<%BANNER%>

The Relative Influences of Predation and Prey Availability on Ardeid Breeding Colony Site Selection


PAGE 1

THE RELATIVE INFLUENCES OF PR EDATION AND PREY AVAILABILITY ON ARDEID BREEDING COLONY SITE SELECTION By MATTHEW JOHN BOKACH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Matthew John Bokach

PAGE 3

This thesis is dedicated to my parents Mark and Cathy, who have weathered the many twists and turns of my lifes last decade with unconditional love and support, leavened with just a touch of legitimate bemusement.

PAGE 4

ACKNOWLEDGMENTS I could not have conducted this research or even made it through graduate school without the support and assistance of a humbling number of other people. Dr. Peter Frederick has been the best advisor I could have asked for. Without his support, patience, and humor, I would not have finished this degree. My data on the wading bird colony sites was collected by a long series of his former students and technicians, and this research is therefore as much a product of their work as mine. Dr. Graeme Cumming and Dr. Jane Southworth, my other two committee members, gave me invaluable advice through nearly every stage of the research. My initial explorations into the daunting task of analyzing these data benefited immensely from Dr. Tim Fiks statistical insights and encouragement. I gratefully acknowledge Dr. Carl Fitz, Mr. Jason Godin, and Mr. Ken Rutchey of the South Florida Water Management District; and, Mr. Troy Mullins of the National Park Service, for providing me with the hydrology and vegetation data used herein. Lorraine Heisler, Les Vilchek, and Shawn Komlos all provided insights concerning the development and appropriate uses of these data. Thanks to Theresa Burcsu and Desiree Price, I was allowed constant access to the computer labs of the University of Florida Geography Department, even when not enrolled in one of their classes; I will always think of Turlington Hall as the home of this research. Lin Cassidy, Andres Guhl, and especially Matt Marsik all gave me valuable assistance with GIS questions. I blame Carl Evans for both complicating and vastly simplifying my life by introducing me to Matlab. The patience and support of my boyfriend Rodney Brown iv

PAGE 5

was a steadying influence through most of this long process. Finally, I thank my family for all of their love and support through the years. I was supported financially for nearly my entire time at the University of Florida by a research assistantship through the College (now School) of Natural Resources and Environment. I extend my most sincere thanks to Dr. Steve Humphrey, Cathy Ritchie, and Meisha Wade for giving me the opportunity to pursue this degree, and for extending my funding for an extra semester. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 2 METHODS...................................................................................................................5 Study Area....................................................................................................................5 Data Sources.................................................................................................................6 Water Depth Grids.................................................................................................6 Vegetation Maps....................................................................................................6 Wