped habitats, they could be identified as a distinct unit. I digitized burned areas using Digital Orthophoto Quarter Quads (DOQQs) from 2003 in ArcGIS (Esri, Redlands, CA) based on p hotographs of the burn taken during flights from 305 610 m altitude. Area of each digitized patch was calculated for each burn. I digitized airboat track length within each survey area using files provided by FWC and DOQQs and calculated airboat track area by multiplying length by 2 m (approximate airboat width). Then, I clipped vegetation types using the vegetation data in Rutchey et al. (2005) in my survey areas and burn units (to determine area of sloughs adjacent to burns). I subtracted airboat track area (approximated to be 80% through grass and 20% through slough) from grass and slough since burned areas were digitized to exclude tracks. I reduced vegetation types from Rutchey et al. (2005) to m y categories, and calculated area of each habitat category for each week surveyed: burn, grass, tree island, slough, sloughs adjacent to burns, and airboat track. I used water depth data over the period of aerial surveys at gaging station 3A S_B to show wa ter level trends within the study area (SFWMD 2012) Water levels are reported relative to NGVD29.
74 Statistical Analysis Prey Due to the low number of potential prey items found, no statistic al analyses were performed on data collected during the prey item survey. From samples collected with a 1 m 2 throw trap, I checked normality of environmental variables (i.e., water depth, sawgrass stem and total stem density, vegetation height, and % perip hyton and vegetation cover), density of aquatic organisms (i.e., fish, crayfish ( Procambarus spp.), grass shrimp ( Palaemonetes paludosa ), amphibians, and aquatic invertebrates (>5 mm total length)), and length of aquatic organisms (fish and crayfish). I al so checked normality of length of fish and abundance of fish, crayfish, and amphibians caught in minnow traps. All environmental variables, crayfish density and length in throw traps and fish standard length in minnow traps and square root transformed aqua tic invertebrate density were normal and tested with a two sample t test for differences between burned and unburned sawgrass. All other variables of aquatic organisms in throw and minnow traps could not be normalized and I used a Kruskal Wallis rank sum t est to test for differences between burned and unburned sawgrass. Statistical Analysis Foraging Observations I constructed models of capture rate (number of captures per min) and capture efficiency (number of captures per attempt) for great egrets a prio ri M odels of capture rate were generalized linear models with a gamma distribution using a log link function I added 0.01 to capture rate in all models because there were zero values in capture rate. M odels of capture efficiency were generalized linear m odels with a quasibinomial distribution due to response type and number of zeroes in the data set. Models were constructed using a combination of water depth (linear or quadratic term), flock size, flock composition (single vs. mixed species), days since b urn, habitat (burn vs. sloughs
75 adjacent to burns), and year. I selected models of capture rate using the corrected AIC c ) and of capture efficiency using corrected QAIC (QAIC c ; Burnham and Anderson 2002) I rescaled AIC values i ) based on the A IC value of the best model (i.e., lowest AIC value), and calculated weighted values (w i ). I i < 2. Percent deviation (%D) was calculated from the null and residual deviances (i.e., %D = (null residual)/null) for m odels of capture rate and capture efficiency. Statistical Analysis Foraging Habitat Selection I calculated habitat selection ratios [((number of birds in each habitat per survey) / total number of birds per survey) / (amount of each habitat per survey / total amount of habitat per survey)] for each year following Manly et al. (2002) A selection ratio of 1 represents use of the habitat equal to its availability. Selection ratios < 1 indicate avoidance while selection ratios > 1 indicate selecti on for the habitat. I compared ratios in each habitat type to expected use via a Chi squared analysis and calculated 95% confidence limits using the Bonferroni correction. All statistical analyses were performed using R 2.10.1 (R Development Core Team 2009) Results Foraging Habitat Selection Great egrets had a high selection ratio (showing selection) for burns in the first 2 weeks (approximately 3.5 weeks after the first burn) of the surveys in 2009 (Fig. 3 2 Table 3 2). In 2010, great egrets similarly selected for burns for approximately 3 weeks after the first prescribed burn (Fig. 3 3 Table 3 3). In some habitat types, great egrets were not observed during surveys and no data point is included (Fig. 3 3). Great egr ets avoided burns when there was no standing water in the burn. G reat egrets only selected
76 for burns in proportion to the availability of burned areas in the survey area immediately after an additional prescribed burn in 2009 and for the last four surveys in 2010 (Tables 3 2 & 3 3). Conversely, g reat egrets strongly avoided unburned sawgrass in both years (Tables 3 2 & 3 3). Much of the use of unburned grass by great egrets occurred in thin strips of sawgrass in the edge between burn and sloughs adjacent to burns. In 2009, the selection ratio for sloughs adjacent to burns increased as water levels declined (Fig. 3 2 Table 3 2). To account for potential differences in water depths between sloughs, I compared great egret use of slough and sloughs adjacent to burns. Great egrets selected for both categories of slough equal to or more than their availability across all surveys (Tables 3 2 & 3 3). Two exceptions occurred in 2009 when great egrets avoided sloughs adjacent to burns in mid March and in the week afte r a new prescribed burn when g reat egrets selected for sloughs adjacent to burns and avoided sloughs even as water levels were dropping and ridge habitats were not available (Fig. 3 2 ). Great egrets used airboat tracks more than available (Figs. 3 2 & 3 3 ) especially in conjunction with a sharp increase in water levels, likely because tracks are typically deeper water than the surrounding sloughs. W hite ibis selected for burns in both years (particularly strongly in 2010), but avoided burns once water leve ls in burned areas were at or below the soil surface of the burned area (Tables 3 4 & 3 5, Figs. 3 4 & 3 5 ). In some habitat types, white ibis were not observed during surveys and no data point is included (Fig. 3 5). Conversely, white ibis strongly avoide d unburned saw grass stands in both years (Tables 3 4 & 3 5). Comparing white ibis use of sloughs adjacent to burns and slough i n 2009, white ibis selected for sloughs adjacent to burns more than their availability and selected for
77 sloughs about equal to th eir availability in the survey area (Table 3 4 & 3 5). In the week following the final prescribed burn in 2009 as water levels were receding, white ibis selected for sloughs adjacent to burns and avoided sloughs (Table 3 4, Fig. 3 4 ) To account for potent ial differences in water depth, I c ompared white ibis use of sloughs adjacent to burns and slough. White ibis selected for sloughs adjacent to burns and slough much more than they were available in 2010, except for one survey in mid April where it appears that ibis were selecting for burns (Fig. 3 5 ). Ibis strongly selected for airboat tracks in 2009 when water levels had receded the most, but only briefly used airboat tracks in 2010 when water levels were dropping (Figs. 3 4 & 3 5). Foraging Observations I observed a total of 104 foraging great egrets in 2009 and 2010. Capture rate of great egrets ranged from 0 3.2 per min with a mean of 0.43 per min across both years and all foraging locations. Capture rates in 2009 in sloughs adjacent to burns were higher than in sloughs and other Everglades habitat s in which capture rates have been previously quantified (Table 3 6). Water depths at foraging locations were deeper in 2010 than 2009 in sloughs adjacent to burns (Table 3 7). In 2009, water levels receded duri ng the sampling period, to the point that no surface water was available in the study area and no foraging was observed after 12 April. However, in 2010, while water levels initially declined, they rose again in early March and remained fairly steady until the end of observations on 30 March. The best model of great egret capture rate included flock size, days since burn, habitat, w ater depth, and flock composition (Table 3 8). Percent deviance of the best model was 26.9%. Great egret capture rate was great er in sloughs adjacent to burns than in burned sawgrass (Table 3 9). Water depth was positively related to capture rate,
78 however, there was also an interaction between depth and habitat that was negatively related to capture rate Capture rate did not diff er with depth in burns, but was negatively related to water depth in sloughs adjacent to burns. Days since burn was positively related to capture rate. In other words, capture rate of great egrets increased with time after the burn occurred. Great egrets f oraging in conspecific only flocks had a higher capture rate than i n multi specific flock s, but capture rate also declined as flock size grew. Effect size for water depth and flock size was very small indicating minimal contribution to the model (Table 3 9 ) and thus are not considered driving factors of capture rate. Capture efficiency (captures/strike) of great egrets ranged from 0 1 with a mean of 0.39 captures per attempt (Table 3 7). Capture efficiencies of this study fell with in the range of other stud ies of great egrets (Table 3 6). The models that best explained capture efficiency included flock size, flock composition, habitat, days since burn, and water depth (Table 3 10). One of the top two models ( c < 2) included an interaction between depth and habitat, however the c value was approximately 2, indicating that the additional variable did not change the likelihood of the model, but increased the c by the penalty term of 2 imposed by AIC for each additional variable in the model. Perce nt deviance of both models was approximately 15.8%. Water depth, foraging in conspecific flocks, and flock size were positively related to capture efficiency, however, effect size for water depth and flock size was very small, indicating minimal contributi on of these variables to the models (Table 3 9). Capture efficiency decreased with days since burn in sloughs adjacent to burns.
79 Prey Item Survey I conducted a single pre burn survey two to four weeks prior to each of two prescribed burns, and two sets of post burn surveys (1 day and 1 week post burn) on three prescribed burns. In pre burn surveys, I found six possible prey items, all spiders, at 6 of 50 points (0.15 items/m 2 ) Mean water depths pre burn were 8.5 cm in Berg burn and 14.5 cm in the 9.5 West burn with water depths at individual sampling locations ranging from 0 22 cm. Surveys immediately after the burn (1 day post burn) yielded 9 prey items at 7 of 75 points (0.12 items/m 2 ) Dead prey included one snail and two millipedes and live prey items i ncluded three spiders, two unidentified invertebrates, and one snail. Similarly, surveys one week after the burn yielded 13 potential prey items (0.22 items/m 2 ; three worms and two millipedes within a sawgrass culm, five spiders, one unidentified invertebr ates, and two snails: one live, one dead). Mean water depths ranged from 6.9 24.3 cm post burn with a range of 0 31 cm at individual sampling locations. Most of these invertebrates are unlikely to be actual prey items sufficient to cause wading birds to fo rage in a recent burn because they were small, hidden in sawgrass, and scarce Prey Density M ost sampling locations were within 10 m of the edge of the slough because water depth often was too shallow for sampling farther into the sawgrass stand Water dep ths in sampled locations were about 5 cm deeper in unburned than burned sawgrass (Table 3 11), suggesting that I inadvertently selected deeper water locations to sample unburned sawgrass despite selecting areas immediately adjacent to the burn that should have similar water depths. Stem density was significantly different, likely because stems of small plants (e.g., Eleocharis sp.) in burned areas were consumed
80 entirely by fire and were not present to be counted. Percent cover of vegetation and periphyton w ere significantly greater in unburned sawgrass. Density and size s of most potential prey items did not differ between burned sawgrass ridges and unburned sawgrass (Table 3 11). Amphibians (Peninsula newt ( Notophthalmus viridescens piaropicola ), siren ( Sire n lacertina ), tadpoles, and adult Florida cricket frog ( Acris gryllus dorsalis )) were at a significantly higher density in burns than unburned sawgrass in throw trap samples. Density of aquatic inver te brates (identified to lowest taxa possible : Belastomati d, Dysticid, Hirudinea, Odonata, Oligochaeta, and Pelocoris femoratus (alligator flea ) ) did not differ among treatments. Discussion Wading birds selected recently burned areas for a number of weeks post burn in greater proportion than was available. The re moval of above water vegetation by fire exposed these sawgrass ridges, which are shallower in water depth than the surrounding sloughs (Loveless 1959) Sawgrass on these ridges grows in dense, tall (mean: 0.8 1.5 m, but up to 3 m) stands (Gunder son 1994) that can inhibit movement and visibility by large animals such as wading birds. Through the removal of sawgrass by fire, this obstruction was eliminated, permitting wading bird access to these areas. Sawgrass starts growing almost immediately af ter a fire, resulting in changing vegetation conditions on burned ridges. While vegetation height increas ed it is unlikely over the duration of this study that wading birds avoided burned areas because of the increasing vegetation heights. Shallow water d epths are preferred by foraging wading birds given similar prey densities between accessible shallow and deep water habitats (Gawlik 2002) and depth may be what primarily attracts birds to these rec ently burned areas. Not only did wading birds show a preference for burned areas, but they also
81 remained in these areas over multiple weeks while water levels dropped and prey populations were depleted through foraging by wading birds and migration of prey to deeper water refuges. Field observations provide evidence that wading birds quit foraging in burned areas when water levels dropped below the marsh surface. One of my hypotheses was that wading birds are attracted to recently burned areas because of pr ey items injured or killed by the fire. I found almost no such potential prey items post burn, which were at levels insufficient to result in wading birds selecting burned areas over sloughs. The lack of potential prey items post burn is not surprising giv en the low density of potential prey items found pre burn. Burns in upland areas similarly yield few dead prey resources after the burn. Instead, many small mammals, insects, and other potential prey frequently flee the flame front, and are targeted during the fire by aerial and ground predators such as hawks, kestrels, and cattle egrets (Komarek 1969, Smallwood et al. 1982, Tewes 1984) Fish densities in burned and unburned sawgrass did not differ, indicating that b urning did not affect fish density in sawgrass. The densities in this study were typically lower than densities in sloughs in the Everglades (Loftus and Eklund 1994, Jordan 1996, Jordan et al. 1997b, Trexler et al. 2 002, Williams and Trexler 2006) Jordan (1996) found lower densities of fish in sawgrass than in sloughs. While water depths were shallower in this study than in other studies of fish density, the values of fish dens ity in this study were similar to a location sampled by Trexler et al. (2005) that also had shallow water depths. I was forced by low water levels to collect the majority of samples near the edge of the ridge due to shallow water levels farther onto the ridge, which may have biased measures of density or assemblage composition. However,
82 these samples are representative of the fish available to wading birds foraging in areas of burned sawgrass just before the burns have no standing water. Thus, while fish densities were not very high in burned areas relative to the sloughs, fish are available to foraging wading birds. Capture rate of great egrets was much higher in sloughs adjacent to burns than in burned a reas. Capture rates in this study fell in the range observed in other areas of the Everglades (Surdick 1998, Sizemore 2009, Lantz et al. 2010, 2011) although capture rates in 2010 were at the low end of those report ed. Fast prey capture rates may indicate that birds spend less time foraging to meet their energetic requirements, however, this metric gives no indication of the quality of the prey being captured or the quantity of prey available. For example, larger pre y (e.g., siren ) may require more handling time than small prey (e.g., small fish) although the caloric intake of large prey is typically much greater than a number of small prey captured in rapid succession. Given that capture rates were higher in sloughs adjacent to burns, it seems contrary to expectations that great egrets preferentially foraged in burned areas. However, sloughs are deeper water habitats than burned sawgrass and at the time that burns were available, many sloughs adjacent to burns may hav e had water levels deeper than appropriate for great egrets. While I observed foraging wading birds in an overlapping range of water depths in both habitat types, water depths in sloughs adjacent to burns where some great egrets foraged were at the upper l imit of water depths in which wading birds can forage and may have limited the inclusion of birds foraging in deep water habitats (Powell 1987, Gawlik 2002)
83 Overall, capture efficiency was greater in burns than in sloughs adjacent to burns despite lower prey densities. As with capture rate, mean values of capture efficiency were in the range of other capture efficiencies reported for great egrets foraging in the Everglades (Su rdick 1998, Sizemore 2009, Lantz et al. 2010, 2011) Greater capture efficiency in burned areas than sloughs adjacent to burns is compatible with the prediction that wading birds select burned ridges over sloughs. Burned areas have less submerged aquatic vegetation and almost no thick periphyton mat (pers. obs., Venne) within the water column, unlike sloughs. This provides less cover for fish and may enhance the ability of predators to see and capture prey. Burned sawgrass ridges provide shallow areas that wading birds appear to prefer more than sloughs that have deeper water and typically have higher prey densities. I found no evidence that the few potential prey items that were killed by the fire were sufficient to cause wading birds to select these areas for the purpose of scavenging. Wading birds appear to be selecting shallow water habitats despite lower capture rate. Habitat rather than foraging conditions may influence habitat selection (Gawlik 2002, Lantz et al 2010, 2011) which would explain why wading birds selected burned areas. While prescribed burns are a small percentage of the Everglades ecosystem, the removal of the sawgrass canopy by these burns provides shallow water habitats in which wading birds c an forage efficiently, albeit not at a fast rate. Regardless, wading birds must capture prey of sufficient caloric value while foraging. Prescribed fires are typically conducted during the dry season when water levels are dropping. Wading birds may have a limited window of opportunity to forage on burned ridges when water depths
84 are appropriate and before vegetation grows too tall. Fires conducted at another time of year may yield different results and should be explored.
85 Table 3 1. Description of prescri bed burns conducted by the Florida Fish and Wildlife Conservation Commission in Water Conservation Area 3A used for wading bird foraging observations and/or prey studies in 2009 2011. Burn a HeatNSmoaks b b Lost Lemon b Hackberry b,c Berg b,c 9.5 West c Apple Camp d Date burned 17 Feb. 09 26 Feb. 09 27 Mar. 09 16 Feb. 10 03 Mar. 10 01 Apr. 10 02 Mar. 11 Size (ha) 1003 931 1039 817 548 690 884 Last Yr Burned e 2004 2005 2005 W 2007 E 2006 1997 N 2005 2006 W Estimated % Habitat Composition Sa wgrass 70 70 85 70 75 70 67 Slough 14 19 7 29 13 25 15 Other 16 11 8 1 12 5 18 Fuel Density (%) Light 30 30 20 40 20 20 15 Moderate 50 55 70 40 25 35 15 Heavy 20 15 10 10 55 45 70 Weather Conditions Dispersion 45 62 70 60 48 42 55 Min. Mixing Ht 3000 400 0 5000 2700 2700 -4000 Onsite Conditions Time Taken 10:55 10:10 11:10 12:00 9:00 9:50 10:49 Wind NE 5/9 NE 6/9 SE 9/16 NW 7 W 5.3/8.9 NE 1.2/3.1 NE 11 RH (%) 60 61 62 52 75 80 60 Air Temp. 72 75 81 64 55 67 80 Flame Length (ft) 7 8 10 4 10 3 15 ROS 2 ft/min a Data taken from burn prescriptions provided by FWC. These are estimated percent habitat compositions. b Burn used for wading bird foraging observations. c Burn used for pre and post burn prey quantification. d Burn used for comparison of prey den sities. e W, E, and N designates burn occurred in west, east, and north, respectively, portion of burn unit in year listed.
86 Table 3 2. Great egret habitat selection ratio (Bonferroni adjusted 95% confidence interval) for 2009. Survey Date Slough adj. Burn a Burn Grass Slough Track Feb. 28 0.718 (0.18 1.26) ns 2.985 (2.03 3.94) 0.265 (0.13 0.40) 1.019 (0.75 1.29) ns 54.44 (34.84 74.04) Mar. 06 0.761 (0.36 1.17) ns 3.200 (2.49 3.91) 0.262 (0.16 0.36) 1.273 (1.07 1.48) ns 28.16 (17.22 39.09) Mar. 12 2.217 (1.75 2.6 8) 1.934 (1.51 2.36) 0.052 (0.03 0.08) 1.866 (1.75 1.98) 24.22 (17.83 30.60) Mar. 20 0.712 (0.50 0.93) ns 0.407 (0.25 0.57) 0.024 (0.01 0.04) 2.191 (2.09 2.29) 41.16 (34.20 48.11) Mar. 28 2.274 (1.78 2.77) 0.753 (0.49 1.02) ns 0.019 (0.00 0.04) 1.764 (1.60 1 .93) 43.83 (33.04 54.63) Apr. 03 4.080 (3.67 4.49) 0.230 (0.13 0.33) 0.041 (0.02 0.06) 1.346 (1.23 1.46) 51.21 (43.36 59.06) Apr. 10 0.935 (0.75 1.12) ns 0.005 (0.00 0.02) 0.016 (0.01 0.03) 2.624 (2.55 2.70) 24.25 (19.65 28.85) ns Chi square p value >0.05 fo r test of habitat selection different than expected a This is sloughs adjacent to burns.
87 Table 3 3. Great egret habitat selection ratio (Bonferroni adjusted 95% confidence interval) for 2010. Week Slough adj. Burn a Burn Grass Slough Track Feb. 15 NA NA 0. 154 (0.05 0.26) 2.434 (2.16 2.71) 20.99 (6.77 32.21) Feb. 26 1.486 (0.53 2.44) ns 1.775 (0.77 2.78) ns 0.540 (0.41 0.67) 1.830 (1.55 2.11) 1.673 (0.00 4.71) ns Mar. 05 2.095 (1.45 2.74) 2.898 (2.30 3.50) 0.479 (0.39 0.57) 1.249 (1.04 1.46) ns 4.366 (0.85 7.88) Mar. 13 2.401 (1.60 3.20) 0.543 (0.20 0.89) ns 0.253 (0.17 0.34) 2.325 (2.07 2.58) 8.977 (3.14 14.82) Mar. 19 3.213 (2.25 4.17) 0.386 (0.08 0.67) ns 0.046 (0.00 0.09) 2.644 (2.39 2.90) 8.167 (2.21 14.12) Mar. 25 2.349 (1.45 3.25) 0 b 0.026 (0.00 0.06) 2.989 (2.75 3.22) 9.188 (2.51 15.87) Mar. 31 2.394 (1.39 3.40) 0.971 (0.41 1.54) ns 0.120 (0.04 0.20) 1.899 (1.57 2.23) 40.89 (26.49 55.29) Apr. 08 2.836 (1.99 3.68) 0.868 (0.42 1.32) ns 0.063 (0.00 0.12) 2.385 (2.01 2.76) 7.934 (0.85 15.02) Apr. 15 2.747 (1.90 3. 59) 0.768 (0.34 1.20) ns 0.064 (0.00 0.13) 2.463 (2.09 2.84) 8.127 (0.87 15.38) Apr. 25 2.486 (1.42 3.55) 1.155 (0.49 1.82) ns 0.185 (0.05 0.32) 2.122 (1.63 2.62) 6.828 (0.00 15.48) ns Chi square p value >0.05 for test of habitat selection different than expe cted NA=not available, this survey occurred pre burn. a This is sloughs adjacent to burns. b Selection ratios of zero indicate that no birds were observed in this habitat.
88 Table 3 4. White ibis habitat selection ratio (Bonferroni adjusted 95% confidence in terval) for 2009. Survey Date Slough adj. Burn a Burn Grass Slough Track Feb. 28 2.567 (1.97 3.17) 3.476 (2.83 4.12) 0.235 (0.15 0.32) 1.123 (0.94 1.30) ns 0 b Mar. 06 2.165 (1.77 2.56) 1.801 (1.45 2.16) 0.432 (0.36 0.51) 1.365 (1.24 1.49) 2.737 (0.52 4.95) ns Mar. 12 3.071 (2.74 3.40) 1.979 (1.71 2.25) 0.097 (0.08 0.12) 1.795 (1.72 1.87) 9.960 (7.33 12.59) Mar. 20 0.591 (0.43 0.75) 0.544 (0.40 0.69) 0.217 (0.18 0.26) 2.259 (2.18 2.34) 13.32 (9.88 16.76) Mar. 28 1.417 (1.13 1.71) 1.792 (1.52 2.06) 0.008 (0.00 0 .02) 1.920 (1.80 2.04) 19.70 (14.34 25.07) Apr. 03 6.151 (5.67 6.64) 0 b 0.050 (0.02 0.08) 0.855 (0.74 0.97) ns 51.14 (42.27 60.00) Apr. 10 2.231 (1.89 2.57) 0 b 0.013 (0.00 0.03) 2.462 (2.36 2.56) 8.456 (4.93 11.98) ns Chi square p value >0.05 for test of hab itat selection different than expected a This is sloughs adjacent to burns. b Selection ratios of zero indicate that no birds were observed in this habitat.
89 Table 3 5. White ibis habitat selection ratio (Bonferroni adjusted 95% confidence interval) for 20 10. Survey Date Slough adj. Burn a Burn Grass Slough Track Feb. 15 NA NA 0.271 (0.13 0.41) 2.428 (2.13 2.73) 5.529 (0.00 13.64) ns Feb. 26 6.443 (5.26 7.62) 5.159 (4.11 6.21) 0.264 (0.20 0.33) 1.226 (1.04 1.41) ns 0.745 (0.00 2.10) ns Mar. 05 1.168 (0.92 1.41) ns 5.410 (5.07 5.75) 0.099 (0.08 0.12) 1.553 (1.45 1.66) 0 b Mar. 13 2.156 (1.51 2.80) 2.416 (1.86 2.97) 0.011 (0.00 0.03) 2.464 (2.26 2.67) 0 b Mar. 19 2.774 (1.96 3.59) 0.563 (0.23 0.90) ns 0.070 (0.02 0.12) 2.785 (2.57 3.00) 0 b Mar. 25 0 b 0.192 (0.00 0.47) ns 0.048 (0.00 0.10) 3.584 (3.44 3.73) 0 b Mar. 31 0 b 0 b 0 b 0 b 0 b Apr. 08 4.061 (0.00 8.84) ns 5.179 (1.53 8.83) 0 b 0 b 0 b Apr. 15 0.241 (0.00 0.85) ns 7.182 (6.10 8.26) 0.083 (0.00 0.23) 0.283 (0.00 0.69) ns 0 b Apr. 25 0 b 0 b 0 b 0 b 0 b ns Chi square p value >0.05 for test of habitat selection different than expected NA=not available, this survey occurred pre burn. a This is sloughs adjacent to burns. b Selection ratios of zero indicate that no birds were observed in this habitat.
90 Table 3 6. C apture rates (capture s per minute) and capture efficiencies (captures per attempt) reported for the great egret ( Ardea alba ) in southern Florida marshes. Year or Capture Capture Study Condition Rate (N) Efficiency (N) Location Surdick (1998) 1996 0.4 (292) NA Everglades 1 997 0.2 (593) NA Sizemore (2009) 2008 0.46 (82) 0.60 (76) Agricultural fields 2009 0.34 (130) 0.47 (115) Lantz et al. (2010) Jan. shallow 0.19 0.29 (35) 0.30 0.60 (29) SAV a density experiment Jan. deep 0.26 1.58 (19) 0.56 1.0 (16) Apr. shallow 0.23 0.75 (12) 0.33 0.75 (11) Lantz et al. (2011) 2008 0 0.66 (12) 0.13 0.34 (11) Emergent vegetation experiment This study 2009 0.59 (60) 0.40 (60) Everglades WCA 3A 2010 0.18 (38) 0.35 (38) a SAV is submerged aquatic vegetation
91 Table 3 7. Capture rates (captures per minute) and capture efficiencies (captures per attempt) of great egret ( Ardea alba ) in 2009 and 2010 in Water Conservation Area 3A of the Everglades, USA. 2009 2010 Variable Burn Slough adj. Burn a Burn Slough adj. Burn a Number of observa tions 17 43 14 24 Mean Capture Rate ( sd) 0.30 (0.3) 0.71 (0.9) 0.07 (0.1) 0.24 (0.2) Range of Capture Rate 0 0.9 0 3.2 0 0.4 0 0.8 Mean Capture Efficiency ( sd) 0.46 (0.4) 0.38 (0.3) 0.18 (0.3) 0.45 (0.4) Range of Capture Efficiency 0 1 0 1 0 1 0 1 Mean Attempts per minute 0.6 (0.5) 1.3 (1.2) 0.3 (0.3) 0.5 (0.4) Water depth (cm) 12.1 (8.9) 16.5 (4.6) 13.9 (4.9) 22.7 (4.4) Range of water depth (cm) 0 21 8 25 7 21 14 30 a Sloughs adjacent to burns
92 Table 3 8. Candidate set of models of great egret capture Information Criterion (AICc) to select generalized linear models constructed with environmental characteristics in foraging locations in Water Conservation Area 3A of the Everglades, USA, 2009 and 2010. Model Model Variables a k b i w i m5a D, FS, FC, Hab, dSB, D*Hab 8 0.0 0.768 m18a D2, FS, FC, dSB, Hab, D2*Hab 8 4.5 0.081 m5 D, FS, Hab, dSB 6 4.9 0.066 m20 Yr, FS, FC, Hab, dSB 7 5.6 0.046 m18 D2, FS, FC, Hab, dSB 7 7.0 0.023 m4 D, FS, Hab 5 9.7 0.006 m7 FS, Hab 4 10.6 0.004 m17 D2, FS, FC, Hab 6 10.9 0.003 m11 FS, FC, Hab, dSB 6 12.2 0.002 m10 FS, FC, Hab, dSB 5 12.8 0.001 m00 Yr, Hab, dSB 5 18.0 < 0.001 m001 D, Hab, Hab*D 5 19.2 < 0.001 m9 FS, dSB 4 20.4 < 0.001 m16 D2, FS, FC 5 23.2 < 0.001 m6 FS 3 24.2 < 0.001 m3 D, FS, FC 5 2 4.6 < 0.001 m19 Yr 3 25.2 < 0.001 m8 FS, FC 4 25.9 < 0.001 m2 D, Hab 4 28.1 < 0.001 m13 Hab, dSB 4 31.1 < 0.001 m12 Hab 3 31.1 < 0.001 m15 D2 3 34.3 < 0.001 m1 D 3 34.7 < 0.001 m14 dSB 3 35.0 < 0.001 a D = depth, D2 = depth squared, dSB = days since burned, FC = flock composition, FS = flock size, Hab = habitat bird was foraging in, Yr = year, D*Hab = interaction of depth and habitat, D2*Hab = interaction of depth squared and habitat b Number of parameters included within the model
93 Table 3 9. Coefficients Information Criteria (AIC c ) and capture efficiency (Effic iency ) selected using corrected quasi AIC (QAIC c ). Models of capture rate use a gamma distribution an d capture efficiency use a quasibinomial distribution. Variable Flock Comp Habitat Intercept Depth Flock Size Single spp. BSL a dSB a D*Hab a a Rate b 2.37 (0.54) 0.003 (0.05) 0.010 (0.007) 0.696 (0.46) 3.99 (0.95) 0.035 (0.02) 0.170 (0.06) 0.0* Ef fic iency 0.269 (0.63) 0.069 (0.03) 0.011 (0.007) 0.232 (0.47) 0.890 (0.49) 0.035 (0.02) 0.0* 0.377 (0.96) 0.078 (0.07) 0.010 (0.007) 0.244 (0.48) 0.740 (1.1) 0.035 (0.02) 0.011 (0.07) 1.98 a ue between best model and the given model b Models of capture rate are (capture ra te + 0.01) = (explanatory variables) because zeroes cannot be log transformed. See Methods for more details.
94 Table 3 10. Candidate set of models of great egret capture efficiency using corrected quasi eralized linear models constructed with environmental characteristics in foraging locations in Water Conservation Area 3A of the Everglades, USA, 2009 and 2010. Model Model Variables a k b QAIC c i w i m5 D, FS, FC, Hab, dSB 7 85.5 0.0 0.304 m5a D, FS, FC, Hab, dSB, D*Hab 8 87.5 1.98 0.113 m4 D, FS, FC, Hab 6 88.0 2.47 0.088 m20 Yr, FS, FC, Hab, dSB 7 88.0 2.52 0.086 m18 D2, FS, FC, dSB 6 88.4 2.87 0.072 m7 FS, Hab 4 89.0 3.47 0.054 m17 D2, F S, FC, Hab 6 89.2 3.72 0.047 m11 FS, FC, Hab, dSB 6 89.3 3.75 0.047 m18a D2, FS, FC, dSB, Hab, D2*Hab 8 89.4 3.92 0.043 m9 FS, dSB 4 89.9 4.36 0.034 m10 FS, FC, Hab 5 90.1 4.62 0.030 m6 FS 3 90.3 4.75 0.028 m3 D, FS, FC 5 90.6 5.05 0.024 m16 D2, FS, FC 5 91.3 5.82 0.017 m8 FS, FC 4 92.0 6.50 0.012 m1 D 3 102.4 16.9 <0.001 m15 D2 3 103.3 17.7 <0.001 m19 Yr 3 103.5 17.9 <0.001 m2 D, Hab 4 103.6 18.1 <0.001 m14 dSB 3 104.1 18.6 <0.001 m001 D, Hab, Hab*D 5 105.0 19.5 <0.001 m12 Hab 3 105.3 19.7 <0.001 m00 Yr, Hab, dSB 5 105.3 19.8 <0.001 m13 Hab, dSB 4 105.8 20.3 <0.001 a D = depth, D2 = depth squared, dSB = days since burned, FC = flock composition, FS = flock size, Hab = habitat bird was foraging in, Yr = year, D*Hab = interaction of depth and habitat, D2* Hab = interaction of depth squared and habitat b Number of parameters included within the model
95 Table 3 11. Mean ( standard deviation) of environmental variables and aquatic organisms in locations sampled with 1 m 2 throw trap and minnow trap in Water Co nservation 3AS of the Everglades, USA, in 2011. Burned Unburned Number Variable Sawgrass Sawgrass t df p a of plots b N 17 13 Water depth (cm) 11.9 (2.8) 15.7 (3.7) 3.02 21.7 0.01 Sawgrass density (stems m 2 ) 33.9 (8.7) 27.8 (8.5) 1.95 26.2 0.06 Stem density (stems m 2 ) 40.9 (9.1) 49.0 (11.2) 2.12 22.7 0.05 Vegetation height (cm) 52.3 (10.1) 131.9 (16.9) 15.1 18.5 <0.01 Vegetation cover (%) 49.0 (14.2) 60.4 (11.8) 2.39 27.8 0.02 Periphyton cover (%) 13.1 (13.6) 25.2 (14.7) 2.79 27.5 0.01 1 m 2 Throw Traps Fish density (m 2 ) 3.6 (5.1) 3.1 (3.3) 0.97 2 ) 2.8 (4.6) 2.6 (2.8) 0.64 Fish (>20 mm) density (m 2 ) 0.8 (1.5) 0.5 (0.6) 0.51 Mean Fish SL (mm) 17.9 (4.6) 16.8 (2.7) 0.95 13,10 Crayfish density (m 2 ) 1.4 (0.8) 1.9 (1.3) 1.16 18.2 0.26 Mean Crayfish length (mm) 31.7 (3.9) 32.8 (2.7) 0.90 27.0 0.38 17,12 Shrimp density (m 2 ) 4.3 (11.7) 1.5 (3.3) 0.88 9,6 Amphibian density (m 2 ) 1.5 (1.4) 0.5 (0.6) 0.01 15,8 Aquatic invert. density (m 2 ) 4.1 (2.6) 3.5 (3.2) 0.72 22.7 0.48 17,12 Minnow Traps Plots sampled 11 12 Fish abundance 1.8 (3.4) 2.1 (2.4) 0.35 Mean Fish SL (mm) 19.5 (6.1) 20.5 (8.3) 0.30 14.7 0.77 9,9 Crayfish abundance 0.09 (0.2) 0.17 (0.3) 0.64 Mean Crayfish length (mm) 36.0 (9.9) 27.2 (5.9) NA 2,3 Amphibian abundance 0.18 (0.2) 0.08 (0.2) 0.27 Mean Amphibian SVL 14.7 (3.8) 21.0 (0.0) NA 5,2 a p values without accompanying values for t and degrees of freedom (df) are from a Kruskal Wallis rank sum test. b Number of plots in w hich the given species was captured. Average lengths were calculated using this N.
96 Figure 3 1. Map of study area including prescribed burns conducted in 2009 2011 used in various components of this study. See Table 3 1 and Methods for details of the b urns and uses.
97 Figure 3 2. Habitat selection ratio (bars represent standard error) for great egrets ( Ardea alba sloughs adjacent to burns. Surface water depth is water level above NGVD29.
98 Figure 3 3. Habitat selection ratio (bars represent standard error) for great egrets ( Ardea alba sloughs adjacent to burns. Surface water depth is water level above NGVD29.
99 Figure 3 4. Ha bitat selection ratio (bars represent standard error) for white ibis ( Eudocimus albus ) in 2009 in the central Everglades, USA. designates sloughs adjacent to burns. Surface water depth is water level above NGVD29.
100 Figure 3 5 Habitat selecti on ratio (bars represent standard error) for white ibis ( Eudocimus albus designates sloughs adjacent to burns. Surface water depth is water level above NGVD29.
101 CHAPTER 4 EFFECTS OF FIRE ON P ERIPHYTON PR IMARY PRODUCTION AND FISH STANDING STOCK IN AN OLIGOTROPHIC WETLAND Introduction Fire is a natural disturbance in many upland and wetland ecosystems that, through combustion of vegetation, exposes the underlying substrate to light and redistributes nutrien ts important to primary production. In uplands, fire typically alters nutrient availability, increase s nutritive content in post fire vegetation, changes vegetative cover and structure, and influence s animal utilization of the landscape (Whelan 1995) Similarly, in wetlands where fire occurs, fire has been shown to remobilize nutrients (Smith et al. 2001, Qian et al. 2009) alter plant cover, structure, and compos ition (Smith and Newman 2001) and promote new vegetative growth (Lugo 1995) with enhanced nutritional content (Smith et al. 1984) Many aquatic invertebrates respond to changes in veget ation post burn via increasing biomass, density, and abundance (de Szalay and Resh 1997, Munro et al. 2009, Beganyi and Batzer 2011) however, alternate hypotheses such as availability of food resources and alterati on of microclimate may better explain use patterns of invertebrates (Hochkirch and Adorf 2007) Most studies concerning the effect of fire on fish generally focus on mountainous wate rsheds where sediment runoff negatively impacts water quality or reduced shading after a wildfire increases stream temperature (Gresswell 1999) neglecting effects on fish of increased food resources due to fire. While a good understanding of how fire affect s nutrient cycling and macrophytes in wetlands has been developed, we do not understand how fire impacts other aspects of wetlands such as periphyton fish, and higher trophic levels.
102 Light is a key factor in determining primary production and composition of the algal assemblage (Mosisch et al. 2001) In temporary ponds and streams, an increase in light increase d algal biomass (Mosisch et al. 2001, Mokany et al. 2008) while low light levels often result in decreased algal biomass (Hillebrand 2005) High light conditions often result in the presence of larger species of algae, which alters algal species composition and growth form of the algal assemblage. In the Everglades in southern Florida substantially less periphyton exists in sawgrass stands than in wet p rairies and sloughs (McCormick et al. 1998) This is attributed to shading from dense macrophyte communities (Gr imshaw et al. 1997, Thomas et al. 2006) However, shading does not change composition of periphyton in the Everglades, but it does reduce gross photosynthesis and percent organic matter at very high levels of shading (98% shade; Thomas et al. 2006) Nutrients, specifically phosphorus (P), also initiate changes in algal biomass and shifts in species composition (Mosisch et al. 2001, Gaiser et al. 2011) Fire alters nutrient availability, typically resulting in increased bioavailability of P (Smith et al. 2001) In the Everglades where P is limited, remobilization of bioavailable P can be crucial for components of the ecosystem such as periphyton. In a P dosing experiment, periphyton biomass increas ed within 18 days at doses of 32 mg P/m 2 /wk (McCormick and Scinto 1999) How ever, at chronic, low level P loads, floating periphyton mats are lost and biomass decreases, as the composition of the algal assemblage shifts from cyanobacteria to other algal species (Gaiser et al. 2004) I ncreases in periphyton P concentrations result in greater productivity of alga e that may outcompete certain diatom taxa (Gaiser et al. 2006) This suggests that even small pulse s of nutrients from
103 a fire in an oligotrophic wetland may be able to affect primary production, and possibly have indirect effects on other trophic levels. An increase in periphyton biomass can provide more food resources to consumers depending on the species composition of the per iphyton mat (Rader and Richardson 1992, Geddes and Trexler 2003) Many algal species employ protective mechanisms (e.g., toxins, calcite encrustation) to avoid herbivory, thereby affecting edibility of the periphyto n mat (Browder et al. 1994, Chick et al. 2008) Increase d algal biomass resulted in a shift in the community of consumers from filter f eeders to algal grazers in temporary ponds (Mokany et al. 2008) Similarly, periphyton rich in green algae and diatoms is a preferred food for wetland herbivores (McCormick and Scinto 1999) Tadpoles increased their growth and weight when eating periphyton rich in green periphyton and diatoms rather than blue green algae (Browder 1981) However, the loss of p eriphyton mats due to repeated P inputs resulted in decomposition of periphyton associated vegetation (eastern purple bladderwort) and changes of faunal (fluctuation in fish biomass) assemblages (Gaiser et al. 2005) Additionally, density of the macroinvertebrate community is reduced without periphyton mats (i.e., no habitat available; Liston et al. 2008) Thus, a pulse of nutrients and increase in light, such as result from fires, may increase biomass and alter algal species composition sufficien tly to alter the aquatic consumer community, including species that serve as key links to higher trophic levels. I conducted a field experiment in which I manipulated light and nutrients in order to determine how fire affects oligotrophic wetlands by alter ing primary production and fish standing stock. I predicted that 1) an increase in light and nutrients would result in more
104 periphyton biomass and cover and 2) additional available resources, assuming an increase in periphyton biomass, would increase total and individual fish size, condition factor, and relative abundance. Methods The Everglades is a large, oligotrophic, P limited wetland in southern Florida, USA (Noe et al. 2001) Sawgrass ( Cladium jamai cense ) is the dominant vegetation and forms contain periphyton mats, submerged aquatic vegetation, and some emergent vegetation (Gunderson 1994) Sawgrass is a fast growing, fire adapted plant with leaves that grow out from the culm and senesce, with stands typically recovering within 2 years post burn (Wade et al. 1980) This growth form, coupled with a high frequency of lightning, promotes fire (Wade et al. 1980) resulting in a wetland system that burns frequently, primarily at the onset of the wet season (Gunderson and Snyder 1994, Slocum et al. 2007) I set up a 2x2 factorial ex periment in which I manipulated nutrients and light in 20 10 m x 10 m plots in sawgrass ridges (Fig. 4 1) Nutrient treatments were either burned (added nutrients from a prescribed burn) or mowed with mowed vegetation removed from plots (no nutrients added ), based on the assumption that a fire temporarily increases concentrations of available nutrients, and that mowing with removal of above water vegetation would mimic the light increase typical following burns but not add nutrients. Light treatments were plots with and without shade houses to mimic natural shading from sawgrass. Shade cloth was selected using light levels measured for photosynthetically active radiation (PAR) using an AccuPAR LP 80 (Decagon Devices, Pullman, WA) in sawgrass at five locatio ns in sawgrass stands in the study area (63
105 95% shading, x = 84% 12%). Treatments were : burned (nutrients and light added N no nutrients, light added L shade house (control with no nutrients and no light added added a fifth treatment in a stand of unmanipulated sawgrass as an experimental eatment in the 2x2 factorial design was manipulated just like the other treatments. These plots served more as a control of the treatment process than a true control. Hereafter, I refer to this fifth treatment as Burned plots were l ocated within the sawgrass ridges ( >1 ha) of a much larger prescribed burn. This ensured that burn effects were representative of the management tool and minimized edge and small plot effects of fire. I used ArcMap (Esri, Redlands, CA) to randomly select e ight points, four in the prescribed burn unit, and four in an adjacent area that was not burned, from which bearings were randomly selected to place plot locations in the next nearest, sufficiently large (>1 ha) sawgrass ridge. Plots were positioned in saw grass 30 45 m from the edge of the ridge to reduce edge effects. The Florida Fish and Wildlife Conservation Commission conducted the prescribed burn on 01 April 2010. The burn unit was approximately 690 ha and incorporated approximately 70% sawgrass, 25% s lough, with woody tree islands, cattail ( Typha spp.), and willow ( Salix spp.) composing the remainder. Approximately 45% of the overall fuel density was considered heavy. The fire was conducted over standing water ( x = 17.5 cm), and was a comp lete burn, leaving sawgrass and buttonbush ( Cephalanthus occidentalis ) stubble standing approximately 32 cm above the water surface, typical of burns with standing water in the Everglades. Between 1 4 April 2010,
106 I set up treatment plots. I mowed an area o f 12 m x 12 m with articulating hedge trimmers to a vegetation height above the marsh surface approximately equivalent to burned vegetation ( x = 47.5 cm). I dragged the mowed vegetation >50 m away to areas of deeper water. The extra area was m owed to reduce refuge for aquatic organisms in standing sawgrass on the edge of the plots. On the day following vegetation removal (burned or mowed), I constructed 10 m x 10 m x 2 m (l x w x h) shade houses of 80% spectrally neutral black knitted cloth (In ternational Greenhouse Company, Danville, IL, USA) in plots without light. I also mock disturbed light treatments that did not get shade houses s ince the shade house (i.e., no light) plots were trampled by people during set up Experimental Control plots w ere not trampled. I collected water samples for analysis of phosphorus from burned plots i n the morning before the burn occurred, in the afternoon shortly after the burn was completed (day 1), and on days 2, 3, 5, 7, 9, 12, and 15. Samples in burned treatm ents (i.e., house treatment set up on P concentrations. I collected samples in mowed treatments 3 days) prior to treatment, immediately after mowing (labeled day 0.5 if a shade house was constructed the following day to complete the treatment), one day after setting up the treatment, and in two E xperimental C ontrol sites the morning before the burn and on day 5. Water samples were placed on ice and processed in the evening of the day sampled. I transferred and acidified 40 mL of water to analyze for total phosphorus (TP) and filtered and acidified 40 mL of water to analyze for soluble reactive phosphorus (SRP). Samples were kept at 4C and analyzed by the National Environmental Laboratory Accreditation Program (NELAP) certified University
107 of Florida Wetland Biogeochemistry Laboratory (Gainesville, FL) within 2.5 months of collection. Additionally, at all water sample locations I collected l arge clumps of periphyton (<1 L), where present, prior to and immediately after the prescribed burn. These samples were ashed and analyzed for TP at the University of Florida Wetland Biogeochemistry Laboratory. I sampled plots once every 10 days starting 2 days after all plots were set up for a total of eight sampling periods from early April to the end of June 2010 In all plots, I measured water depth, average and maximum vegetation height, and I haphazardly placed a 0.25 m 2 quadrat to estimate percent v egetation cover, percent periphyton cover, and percent periphyton collected. Percent vegetation cover was estimated as percent of area within the water column filled by vegetation, typically using the consensus of two observers. Percent periphyton cover wa s visually estimated as percent of marsh bottom, vegetative material in water column, and water surface covered by periphyton. The periphyton I collected was stored in plastic bags on ice and transferred to an approximately 20C freezer within 7 h of coll ection. I sampled the fish assemblage using minnow traps. In each plot I set 3 Gee minnow traps (23 x 45 cm, 3.2 mm mesh, Memphis Net & Twine Co., Inc., Memphis, TN) for 2 hours. After 2 hours, I identified and measured total length (TL, 1 mm) and mass ( 0.1 g) of each aquatic organism captured. During the last sampling period at the end of June (approximately day 90 of the study), I also used 1 m 2 throw traps to sample all mowed and burned plots. I threw two traps in each plot, cleared traps following m ethods of Jordan et al. (1997a) and measured TL of the organisms
108 captured. I used equations relating TL to standard length (SL) for individual fish specie s (D. Gawlik, pers. comm.) to convert measurements. I measured PAR in three locations in each plot once per month between April and June, 2010 Readings were taken between 1000 and 1400 h to standardize the sun azimuth every minute for 15 minutes with some exceptions due to equipment difficulties. I used as many of the readings as possible so plots had 3 15 readings. Readings at all locations in each plot were pooled and averaged to calculate percent shading in the plots. Periphyton samples were analyzed fo r chlorophyll a following the methods of Sartory and Grobbelaar (1984) at a NELAP certified University of Florida laboratory (Gainesville, FL). A small (~20 mg, wet we ight) subsample of periphyton was weighed and processed while the remaining periphyton from each plot was used to determine wet:dry weight ratios to calculate periphyton biomass. From this, I calculated corrected chlorphyll a ( g/g) and periphyton mass (g) on a dry weight (dw) basis per area (m 2 ). I tested normality of e nvironmental variables (vegetation variables, water depth, and percent shading) with a Shapiro Wilk normality test. For percent periphyton cover, average and maximum vegetation height and pe rcent vegetation cover, a transformation did not achieve normality, so I rank transformed data and analyzed for differences among treatments using Kruskal Wallis rank sum tests. I analyzed water depth, percent vegetation cover, and percent shading with an analysis of covariance (ANCOVA) using sampling period as the covariate. For TP and SRP, I compared daily concentrations post burn to pre burn concentrations using paired Wilcoxon signed rank tests.
109 Fish richness and Shannon Wiener diversity were not normal ly distributed so data were rank transformed and analyzed for differences among treatments using Kruskal Wallis tests. I analyzed f ish metrics (standard length, mass, condition factor, abundance, and relative abundance) for species that were caught in at l east 80% of plot sampling period combinations. I analyzed all species combined and individual species using generalized least squares ( gls ) in R (R Development Core Team 2009) Due to repeated sampling of the same plo ts, I used models incorporating autoregressive variance covariance structure and compared models with and without the assumption of heterogeneous variances to determine if there were significant differences among treatments or between sampling periods. I i nspected histograms of the residuals and plots of the residuals versus predicted values to determine if transformation was necessary. If there was a significant difference among treatments, I set up three contrasts to compare Experimental Control vs. treat ments, light vs. no light, and nutrients vs. no nutrients. To maintain orthogonality of the contrasts, I omitted the Experimental Control treatment from the latter two contrasts. Relative abundance of fish was organized into a species x site matrix. I four th root transformed each response variable to reduce the weight of dominant species on more rare species. I conducted an analysis of similarities (ANOSIM; Clarke 1993) for each of these community matrices, using 999 permutations and Euclidean distances for mass and conditi abundance. Permutations were limited to within each period since there were significant differences among some periods and response variables.
110 Results Nutrients Immediately following c ompletion of plot treatments, TP concentrations in water increased significantly (F=12.2, df=1, p<0.001) with concentrations in the N utrient treatments (i.e., burned) as high as 0.161 mg/L ( x = 0.077 mg/L; Fig. 4 2). The day after the burn (da y 2), the average concentration was 0.024 mg/L (max. = 0.053 mg/L). TP concentrations leveled off at approximately 0.014 mg/L, remaining significantly different than the average pre burn TP concentration (0.008 mg/L) although the average difference was onl y 0.006 mg/L. On day 5, there was an elevation to 0.034 mg/L after concentration had declined from the initial peak. On day 15, TP concentrations dropped to 0.005 mg/L, lower than pre burn concentrations. Concentrations in N o N utrients and Experimental Con trol treatments were not significantly different than pre treatment TP concentrations. SRP concentrations in nutrient treatments spiked significantly (F=6.64, df=1, p=0.012) immediately post treatment to an average of 0.047 mg/L (max. = 0.163 mg/L; Fig. 4 2). SRP concentrations returned to close to detection limits the following day ( x = 0.003 mg/L, max. = 0.013 mg/L) and remained low thereafter with the exception of day 5, corresponding with a spike in TP concentrations, when SRP concentration s were elevated to 0.020 mg/L. Environmental Factors Water depth was significantly shallower in the N utrients O nly treatment plots compared to both treatments without nutrients, but did not change much over the course of the experiment (Tables 4 1 & 4 2). Following treatment, s haded treatments (No N utrients or L ight and N utrients O nly), Experimental Control and L ight O nly, and light treatments ( L ight O nly and N utrients + L ight) were not significantly different in
111 percent shading but all other treatments we re significantly different (Table 4 2) from each other, indicating that the intended light treatments were effective. Percent vegetation cover did not differ among treatments over the course of the experiment, however average and maximum vegetation heights were significantly higher in Experimental Control than all other treatments (Table 4 1 & 4 3). Vegetation grew over time, driven primarily by all manipulated plots and excluding Experimental C ontrol (t= 9.065 df=1,125, p<0.001, adjusted R 2 =0.39). Periphyt on Percent periphyton cover was greater in the N utrient + L ight treatment than in all other treatments (Table 4 3). Similarly, on a dry weight (dw) basis, periphyton mass per area and percent periphyton cover were greater in the N utrient + L ight treatment than in all other treatments (Table 4 3). Concentrations of chlorophyll a were not significantly different among treatments (Table 4 3). Overall Fish Metrics I captured 10 species of fish 2 species of invertebrates, 3 species of amphibians, 1 species of re ptile, and 2 species of crustaceans in minnow traps (Table 4 4). Four species of fish (eastern mosquitofish ( Gambusia holbrooki ), least killifish ( Heterandria formosa ), Flagfish ( Jordanella floridae ), sailfin molly ( Poecilia latipinna )) were captured in ne arly all plots during the study. Three of the fish species were captured very infrequently. The number of fish captured was significantly higher in the N utrient + L ight treatment compared to the N o N utrients or L ight treatment, but did not differ from the other treatments (Tables 4 1 & 4 5 ). Overall fish mass, length, condition factor, relative abundance, and richness were not significantly different among treatments (Tables 4 1,
112 4 3 & 4 5 ). Diversity of fishes was significantly higher in treatments with li ght and lowest in the Experimental Control and N utrients O nly treatments (Tables 4 1 & 4 3). In the final period when throw traps were also used to sample plots, fish density was greater in plots with light (Tables 4 6 & 4 7). Catch per unit effort of min now traps and density of fish in throw traps in this final period were related (Adjusted R 2 =0.257; Fig. 4 3). There were no differences associated with changes in nutrients (Table 4 6). Standard length of all fish and individual species of fish did not dif fer among treatments with the exception of Everglades pygmy sunfish and marsh killifish (Table 4 7). Everglades pygmy sunfish were significantly longer in treatments with Light ( L ight O nly and N utrients + L ight) than No Light ( Nutrients O nly and No N utrien ts or L ight) while marsh killifish were longer in Nutrient treatments ( Nutrients O nly and N utrients + L ight) than No Nutrients treatments ( L ight O nly and No N utrients or L ight; Tables 4 6 & 4 8). However, there was a significant interaction between light a nd nutrients for marsh killifish length where lengths under conditions of Light were greater in Nutrient than No Nutrient treatments, but lengths in No Light treatments were similar with and without nutrients. I used contrasts to compare differences of com bined means of factors for four fish species that were captured in 80% or more of the plots across all sampling periods (Table 4 8). Flagfish were smaller (length and mass) and had lower relative abundance in Experimental C ontrol treatments than the combin ed mean of all other treatments (Tables 4 6 & 4 8, Fig. 4 4). Additionally, flagfish were bigger (length and mass) in No Nutrient treatments compared to Nutrient treatments. I caught more flagfish in Light treatments than no light treatments. Sailfin molli es were heavier in No Nutrient
113 treatments compared to Nutrient treatments (Tables 4 6 & 4 8, Fig. 4 5). Least killifish were longer in No Light treatments than Light treatments and longer in Nutrient treatments than No Nutrient treatments (Tables 4 6 & 4 8 Fig. 4 6). Least killifish had a higher condition factor in No Nutrient treatments than Nutrient treatments. In Experimental Control plots, least killifish had higher relative abundance than treatment plots, and in plots with Nutrients, relative abundanc e of least killifish was lower than No Nutrient treatment plots. Eastern mosquitofish showed no differences among treatments (Tables 4 6 & 4 8 Fig. 4 7). Marsh killifish and golden topminnow were captured in 124 and 93, respectively, out of 160 plot sampl ing period combinations (Table 4 4), an insufficient number of plots to be analyzed. However, golden topminnows tended to be captured more frequently in the later sampling periods and in nutrient enriched treatments (data not shown). Fish Community Respons e I used community dissimilarity matrices of relative abundance to determine if treatments had an effect on the fish community sampled. While p values were significant, the R statistic (indicative of the strength of between vs. within group treatment diffe rences) was close to zero, indicating that relative abundance of fish communities sampled within a treatment were similar to fish communities in other treatments (Table 4 9 ). The inclusion of crustaceans (i.e., riverine grass shrimp ( Palaemonetes paludosa) and crayfish ( Procambarus spp.)) did not change results. Upon closer inspection of effects from nutrients and light, a similar pattern of significant p values and an R statistic close to zero was revealed for nutrient effects and light effects on fish rel ative abundance.
114 Discussion Contrary to my hypothesis, the fish assemblage showed a limited response to prescribed fire in the Everglades, despite a spike in P concentrations in water and an increase in periphyton cover and biomass in the burn. Post burn T P and SRP concentrations in water spiked approximately an order of magnitude above pre burn concentrations for less than 24 hours, indicating that nutrient availability to biota is short lived after fire in the Everglades. Similarly, P concentrations in nu trient enriched, cattail dominated areas of the Everglades spiked relatively quickly and then dropped to pre burn concentrations (Miao et al. 20 10) albeit at slower rates than in the sawgrass dominated marshes sampled in the present study. Absorption by periphyton was likely the primary mechanisms for decreasing concentrations of P following the spike on day 1 post burn (Noe et al. 2001, Saiers et al. 2003) Periphyton readily uptakes P, in accordance w ith the loading rate and duration that P is available (Newman et al. 2004) In two periphyton samples I collected at the same site pre and immediately post b urn, periphyton tissue P concentrations increased by 0.027 and 0.073 mg/kg TP to 0.205 and 0.236 mg/kg, respectively. This elevation in periphyton P concentration reflected P concentration increases in water immediately post burn at the same sites (0.036 a nd 0.040 mg/L TP, respectively). Thus, fire is an important process for remobilizing P and making P readily available to biota at the base of the aquatic trophic web. Phosphorus in water can also diminish by flowing out of the burn, but this is not likely a primary mechanism by which P concentrations in water decreased within the study site. P flow post burn has been detected at least 100 m downstream of burns (Miao et al. 2010) M ost sample locations in this study were in the middle or downstream portions of the burn. Based on water flow rates that range from 0.2 7.9
115 mm/s in sloughs in central Water Conservation Area 3A of the Everglades (Harvey et al. 2009) the sites I sampled sh ould have had elevated P concentrations equivalent to the day 1 spike on day 2 or later, even under high flow rate conditions. Instead, concentrations dropped rapidly, indicating that biotic uptake reduced P concentrations in water. Sawgrass stores more TP in belowground parts of the plant that are associated with resource storage th an in leaves (Miao and Sklar 1998) Fire released pulses of P depend on the concentrations of P in the parts of the plant burned. Prescribed burns are ty pically conducted with standing water covering the belowground portion of sawgrass and only burn the aboveground portion of sawgrass. Thus, prescribed burns remobilize limited concentrations of P and are short lived due to low concentrations in sawgrass le aves TP concentrations in water in this study were much lower than TP concentrations released after cattail in the Everglades, which stores more P, was burned (Miao et al. 2010) Conversely, wildfires typically occur when water levels are below the marsh surface and often burn above and belowground portions of sawgrass, releasing much more P than prescribed burns. In an oligotrophic wetland, an y remobilization of nutrients, particularly a limiting nutrient such as P in the Everglades, can result in a boost in primary production. Increases in periphyton cover and periphyton mass per area (dw) after a fire indicate that the release of P and light post burn was sufficient to result in a significant response of periphyton. Thomas et al. (2006) saw no difference in periphyton mat composition or daily gross photosynthesis (GPP) under a similar range of light condit ions as used in this study. However, p ast studies of nutrient or light effects on
116 periphyton in the Everglades have primarily focused on thick mats of periphyton (e.g., Newman et al. 2004, Gaiser et al. 2005, Thomas et al. 2006) In other aquatic systems, light typically results in increased algal biomass (Mosisch et al. 2001, Mokany et al. 2008) Periphyton in a recently burned area with no established periphyton mat, such as plots in this study, may react differently to changes in light conditions than an established periphyton assemblage in a thick mat. Periphyton collected in sawgrass ridges in this study generally grew as a thin epipelic layer, which may be more available to herbivores than when growing within a thick, complex mat structure (Geddes and Trexler 2003, Chick et al. 2008) Periphyton mats in very oligotrophic areas of the Ev erglades in water ) tend to be composed primarily (49 83%) of cyanobacteria (McCormick and O'Dell 1996) Edible, preferred species such as diatoms grow in pockets created during cyanobacterial growth (Geddes and Trexler 2003) Nutrient enrichment can alter species com position or structure of the mat and thereby increase edibility of periphyton (Geddes and Trexler 2003, Chick et al. 2008) While I do not have species composition data to confirm edibility, the increase of periphyton in burned sawgrass stands may have provide d an additional food resource for herbivorous species where periphyton was previously limited or non existent. Total fish assemblage did not respond to prescribed f ire in the Everglades despite an increase in periphyton cover and biomass in burns. Over the duration of this experiment, fish showed a lot of variability in all metrics and no consistent trend, indicating that burns increased overall fish size or conditio n factor. While it is possible that the duration of the experiment was insufficient to capture all effects of fire on the
117 fish assemblage using a prescribed burn, we saw no indication of a lag in response. Total fish abundance increased temporarily (about three weeks) after a burn, but this did not translate into increases in overall fish size or condition factor. Fish are highly mobile organisms that can respond relatively quickly to changes in the environment (DeAng elis et al. 2010, Obaza et al. 2011) Thus, fish may have concentrated in the burn for the first three weeks and then left rather than staying in this habitat for three months or more Additionally, s pecies respond to environmental changes differently bas ed on availability of preferred food resources (e.g., Reimer 1970) and predation risk (e.g., Dorn et al. 2006) Metrics characterizing the total fish assemblage a re a composite of species with these different life history traits. The contrasting behaviors of species likely diluted responses of individual species, resulting in no overall trend. Thus, aggregating the entire assemblage may have obscured responses by i ndividual species to burned areas. Responses by individual species to burns may be more informative. As expected, individual species of fish responded differently to treatments. Least killifish, the smallest fish species captured, were 1 2 mm longer (6 11% length increase) in burned than unburned areas, a biologically significant size difference for this species. Larger female least killifish produce more broods and more juveniles per brood than smaller females (Leips and Travis 1999) Thus, the increase of nutrients in burns could incr ease reproductive output of this species via increasing female size. Conversely, smaller flagfish and sailfin mollies were captured in burns. D ifferences in size for these species, and also for least killifish, may not be due to growth (Travis et al. 1989) but rather related to habitat choice by different size classes.
118 Differences in the abundance of species such as Everglades pygmy sunfish captured by minnow and throw traps in treatment plots illustrate that interpretation of treatment effects can be biased by the capture method used. I used minnow traps to minimize sampling disturbance of the plots (i.e., increased nutrient concentrations from re suspended sediment; Rozas and Minello 1997) and to avoid obfuscating effects that may have occurred as a result of treatment. Passive sampling devices such as minnow traps are biased because they do not sample a standardized area, are selective in terms of species captured, and do not have high capture efficiency (e.g., Blaustein 1989, He and Lodge 1990, Layman and Smith 2001, Obaza et al. 2011) For some species though, minnow traps can provide an accurate index of abundance (He and Lodge 1990) However, I expect that repeated use of throw traps, in which I pulled all sawgrass (and thus underwater structure) to clear traps, would have changed the habitat and treatment effects in the plots almost immediately, resulting in samples not representative of the treatments in which samples were taken. Lof tus and Eklund (1994) illustrated differences in the fish community between the surrounding marsh and sampling locations due to long te rm use of a drop trap in the same locations, thereby lowering marsh elevations and changing fish response. While minnow traps are not ideal for measuring fish response to changes in the environment, they were the best available method given the study objec tives and constraints and provided similar responses to throw traps. Despite the limitation of minnow traps, the results of this study provide an initial understanding of how fish respond to changes related to fire burning wetlands.
119 In conclusion, the fish assemblage showed a limited response to prescribed fire in the Everglades, contrary to my hypothesis, despite a spike in P concentrations in water in the burn and an increase in periphyton cover and biomass in burns. Concentrations of nutrients available for biotic uptake were limited by concentrations in aboveground plant parts and may result in limited trophic level effects such as we saw in this study. Concentrations of P in peat are much higher than in sawgrass (Noe et al. 2001) representing a significant source of P in the Everglades. Wildfires typically occur when water levels are below the marsh surface and frequently burn the entire sawgrass plant and the peat (Wade et al. 1980) Due to higher concentrations of P stored in sawgrass and peat below the marsh surface, I would expect a stronger response by periphyton and fish to higher P concentrations available after wildfires that burn peat, however this has not been quantified and should be studied. Similarly, effects of fires in other oligotrophic wetlands are likely influenced by concentrations of the limiting nutrient released during fires.
120 Table 4 1. Mean ( standard deviation) of environmental variables measured in plots. Experimental No Nutrients Light Nutrients Nutrients Variable Control or Light Only Only + Light Water Depth (cm)* 17.1 (3.6) bc 17.3 (3.1) c 19.3 (3.8) c 15.2 (3. 1) ab 17.4 (3.9) bc Shading (%)* 63 (27) b 89 (4) a 36 (26) c 87 (6) a 50 (18) bc Median Vegetation Height (cm)* 165 a 71 c 83 b 69.5 c 89 b Vegetation Height range (cm) 142 215 37 103 27 122 39 98 32 120 Median Max. Vegetation Height (cm)* 200 a 101 cd 114 b 94.5 d 104 b c Max. Vegetation Height range (cm) 161 350 51 150 56 177 56 175 43 157 Vegetation Cover (%) 40.7 (28) 27.8 (19) 29.0 (22) 26.3 (9.4) 25.3 (18) Periphyton Cover (%)* 21.4 (24) b 8.1 (14) c 13.5 (14) b 11.3 (14) c 26.6 (23) a Periphyton Mass per Area (g dw/m 2 )* 41.5 (65) b 2.8 (5.4) c 12.5 (26) b 8.6 (14) b 50.5 (83) a Average Chlorophyll a (g/g dw) 558 (630) 771 (543) 687 (387) 1084 (920) 689 (366) Chlorophyll a (g/g dw) 0 2134 0 2486 0 1474 0 3590 0 1400 Average Fish Standard Length (mm) 25.4 (4.4) 25.5 (2.5) 25 .8 (2.6) 26.5 (2.8) 26.2 (2.8) Average Fish Mass (g) 0.54 (0.36) 0.52 (0.22) 0.56 (0.23) 0.59 (0.26) 0.56 (0.24) Average Fish Condition Factor 2.18 (0.20) 2.24 (0.21) 2.32 (0.27) 2.22 (0.19) 2.25 (0.22) Average Fish Abundance 48.7 ab 37.3 b 55.4 ab 50.2 ab 63. 8 a Average Relative Abundance 20.4 20.6 17.4 18.6 18.0 Richness 4.9 (1.5) 4.8 (1.2) 5.7 (1.2) 5.4 (1.2) 5.5 (1.2) Shannon Diversity* 1.09 bc 1.03 c 1.27 a 1.03 c 1.17 ab Significant difference among treatments, differences denoted by letters. Summary of stati stical results provided on Table 4 2 and Table 4 3.
121 Table 4 2. Summary of ANCOVAs testing differences due to treatment and period. Treatment Period Interaction Residuals Variable df F p df F p df F p Water Depth (cm) 4 5.48 <0.001 1 1.24 0.267 4 1.59 0.180 148 Shading (%) 4 12.0 <0.001 1 2.46 0.128 4 0.257 0.903 29
122 Table 4 3. Summary of response s of biotic variables to treatments (Kruskal Wallis rank sum test) Treatment Variable df 2 p Periphyton Periphyton Cover (%) 4 16.7 0.002* Periphyton Mass per Area (g dw/m 2 ) 4 18.7 0.001* Chlorophyll a (corrected) (g/g dw) 4 7.04 0.134 Vegetation Vegetation Cover (%) 4 6.12 0.190 Vegetation Height (cm) 4 83.6 <0.001* Max. Vegetation Height (cm) 4 76.5 <0.001* Fish Richness 4 9.30 0.054 Diversity 4 11.6 0.021* Relative Abundance 4 5.95 0.203
123 Table 4 4. Frequency of capture of aquatic organisms in minnow traps by treatment plot and species in the Everglades, 2010. Experimental No Nutrients Light Nutrients Nutrients Species Control a or Light Only Only + Light Fish Elassoma evergladei (Everglades pygmy sunfish) 0 0 0 1 0 Esox americanus (redfin pickerel) 0 0 2 0 0 Fundulus chrysotus (golden topminnow) 18 11 16 25 24 F. conflu entus (marsh killifish) 21 23 27 34 25 Gambusia holbrooki (eastern mosquitofish) 32 32 32 31 32 Heterandria formosa (least killifish) 28 29 31 19 23 Jordanella floridae (flagfish) 23 28 31 31 31 Lepomis punctatus (spotted sunfish) 2 2 1 3 3 Lucania go odei (bluefin killifish) 8 4 16 9 16 Poecilia latipinna (sailfin molly) 25 26 28 25 24 Crustaceans Palaemonetes paludosa (riverine grass shrimp) 18 19 23 14 21 Procambarus spp. (crayfish) 10 18 13 9 5 Herpetofauna Nerodia fasciata (Florida water snake ) 2 1 0 1 1 Notophthalamus viridescens piaropicola (peninsula newt) 0 0 1 0 0 Siren lacertina (greater siren) 1 1 0 0 1 Rana grylio (pig frog) 3 3 3 0 3 Macroinvertebrates Belastomatid (giant water bug) 1 0 0 2 1 Dytiscidae (predaceous diving beetle) 6 3 1 5 9 a Number of plot and period combinations in which each species was captured at least once. Total possible plot and period combi nations per treatment is 32.
124 Table 4 5. Summary of generalized least squares regression examining response of fish me asures to treatment and sampling period. Treatment Period Interaction Treatment Contrast a Variable b Model c F p F p F p All Fish Standard Length d arh1 1.21 0.310 2.02 0.058 0.910 0.600 -Mass e (log) arh1 0.634 0.640 1.45 0.193 0.789 0.762 -Condition Factor arh1 0.420 0.794 4.76 <0.001* 1.73 0.023 -Abundance (4 th rt) arh1 1.91 0.113 2.49 0.020* 0.923 0.581 -G. holbrooki Standard Length arh1 1.36 0.254 2.44 0.023* 1.10 0.348 -Mass (sqrt) ar1 1.31 0.271 1.16 0.333 1.00 0.473 -Con dition Factor arh1 1.03 0.394 8.58 <0.001* 1.05 0.410 -Abundance (sqrt) arh1 1.17 0.328 3.37 0.003* 0.865 0.661 -Relative Abundance ar1 0.978 0.423 2.52 0.019* 0.818 0.725 -J. floridae Standard Length arh1 4.69 0.002* 2.69 0.014* 0.737 0.822 cont rol, nutrients Mass arh1 6.41 <0.001* 2.22 0.039* 0.850 0.681 control, nutrients Condition Factor arh1 1.18 0.323 2.45 0.028* 0.734 0.825 -Abundance arh1 4.17 0.003* 0.864 0.537 1.47 0.079 light Relative Abundance (sqrt) arh1 4.74 0.001* 1.52 0.167 1.23 0.223 control P. latipinna Standard Length ar1 2.35 0.061 3.06 0.006* 0.819 0.720 -Mass ar1 3.43 0.012* 1.95 0.071 1.01 0.469 nutrients Condition Factor arh1 0.729 0.574 1.79 0.100 1.69 0.034* -Abundance arh1 1.98 0.102 3.11 0.005* 0.799 0 .749 -Relative Abundance ar1 2.31 0.062 4.32 <0.001* 1.26 0.196 -
125 Table 4 5. Continued Treatment Period Interaction Treatment Contrast a Variable b Model c F p F p F p H. formosa Standard Length arh1 6.83 <0.001* 1.91 0.077 1.78 0.022* nut rients, light Mass ar1 1.67 0.165 1.96 0.069 1.81 0.019* -Condition Factor arh1 2.90 0.026* 2.44 0.024* 1.45 0.096 nutrients Abundance arh1 2.32 0.061 4.13 <0.001* 1.10 0.347 -Relative Abundance (sqrt) arh1 4.50 0.002* 3.53 0.002* 1.30 0.168 contr ol, nutrients a Treatment contrasts refer to differences seen among treatments. control = Experimental Control vs. other treatments, nutrient s = nutrients vs. no nutrients, light = light vs. no light b Transformation of dependent variable given in parenthese s. If there is nothing in parentheses, variable was not transformed. sqrt = square root, 4 th rt = fourth root, log = log c arh1 = autoregressive with heterogeneous variances, ar1 = autoregressive; degrees of freedom are treatment = 4, period = 7, a nd intera ction = 28 d Standard length is in millimeters e Mass is in grams
126 Table 4 6. Mean ( standard deviation) of fish captured in 1 m 2 throw traps. No Nutrients Light Nutrients Nutrients Variable or Light Only Only + Light Density* 17.1 (4.8) 36.0 (17.1) 23.4 (9.6) 29.3 (9.2) Standard Length (mm) All fish 16.2 (7.5) 15.9 (6.7) 17.9 (10.0) 14.3 (4.0) E. evergladei 14.0 (1.0) 16.5 (2.0) 11.1 (1.2) 14.7 (4.3) F. chrysotus 21.5 (15.6) 10.5 43.2 12.8 (2.2) F. confluentus 14.7 (2.5) 10.7 (4.4) 15.5 (2.3) 21.5 (3.0) G. holbrooki 13.7 (2.7) 11.9 (3.4) 12.5 (3.1) 13.4 (2.8) H. formosa 11.0 (0.8) 11.4 (0.9) 11.8 (1.4) 11.8 (0.9) J. floridae 20.9 (8.2) 28.3 (5.7) 22.7 (8.6) 16.2 (5.5) L. goodei -14.8 (6.2) 16.2 (1.5) 14.1 (3.8) L. punctatus -27.0 47.8 (9.2 ) -P. latipinna 22.3 (9.3) 20.2 (6.5) 19.1 (3.0) 12.8 (1.4) Significant difference among treatments. Statistical summary provided on Table 4 7.
127 Table 4 7. Summary of Analysis of Variances (ANOVA) examining response of all and individual fish species captured in throw traps to light ( L ight vs. No L ight) and nutrient treatments ( N utrients vs. N o Nutrients). Light Nutrients Interaction df a Variable F p F p F p Density 4.98 0.046* 0.002 0.965 1.37 0.264 1,1,1,12 Standard Leng th (mm) b E. evergladei 6.33 0.024* 4.64 0.048* 0.170 0.686 1,1,1,15 F. confluentus 0.152 0.706 14.6 0.004* 5.42 0.045* 1,1,1,9 G. holbrooki 0.189 0.667 0.106 0.747 2.17 0.152 1,1,1,27 H. formosa 0.249 0.622 3.24 0.083 0.203 0.656 1,1,1,28 J. floridae 0.309 0.585 2.99 0.100 4.99 0.038* 1,1,1,19 L. goodei 0.733 0.417 0.008 0.933 NA 1,1,1,8 P. latipinna (log) c 3.37 0.081 3.78 0.065 1.68 0.209 1,1,1,21 a Degrees of freedom for light, nutrients, interaction, and residuals, respectively. b F chrysotus were not normally distributed and were analyzed for differences between light and between nutrient treatments using a Kruskal Wallis rank sum test. P values were > 0.18 and are not included on this table. c P. latipinna standard lengths were log transformed to meet assumptions of normality.
128 Table 4 8. Characteristics of fish species caught in at least 80% of plots sampled; mean ( standard deviation). Experimental No Nutrients Light Nutrients Nutrients Variable Control or Light Only Only + Light G. holbrooki Standard Length (mm) 22.6 (1.6) 23.4 (1.3) 23.3 (2.2) 23.5 (1.1) 23.2 (1.4) Mass (g) 0.26 (0.06) 0.28 (0.05) 0.28 (0.05) 0.29 (0.05) 0.28 (0.06) Condition Factor 1.97 (0.26) 1.98 (0.16) 2.12 (0.73) 1.98 (0.18) 1.98 (0.28) Abundance 2 8.9 (27.4) 23.0 (11.1) 30.6 (22.8) 34.5 (27.1) 36.5 (23.4) Relative Abundance 56.9 (18.7) 63.0 (17.5) 52.6 (15.3) 62.9 (20.2) 56.5 (15.2) J. floridae Standard Length* 25.7 (2.4) 27.9 (2.3) 29.0 (2.8) 26.2 (3.6) 26.9 (2.3) Mass* 0.61 (0.19) 0.77 (0.19) 0.90 (0.27) 0.63 (0.23) 0.67 (0.18) Condition Factor 3.4 (0.36) 3.4 (0.44) 3.4 (0.48) 3.3 (0.32) 3.3 (0.33) Abundance* 5.1 (7.8) 4.2 (4.0) 7.3 (6.7) 5.2 (4.9) 11.7 (12.8) Relative Abundance* 7.2 (6.8) 10.0 (8.0) 13.2 (7.2) 11.5 (11.3) 16.4 (11.8) P. la tipinna Standard Length 24.5 (3.8) 25.9 (6.3) 27.8 (3.8) 24.3 (4.0) 24.7 (3.7) Mass* 0.43 (0.16) 0.55 (0.33) 0.62 (0.23) 0.44 (0.18) 0.41 (0.18) Condition Factor 2.51 (0.37) 2.55 (0.58) 2.51 (0.33) 2.65 (0.43) 2.47 (0.40) Abundance 5.7 (5.6) 3.8 (4.5) 6.6 (6.1) 4.3 (4.2) 3.4 (4.9) Relative Abundance 11.8 (11.9) 9.4 (9.1) 11.8 (9.4) 9.7 (10.5) 5.4 (7.3) H. formosa Standard Length* 17.4 (1.5) 17.9 (1.4) 17.2 (1.8) 19.0 (1.8) 18.5 (1.9) Mass 0.13 (0.04) 0.15 (0.05) 0.13 (0.05) 0.16 (0.06) 0.15 (0.05) Condition Factor* 2.42 (0.54) 2.53 (0.48) 2.51 (0.78) 2.30 (0.38) 2.20 (0.45) Abundance* 3.7 (3.0) 3.0 (3.0) 4.8 (6.2) 1.8 (2.4) 3.4 (4.2) Relative Abundance* 11.2 (9.3) 8.8 (8.1) 9.5 (9.1) 3.5 (4.2) 5.5 (6.7) Letters indicate significant difference of means between treatments grouped by factor (e.g., light vs. no light).
129 Table 4 9. Summary of ANOSIM (Analysis of Similarities) results testing differences of relative abundance. The R statistic ranges from 1 to 1 with 0 indicating random grouping of replicates in groups and 1 indicating replicates within a site are similar compared to replicates from other sites. A p value <0.05 is used to indicate significance (*) Fish Only Fish and Crustaceans Comparison R statistic p value R statistic p value A ll Treatments 0.063 0.001 0.064 0.001 Exp. Ctrl vs. Treatment 0.108 0.026 0.109 0.030 Period 0.055 0.001 0.055 0.001 Nutrients 0.072 0.001 0.072 0.003 Light 0.010 0.261 0.011 0.274
130 Figure 4 1. Experimental design showing burned (Nutrient) treat ments (top row) vs. unburned (No Nutrient) treatments (second row) and Light treatments (first column) vs. No Light treatments (second column). Experimental Control treatment is the unmanipulated version of the No Nutrients or Light treatment.
131 Figure 4 2. Co ncentrations of total phosphorus (TP) and soluble reactive phosphorus (SRP) in water sampled collected pre burn (day 0) and post burn (days 0.5 15) in burned plots (Nutrients), mowed with vegetation removed (Light Only), and mowed with vegetation rem oved and a shade house constructed (No NL = No Nutrients or Light) in northern Water Conservation Area 3A South of the Everglades, Florida, USA. *B = concentration in burn on that day is significantly different than pre burn phosphorus concentration
132 Fig ure 4 3. Linear relationship of minnow trap catch per unit effort (CPUE) and throw trap density sampled during the final sampling period (period 8) in treatment plots (n=16). Experimental control plots were not sampled with throw traps. Adjusted R 2 value p rovided on the figure.
133 Figure 4 4. Standard length (mm), mass (g), condition factor, and abundance of Flagfish ( Jordanella floridae ) captured in minnow traps in plots post burn in northern Water Conservation Area 3A South of the Everglades, Florida, US A. N+L = Nutrients + Light, N = Nutrients Only, Control = Experimental Control, Light = Light Only, No NL = No Nutrients or Light
134 Figure 4 5. Standard length (mm), mass (g), condition factor, and abundance of Sailfin Mollies ( Poecilia latipinna ) captur ed in minnow traps in plots post burn in northern Water Conservation Area 3A South of the Everglades, Florida, USA. N+L = Nutrients + Light, N = Nutrients Only, Control = Experimental Control, Light = Light Only, No NL = No Nutrients or Light
135 Figure 4 6. Standard length (mm), mass (g), condition factor, and abundance of Least Killifish ( Heterandria formosa ) captured in minnow traps in plots post burn in northern Water Conservation Area 3A South of the Everglades, Florida, USA. N+L = Nutrients + Light, N = Nutrients Only, Control = Experimental Control, Light = Light Only, No NL = No Nutrients or Light
136 Figure 4 7. Standard length (mm), mass (g), condition factor, and abundance of Eastern Mosquitofish ( Gambusia holbrooki ) captured in minnow traps in pl ots post burn in northern Water Conservation Area 3A South of the Everglades, Florida, USA. N+L = Nutrients + Light, N = Nutrients Only, Control = Experimental Control, Light = Light Only, No NL = No Nutrients or Light
137 CHAPTER 5 CONCLUSIONS Fire is a nat ural process in the Everglades, important for recycling nutrients and maintaining vegetative communities. While wildfires typically occur at the onset of the wet season (Slocum et al. 2007) prescribed fires are frequently conducted to reduce fuel loads and manage habitat for wildlife (Marsha Ward, FWC, pers. com.). Frequent fires remove tall, dense stands of sawgrass, op ening areas of previously inaccessible, shallow water marsh to foraging wading birds. Prescribed burns are conducted during the dry season when water levels are declining, limiting the length of time these shallow burned areas are available as foraging hab itat for wading birds. Areas of shallow water are preferred habitat for wading birds given conditions of similar prey densities (Gawlik 2002) Wading birds preferred burned areas for the first 2 3 w eeks post burn (Chapter 3). Great egrets had higher capture efficiency in these burned sawgrass ridges than in the surrounding sloughs, but had a higher capture rate in sloughs than in burns because they made more strikes. Over multiple weeks post burn, pr ey densities do not appear to be greater in burned areas than the adjacent sloughs, suggesting that wading bird preference of burned areas is based on water depth and prey accessibility. Fish response to burns was limited, despite an increase in P and peri phyton biomass. Fish abundance in burns appeared to increase temporarily in response to light and nutrients increased by the burn (Chapter 3). Additionally, select individual fish species increased in size in burns and may increase reproductive output, and thus abundance, of this species. However, sampling of burned ridges indicate that prey densities are lower on recently burned ridges than in sloughs (Chapter 4). Overall, the whole fish community did not increase in size, but did briefly increase in abund ance.
138 From the perspective of a wading bird, changes in the whole fish community are likely a better representation of composite diet that wading birds eat rather than changes of individual species. Thus, prescribed burns do not appear to enhance the calor ic intake of wading birds foraging in burns. These studies add to the limited body of knowledge about fire effects on wetland dependent wildlife (Chapter 2), expanding our understanding of how fire impacts foraging opportunities and resources for wading bi rds in the Everglades. The response by wading birds to fire is likely to occur in other wetlands when shallow water areas of marsh are exposed for foraging after a burn. Fire in other wetlands can be expected to release nutrients although the effects of bi oavailable nutrients are dependent on the concentration of bioavailable nutrients released and the concentration already available in the wetland. I expect a stronger response by primary producers to nutrient release to occur in oligotrophic wetlands than in nutrient enriched wetlands. However, the primary result of this research is that prey availability, rather than prey biomass, appears to drive the preference foraging wading birds exhibit for burned areas. Changes in season and severity of the fire will alter these responses and should be explored.
139 APPENDIX A WHITE IBIS ( EUDOCIMUS ALBUS ) AND SNOWY EGRET ( EGRETTA THULA ) CAPTURE EFFICIENCI ES AND CAPTURE RATES Table A 1. Summary of capture rates and capture efficiencies reported for white ibis ( Eudocimus albus ) and snowy egret ( Egretta thula ) in southern Florida marshes. Capture Capture Study Condition Rate (N) Efficiency (N) Location White ibis Surdick (1998) 1996 1.4 (151) Everglades 1997 0.6 (219) Lantz et al. (2010) January 0.79 1.02 (71) S AV a density experiment April 1.78 2.24 (135) This study 2009 1.6 (43) 0.03 (43) Everglades WCA 3A b 2010 0.74 (18) 0.01 (18) Snowy egret Surdick (1998) 1996 1.02 (213) Everglades 1997 0.6 (206) Lantz et al. (2010) January 0.28 0.53 (35) 0.14 0.30 (33) SAV density experiment April 0.90 1.49 (124) 0.23 0.30 (123) Lantz et al. (2011) 0.78 1.35 (92) 0.20 0.41 (89) Emergent vegetation experiment This study 2009 1.5 (13) 0.4 (13) Everglades WCA 3A 2010 0.12 (5) 0.07 (5) a Submerged aquatic veget ation b Water Conservation Area 3A
140 Table A 2. Capture rate (captures per minute) and capture efficiency (captures per attempt) of white ibis ( Eudocimus albus ) in 2009 and 2010 in Water Conservation Area 3A of the Everglades, USA. 2009 2010 Variable B urn Slough adj. Burn a Burn Slough adj. Burn Number of observations 5 38 8 10 Mean capture rate ( sd) 1.2 (1.3) 1.7 (1.2) 0.67 (0.4) 0.79 (0.6) Range of capture rate 0 3.2 0 6.8 0 1.2 0 2.2 Mean capture efficiency ( sd) 0.03 (0.03) 0.03 (0.02) 0.01 (0.008 ) 0.01 (0.01) Range of capture efficiency 0 0.08 0 0.14 0 0.02 0 0.05 Average attempts per minute 38.7 (9.2) 61.7 (13.5) 61.0 (8.9) 55.0 (10.2) Water depth (cm) 3.0 (3.7) 14.7 (4.1) 8.8 (1.7) 20.9 (3.5) Range of water depth (cm) 0 9 3 25 7 11 17 27 a Slough s adjacent to burns
141 LIST OF REFERENCES Almario BS, Marra PP, Gates JE, Mitchell L (2009) Effects of prescribed fire on depredation rates of natural and artificial seaside sparrow nests. The Wilson Journal of Ornithology 121:770 777 A ustin JE, Buhl DA (2011) Nest survival of American coots relative to grazing, burning, and water depths. Avian Conservation and Ecology 6:1 14 Babbitt LH, Babbitt CH (1951) A herpetological study of burned over areas in Dade County, Florida. Copeia 1951:79 Baker JL (1974) Preliminary studies of the dusky seaside sparrow on the St. Johns National Wildlife Refuge. Proceedings of the Annual Conference, Southeastern Association of Game and Fish Commissioners 27:207 214 Baldwin HQ, Grace JB, Barrow J, Wylie C., Rohwer FC (2007) Habitat relationships of birds overwintering in a managed coastal prairie. The Wilson Journal of Ornithology 119:189 197 Bancroft GT, Gawlik DE, Rutchey K (2002) Distribution of wading birds relative to vegetation and water depths in the n orthern Everglades of Florida, USA. Waterbirds 25:265 277 Beganyi SR, Batzer DP (2011) Wildfire induced changes in aquatic invertebrate communities and mercury bioaccumulation in the Okefenokee Swamp. Hydrobiologia 669:237 247 Benson TJ, Dinsmore JJ, Hohma n WL (2007) Responses of plants and arthropods to burning and disking of riparian habitats. Journal of Wildlife Management 71:1949 1957 Bishop DC, Haas CA (2005) Burning trends and potential negative effects of suppressing wetland fires on flatwoods salama nders. Natural Areas Journal 25:290 294 Blaustein L (1989) Effects of various factors on the efficiency of minnow traps to sample mosquitofish ( Gambusia affinis ) and green sunfish ( Lepomis cyanellus ) populations. Journal of the American Mosquito Control As sociation 5:29 35 Bowland JM, Perrin MR (1993) Wetlands as reservoirs of small mammal populations in the Natal Drakensberg. South African Journal of Wildlife Research 23:39 43 Bray MP (1984) An evaluation of heron and egret marsh nesting habitat and possib le effects of burning. The Murrelet 65:57 59
142 Brennan EK, Smith LM, Haukos DA, LaGrange TG (2005) Short term response of wetland birds to prescribed burning in Rainwater Basin wetlands. Wetlands 25:667 674 Browder JA (1981) Perspective on the ecological cau ses and effects of algal composition of southern Everglades periphyton, South Florida Research Center, Homestead. Browder JA, Gleason PJ, Swift DR (1994) Periphyton in the Everglades: Spatial variation, environmental correlates, and ecological implication s. p. 379 418. In Davis SM, and Ogden JC (eds), Everglades: the ecosystem and its restoration. St. Lucie Press Delray Beach Burnham KP, Anderson DR (2002) Model selection and multimodel inference: a practical information theoretic approach 2nd edition. Spr inger Science+Business Media, Inc., New York Cartwright BW (1942) Regulated burning as a marsh management technique. p. 257 263. Transactions of the 7th North American Wildlife Conference. Chick JH, Geddes P, Trexler JC (2008) Periphyton mat structure med iates trophic interactions in a subtropical marsh. Wetlands 28:378 389 Clarke KR (1993) Non parametric multivariate analyses of changes in community structure. Australian Journal of Ecology 18:117 143 Conway CJ, Nadeau CP, Piest L (2010) Fire helps restore natural disturbance regime to benefit rare and endangered marsh birds endemic to the Colorado River. Ecological Applications 20:2024 2035 Cox J, Widener B (2008) Lightning season burning: friend or foe of breeding birds? Tall Timbers Research Station, Tal lahassee Curnutt JL, Mayer AL, Brooks TM, Manne L, Bass J, Oron L., Fleming DM, Pimm SL (1998) Population dynamics of the endangered Cape Sable seaside sparrow. Animal Conservation 1:11 21 de Szalay F, Resh V (1997) Responses of wetland invertebrates and p lants important in waterfowl diets to burning and mowing of emergent vegetation. Wetlands 17:149 156 DeAngelis DL, Trexler JC, Cosner C, Obaza A, Jopp F (2010) Fish population dynamics in a seasonally varying wetland. Ecological Modelling 221:1131 1137 DeB ano LF, Neary DG, Ffolliott PF (1998) Fire: its effect on soil and other ecosystem resources. John Wiley & Sons, Inc., Hoboken
143 Devereux CL, Whittingham MJ, Krebs JR, Fernandez Juricic E, Vickery JA (2006) What attracts birds to newly mown pasture? Decoupli ng the action of mowing from the provision of short swards. Ibis 148:302 306 Dorn N, Trexler J, Gaiser E (2006) Exploring the role of large predators in marsh food webs: evidence for a behaviorally mediated trophic cascade. Hydrobiologia 569:375 386 Enge K M (2005) Herpetofaunal drift fence surveys of steephead ravines in the Florida Panhandle. Southeastern Naturalist 4:657 678 Epanchin PN, Heath JA, Frederick PC (2002) Effects of fires on foraging and breeding wading birds in the Everglades. Wilson Bulletin 114:139 141 Gabrey SW, Afton AD (2000) Effects of winter marsh burning on abundance and nesting activity of Louisiana seaside sparrows in the Gulf Coast Chenier Plain. Wilson Bulletin 112:365 372 Gabrey SW, Afton AD (2001) Plant community composition and biomass in Gulf Coast Chenier Plain marshes: Responses to winter burning and structural marsh management. Environmental Management 27:281 293 Gabrey SW, Afton AD (2004) Composition of breeding bird communities in Gulf Coast Chenier Plain marshes: Effects o f winter burning. Southeastern Naturalist 3:173 185 Gabrey SW, Afton AD, Wilson BC (1999) Effects of winter burning and structural marsh management on vegetation and winter bird abundance in the Gulf Coast Chenier Plain, USA. Wetlands 19:594 606 Gabrey SW, Afton AD, Wilson BC (2001) Effects of Structural Marsh Management and winter burning on plant and bird communities during summer in the Gulf Coast Chenier Plain. Wildlife Society Bulletin 29:218 231 Gabrey SW, Wilson BC, Afton AD (2002) Success of artific ial bird nests in burned Gulf Coast Chenier Plain Marshes. Southwestern Naturalist 47:532 538 Gagnon PR (2009) Fire in floodplain forests in the Southeastern USA: Insights from disturbance ecology of native bamboo. Wetlands 29:520 526 Gaiser EE, Childers D L, Jones RD, Richards JH, Scinto LJ, Trexler JC (2006) Periphyton responses to eutrophication in the Florida Everglades: Cross system patterns of structural and compositional change. Limnology and Oceanography 51:617 630 Gaiser EE, McCormick PV, Hagerthey SE, Gottlieb AD (2011) Landscape patterns of periphyton in the Florida Everglades. Critical Reviews in Environmental Science and Technology 41:92 120
144 Gaiser EE, Scinto LJ, Richards JH, Jayachandran K, Childers DL, Trexler JC, Jones RD (2004) Phosphorus in periphyton mats provides the best metric for detecting low level P enrichment in an oligotrophic wetland. Water Research 38:507 516 Gaiser EE, Trexler JC, Richards JH, Childers DL, Lee D, Edwards AL, Scinto LJ, Jayachandran K, Noe GB, Jones RD (2005) Casca ding ecological effects of low level phosphorus enrichment in the Florida Everglades. J Environ Qual. 34:717 723 Gauthier G, Bty J, Giroux J F, Rochefort L (2004) Trophic interactions in a high arctic snow goose colony. Integrative and Comparative Biology 44:119 129 Gawlik DE (2002) The effects of prey availability on the numerical response of wading birds. Ecological Monographs 72:329 346 Geddes P, Trexler JC (2003) Uncoupling of omnivore mediated positive and negative effects on periphyton mats. Oecologi a 136:585 595 Giles LW, Marshall DB (1954) A large heron and egret colony on the Stillwater Wildlife Management Area, Nevada. The Auk 71:322 325 Givens LS (1962) Use of fire on southeastern wildlife refuges. p. 121 126. Proceedings of the 1st Annual Tall T imbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee Gorman TA, Haas CA (2011) Seasonal microhabitat selection and use of syntopic populations of Lithobates okaloosae and Lithobates clamitans clamitans Journal of Herpetology 45:313 3 18 Gresswell RE (1999) Fire and aquatic ecosystems in forested biomes of North America. Transactions of the American Fisheries Society 128:193 221 Grimshaw HJ, Wetzel RG, Brandenburg M, Segerblom K, Wenkert LJ, Marsh GA, Charnetzky W, Haky JE (1997) Shadin g of periphyton communities by wetland emergent macrophytes: Decoupling of algal photosynthesis from microbial nutrient retention. Archiv Fur Hydrobiologie 139:17 27 Gunderson LH (1994) Vegetation of the Everglades: Determinants of community composition. p 323 340. In Davis SM, and Ogden JC (eds), Everglades: the ecosystem and its restoration. St. Lucie Press Delray Beach Gunderson LH, Snyder JR (1994) Fire patterns in the southern Everglades. p. 291 305. In Davis SM, and Ogden JC (eds), Everglades: the ec osystem and its restoration. St. Lucie Press Delray Beach Guscio CG, Hossack BR, Eby LA, Corn PS (2008) Post breeding habitat use by adult boreal toads (Bufo boreas) after wildfire in Glacier National Park, USA. Herpetological Conservation and Biology 3:55 62
145 Hanowski JM, Christian DP, Nelson MC (1999) Response of breeding birds to shearing and burning in wetland brush ecosystems. Wetlands 19:584 593 Harvey JW, Schaffranek RW, Noe GB, Larsen LG, Nowacki DJ, O'Connor BL (2009) Hydroecological factors governi ng surface water flow on a low gradient floodplain. Water Resources Research 45:1 20 Haszard S, Clark RG (2007) Wetland use by white winged scoters ( Melanitta fusca ) in the Mackenzie Delta Region. Wetlands 27:855 863 He X, Lodge DM (1990) Using minnow trap s to estimate fish population size: the importance of spatial distribution and relative species abundance. Hydrobiologia 190:9 14 Hillebrand H (2005) Light regime and consumer control of autotrophic biomass. Journal of Ecology 93:758 769 Hirano T, Kimijima M, Kobori M (2003) The effects of wildfire on the habitat use of wintering marsh harriers at Watarase marsh. Strix 21:71 79 Hochbaum GS, Kummen LT, Caswell FD (1985) Effects of agricultural burning on occupancy rates of small wetlands by breeding ducks. C anadian Wildlife Service Progress Notes 155:1 3 Hochkirch A, Adorf F (2007) Effects of prescribed burning and wildfires on Orthoptera in Central European peat bogs. Environmental Conservation 34:225 235 Hoffpauir CM (1961) Methods of measuring and determin ing the effects of marsh fires. p. 142 161. Proceedings of the Annual Conference of the Southeastern Association of Game and Fish Commission. Hood GA, Bayley SE (2003) Fire and beaver in the boreal forest grassland transition of western Canada A case st udy from Elk Island National Park, Canada. Lutra 46:235 241 Hood GA, Bayley SE, Olson W (2007) Effects of prescribed fire on habitat of beaver ( Castor canadensis ) in Elk Island National Park, Canada. Forest Ecology and Management 239:200 209 Hossack BR, Co rn PS (2007) Responses of pond breeding amphibians to wildfire: Short term patterns in occupancy and colonization. Ecological Applications 17:1403 1410 Hossack BR, Corn PS (2008) Wildfire effects on water temperature and selection of breeding sites by the boreal toad ( Bufo boreas ) in seasonal wetlands. Herpetological Conservation and Biology 3:46 54
146 Hossack BR, Eby LA, Guscio CG, Corn PS (2009) Thermal characteristics of amphibian microhabitats in a fire disturbed landscape. Forest Ecology and Management 25 8:1414 1421 Isacch JP, Holz S, Ricci L, Martnez MM (2004) Post fire vegetation change and bird use of a salt marsh in coastal Argentina. Wetlands 24:235 243 Jordan F, Coyne S, Trexler JC (1997a) Sampling fishes in vegetated habitats: Effects of habitat st ructure on sampling characteristics of the 1 m throw trap. Transactions of the American Fisheries Society 126:1012 1020 Jordan F, Jelks HL, Kitchens WM (1997b) Habitat structure and plant community composition in a northern Everglades wetland landscape. W etlands 17:275 283 Jordan J, Carroll Frank (1996) Spatial ecology of decapods and fishes in a northern Everglades wetland mosaic. PhD Dissertation, University of Florida, Gainesville Kantrud HA, Krapu GL, Swanson GA (1989) Prairie basin wetlands of the Dak otas: A community profile. p. 121. Report 85(7.28). United States Fish and Wildlife Service Biological Sciences, Washington Kern RA, Shriver WG, Bowman JL, Mitchell LR, Bounds DL (2012) Seaside sparrow reproductive success in relation to prescribed fire. J ournal of Wildlife Management 76:932 939 Kirby RE, Lewis SJ, Sexson TN (1988) Fire i n North American wetland ecosystems and fire wildlife relations: An annotated bibliography. p. 146. US Fish and Wildlife Service Komarek EV, Sr. (1969) Fire and animal beha vior. p. 160 207. Proceedings of the 9th Annual Tall Timbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee Krebs JR (1974) Colonial nesting and social feeding as strategies for exploiting food resources in the great blue heron ( Ardea herodias ). Behaviour 51:99 134 La Puma DA, Lockwood JL, Davis MJ (2007) Endangered species management requires a new look at the benefit of fire: The Cape Sable seaside sparrow in the Everglades ecosystem. Biological Conservation 136:398 407 Laderman AD (1 989) The ecology of the Atlantic white cedar wetlands: a community profile. p. 114. U.S. Fish and Wildlife Service National Wetlands Research Center, Washington Lantz SM, Gawlik DE, Cook MI (2010) The effects of water depth and submerged aquatic vegetation on the selection of foraging habitat and foraging success of wading birds. The Condor 112:460 469
147 Lantz SM, Gawlik DE, Cook MI (2011) The effects of water depth and emergent vegetation on foraging success and habitat selection of wading birds in the Everg lades. Waterbirds 34:439 447 Layman CA, Smith DE (2001) Sampling bias of minnow traps in shallow aquatic habitats on the Eastern Shore of Virginia. Wetlands 21:145 154 Legare M, Hill H, Farinetti R, Cole FT (1998) Marsh bird response during two prescribed fires at the St. Johns National Wildlife Refuge, Brevard County, Florida. p. 114. In Pruden TL, and Brennan LA (eds.), Fire in ecosystem management: shifting the paradigm from suppression to prescription. Tall Timbers Fire Ecology Conference Proceedings, N o. 20, Tall Timbers Research Station, Tallahassee Leips J, Trav is J (1999) The comparative exp ression of l ife h istory t raits and i ts r elationship to the n umerical d ynamics of f our p opulations of the l east k illifish. Journal of Animal Ecology 68:595 616 Lis ton SE, Newman S, Trexler JC (2008) Macroinvertebrate community response to eutrophication in an oligotrophic wetland: An in situ mesocosm experiment. Wetlands 28:686 694 Loftus WF, Eklund A M (1994) Long term dynamics of an Everglades small fish assemblag e. p. 461 483. In Davis SM, and Ogden JC (eds), Everglades: the ecosystem and its restoration. St. Lucie Press Delray Beach Loveless CM (1959) A study of the vegetation in the Florida Everglades. Ecology 40:2 9 Loveless CM, Ligas FJ (1959) Range conditions life history, and food habits of the Everglades deer herd. p. 201 215. Transactions of the 24th North American Wildlife Conference. Lugo AE (1995) Fire and wetland management. p. 1 9. In Cerulean SI, and Engstrom RT (eds.), Fire in wetlands: a managemen t perspective. Proceedings of the Tall Timbers Fire Ecology Conference, No. 19. Tall Timbers Research Station, Tallahassee Lynch JJ (1941) The place of burning in management of the Gulf Coast wildlife refuges. Journal of Wildlife Management 5:454 457 Lynch JJ, O'Neil T, Lay DW (1947) Management significance of damage by geese and muskrats to Gulf Coast marshes. The Journal of Wildlife Management 11:50 76 Lyon LJ, Hooper RG, Telfer ES, Schreiner DS (2000) Fire effects on wildlife foods. p. 51 58. In Smith JK (ed), Wildland fire in ecosystems: effects of fire on fauna. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Ogden
148 Manly BFJ, McDonald LL, Thomas DL, McDonald TL, Erickson WP (2002) Resource selection by animals: statistic al design and analysis for field studies 2nd edition. Kluwer Academic Publishers, Dordrecht McAtee JW, Scifres CJ, Drawe DL (1979) Improvement of gulf cordgrass range with burning or shredding. Journal of Range Management 32:372 375 McCormick PV, O'Dell MB (1996) Quantifying periphyton responses to phosphorus in the Florida Everglades: A synoptic experimental approach. Journal of the North American Benthological Society 15:450 468 McCormick PV, Scinto LJ (1999) Influence of phosphorus loading on wetlands pe riphyton assemblages: A case study from the Everglades. p. 301 319. In Reddy KR, O'Connor GA, and Schelske CL (eds), Phosphorus biogeochemistry in subtropical ecosystems. Lewis Publishers, Boca Raton McCormick PV, Shuford RBE, III, Backus JG, Kennedy WC (1 998) Spatial and seasonal patterns of periphyton biomass and productivity in the northern Everglades, Florida, USA. Hydrobiologia 362:185 208 McWilliams SR, Sloat T, Toft CA, Hatch D (2007) Effects of prescribed fall burning on a wetland plant community, w ith implications for management of plants and herbivores. Western North American Naturalist 67:299 317 Means DB, Dodd CK, Jr., Johnson SA, Palis JG (2004) Amphibians and fire in longleaf pine ecosystems: Response to Schurbon and Fauth. Conservation Biology 18:1149 1153 Means DB, Moler PE (1979) The pine barrens treefrog: Fire, seepage bogs, and management implications. p. 77 83. In Odom RR, and Landers L (eds.), Proceedings of the rare and endangered wildlife symposium. Georgia Department of Natural Resourc es, Game and Fish Division, Athens Miao S, Edelstein C, Carstenn S, Gu B (2010) Immediate ecological impacts of a prescribed fire on a cattail dominated wetland in Florida Everglades. Fundamental and Applied Limnology 176:29 41 Miao SL, Sklar FH (1998) Bio mass and nutrient allocation of sawgrass and cattail along a nutrient gradient in the Florida Everglades. Wetlands Ecology and Management 5:245 263 Mitchell LR, Gabrey S, Marra PP, Erwin RM (2006) Impacts of marsh management on coastal marsh birds habitats Studies in Avian Biology 32:155 175 Mitsch WJ, Gosselink JG (2007) Wetlands 4th edition. John Wiley and Sons, Inc., Hoboken
149 Mokany A, Wood JT, Cunningham SA (2008) Effect of shade and shading history on species abundances and ecosystem processes in tempo rary ponds. Freshwater Biology 53:1917 1928 Moseley KR, Castleberry SB, Schweitzer SH (2003) Effects of prescribed fire on herpetofauna in bottomland hardwood forests. Southeastern Naturalist 2:475 486 Mosisch TD, Bunn SE, Davies PM (2001) The relative imp ortance of shading and nutrients on algal production in subtropical streams. Freshwater Biology 46:1269 1278 Mueller H (1999) Wilson's snipe ( Gallinago delicata ). In Poole A (ed.), The Birds of North America Online. Cornell Lab of Ornithology, Ithaca Munro NT, Kovac K J, Niejalke D, Cunningham RB (2009) The effect of a single burn event on the aquatic invertebrates in artesian springs. Austral Ecology 34:837 847 Newman S, McCormick PV, Miao SL, Laing JA, Kennedy WC, O'Dell MB (2004) The effect of phosphorus enrichment on the nutrient status of a northern Everglades slough. Wetlands Ecology and Management 12:63 79 Niehaus AC, Wilson RS, Franklin CE (2006) Short and long term consequences of thermal variation in the larval environment of anurans. Journal of A nimal Ecology 75:686 692 Noe GB, Childers DL, Jones RD (2001) Phosphorus biogeochemistry and the impact of phosphorus enrichment: Why is the Everglades so unique? Ecosystems 4:603 624 Norton DA, De Lange PJ (2003) Fire and vegetation in a temperate peat bo g: implications for the management of threatened species. Conservation Biology 17:138 148 Nyman JA, Chabreck RH (1995) Fire in coastal marshes: History and recent concerns. p. 134 141. In Cerulean SI, and Engstrom RT (eds.), Fire in wetlands: a management perspective. Proceedings of the Tall Timbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee Obaza A, DeAngelis DL, Trexler JC (2011) Using data from an encounter sampler to model fish dispersal. Journal of Fish Biology 78:495 513 Orvil le RE, Huffines GR (2001) Cloud to ground lightning in the United States: NLDN results in the first decade, 1989 98. Monthly Weather Review 129:1179 1193 Perkins CJ (1968) Controlled burning in the management of muskrats and waterfowl in Louisiana coastal marshes. p. 269 280. Proceedings of the 8th Annual Tall Timbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee
150 Pierce RL, Gawlik DE (2010) Wading bird foraging habitat selection in the Florida Everglades. Waterbirds 33:494 503 Pilliod DS, Bury RB, Hyde EJ, Pearl CA, Corn PS (2003) Fire and amphibians in North America. Forest Ecology and Management 178:163 181 Powell GVN (1987) Habitat use by wading birds in a subtropical estuary: Implications of hydrography. The Auk 104:740 749 Qian Y, Miao SL, Gu B, Li YC (2009) Effects of burn temperature on ash nutrient forms and availability from cattail ( Typha domingensis ) and sawgrass ( Cladium jamaicense ) in the Florida Everglades. Journal of Environmental Quality 38:451 464 R Development Core Team (2009) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna Rader RB, Richardson CJ (1992) The effects of nutrient enrichment on algae and macroin vertebrates in the Everglades: A review. Wetlands 12:121 1 35 Reardon J, Hungerford R, Ryan K (2007) Factors affecting sustained smouldering in organic soils from pocosin and pond pine woodland wetlands. International Journal of Wildland Fire 16:107 118 Reimer RD (1970) A food study of Heterandria formosa Agassiz. American Midland Naturalist 83:311 315 of prescribed burning on amphibian diversity. Conservation Biology 18:1154 1155 Rozas LP Minello TJ (1997) Estimating d ensities of s mall f ishes and d ecapod c rustaceans in s hallow e stuarine h abitats: A r eview of s ampling d esign w ith f ocus on g ear s election. Estuaries 20:199 213 Russell KR, Van Lear DH, Guynn Jr. DC (1999) Prescribed fire effects on herpetofauna: Review and management i mplications. Wildlife Society Bulletin 27:374 384 Rutchey K, Vilchek L, Love M (2005) Development of a vegetation map for Water Conservation Area 3, South Florida Water Managment District, West Palm Beach. Technical Publication ERA #421 Saab VA, Powell HDW (2005) Fire and avian ecology in North America: Process influencing pattern. Studies in Avian Biology:1 13 Saiers JE, Harvey JW, Mylon SE (2003) Surface water transport of suspended matter through wetland vegetation of the Florida Everglades. Geophysical Research Letters 30:1 5
151 Sartory DP, Grobbelaar JU (1984) Extraction of chlorophyll a from freshwater phytoplankton for spectrophotometric analysis. Hydrobiologia 114:177 187 Schlichtemeier G (1967) Marsh burning for waterfowl. p. 41 46. Proceedings of the 6th Annual Tall Timbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee Schurbon JM, Fauth JE (2003) Effects of prescribed burning on amphibian diversity in a southeastern U.S. national forest. Conservation Biology 17:1338 1349 Schurbon JM, Fauth JE (2004) Amphibians and fire in longleaf pine ecosystems: Response to Schurbon and Fauth. Conservation Biology 18:1156 1159 SFWMD (2012) DBHYDRO www.sfwmd.gov/dbhydro West Palm Beach Singleton JR ( 1951) Production and utilization of waterfowl food plants on the East Texas Gulf Coast. Journal of Wildlife Management 15:46 56 Sizemore GC (2009) Foraging quality of flooded agricultural fields within the Everglades Agricultural Area for wading birds (Cic oniiformes). Master of Science, University of Florida, Gainesville Slocum MG, Platt WJ, Beckage B, Panko B, Lushine JB (2007) Decoupling natural and anthropogenic fire regimes: A case study in Everglades National Park, Florida. Natural Areas Journal 27:41 55 Smallwood JA, Woodrey M, Smallwood NJ, Kettler MA (1982) Foraging by cattle egrets and American kestrels at a fire's edge. Journal of Field Ornithology 53:171 172 Smith LM, Kadlec JA, Fonnesbeck PV (1984) Effects of prescribed burning on nutritive quali ty of marsh plants in Utah. Journal of Wildlife Management 48:285 288 Smith SM, Newman S (2001) Growth of southern cattail ( Typha domingensis pers. ) seedlings in response to fire related soil transformations in the Northern Florida Everglades. Wetlands 21: 363 369 Smith SM, Newman S, Garrett PB, Leeds JA (2001) Differential effects of surface and peat fire on soil constituents in a degraded wetland of the northern Florida Everglades. Journal of Environmental Quality 30:1998 2005 Stevenson JO, Meitzen LH (194 6) Behavior and food habits of Sennett's white tailed hawk in Texas. The Wilson Bulletin 58:198 205 Stolen ED (2006) Habitat selection and foraging success of wading birds in impounded wetlands in Florida. PhD, University of Florida, Gainesville
152 Surdick JA (1998) Biotic and abiotic indicators of foraging site selection and foraging success of four Ciconiiform species in the freshwater Everglades of Florida. Master's Degree, University of Florida, Gainesville Taylor DL (1983) Fire management and the Cape Sab le seaside sparrow. p. 147 152. In Quay TL, Funderburg Jr. JB, Lee DS, Potter EF, and Robbins CS (eds.), The seaside sparrow, its biology and management. North Carolina Biological Survey, Raleigh Tewes ME (1984) Opportunistic feeding by white tailed hawks at prescribed burns. The Wilson Bulletin 96:135 136 Thomas S, Gaiser EE, Tobias FA (2006) Effects of shading on calcareous benthic periphyton in a short hydroperiod oligotrophic wetland (Everglades, FL, USA). Hydrobiologia 569:209 221 Toland BR (1987) The effect of vegetative cover on foraging strategies, hunting success and nesting distribution of American kestrels in central Missouri. Journal of Raptor Research 21:14 20 Travis J, Farr JA, McManus M, Trexler JC (1989) Environmental effects on adult growth patterns in the male sailfin molly, Poecilia latipinna (Poeciliidae). Environmental Biology of Fishes 26:119 127 Trexler JC, Loftus WF, Jordan F, Chick JH, Kandl KL, McElroy TC, Bass OL, Jr. (2002) Ecological scale and its implications for freshwater fishe s in the Florida Everglades. p. 153 181. In Porter JW, and Porter KG (eds), The Everglades, Florida Bay, and Coral Reefs of the Florida Keys: An Ecosystem Sourcebook. CRC Press, Boca Raton Trexler JC, Loftus WF, Perry S (2005) Disturbance frequency and com munity structures in a twenty five year intervention study. Oecologia 145:140 152 Tucker JW, Jr., Robinson WD (2003) Influence of season and frequency of fire on Henslow's sparrows ( Ammodramus henslowii ) wintering on Gulf Coast pitcher plant bogs. The Auk 120:96 106 Underwood AJ (1994) On beyond BACI: Sampling designs that might reliably detect environmental disturbances. Ecological Applications 4:3 15 van der Graaf S, Stahl J, Klimkowska A, Bakker JP, Drent RH (2006) Surfing on a green wave how plant growth drives spring migration in the Barnacle Goose Branta leucopsis Ardea 94:567 577 Vergeichik L, Kozuli n A (2006) Breeding ecology of a quatic warblers Acrocephalus paludicola in their key habitats in SW Belarus. Acta Ornithologica 41:153 161
153 Vickery JA, Tallowin JR, Feber RE, Asteraki EJ, Atkinson PW, Fuller RJ, Brown VK (2001) The management of lowland neutral grasslands in Britain: effects of agricult ural practices on birds and their food resources. Journal of Applied Ecology 38:647 664 Vogl RJ (1973) Effects of fire on the plants and animals of a Florida wetland. American Midland Naturalist 89:334 347 Wade D, Ewel J, Hofstetter R (1980) Fire in South Florida ecosystems. p. 125. US Department of Agriculture, Forest Service, Southeastern Forest Experiment Station, Asheville Walters JR, Beissinger SR, Fitzpatrick JW, Greenberg R, Nichols JD, Pulliam HR, Winkler DW (2000) The AOU Conservation Committee rev iew of the biology, status and management of Cape Sable seaside sparrows: Final report. The Auk 117:1093 1115 Ward P (1968) Fire in relation to waterfowl habitat of the Delta marshes. p. 254 267. Proceedings of the 8th Annual Tall Timbers Fire Ecology Conf erence. Tall Timbers Research Station, Tallahassee Warren SD, Scifres CJ, Teel PD (1987) Response of grassland arthropods to burning: A review. Agriculture, Ecosystems and Environment 19:105 130 Watts AC, Kobziar LN, Snyder JR (2012) Fire reinforces struct ure of pondcypress ( Taxodium distichum var. imbricarium ) domes in a wetland landscape. Wetlands 32:439 448 Wharton CH, Kitchens WM, Pendleton EC, Sipe TW (1982) The ecology of bottomland hardwood swamps of the Southeast: a community profile. p. 133. U.S. F ish and Wildlife Service, Biological Services Program, Washington Whelan RJ (1995) The ecology of fire. Cambridge University Press, Cambridge Williams A, Trexler J (2006) A preliminary analysis of the correlation of food web characteristics with hydrology and nutrient gradients in the southern Everglades. Hydrobiologia 569:493 504 Wu Q M, Zou H F (2011) Nest site selection pattern of Grus japonensis in Zhalong Nature Reserve of northeast China. Journal of Forestry Research 22:281 288 Zontek F (1966) Prescri bed burning on the St. Marks National Wildlife Refuge. p. 195 201. Proceedings of the 5th Annual Tall Timbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee Zou H F, Wu Q M, Ma J Z (2003) The nest site selection of red crowned crane in Zhalong Nature Reserve after burning and irrigating. Journal of Northeast Normal University Natural Sciences Edition 35:54 59
154 BIOGRAPHICAL SKETCH Louise S. Venne grew up in Wisconsin. She attended the University of Wisconsin Stevens Point where she earn ed Bachelor of Science degrees in Wildlife and in Chemistry. She then attended Texas Tech University for a Master of Science degree in Environmental Toxicology studying land use effects on amphibian community composition in playa wetlands. After working fo r a year as an environmental consultant, Louise enrolled in the Department of Wildlife Ecology and Conservation at University of Florida (UF). She was one of the fellows in the National Science Foundation funded Integrative Graduate Education and Research Traineeship programs at UF titled received her Ph.D. from the University of Florida in August 2012.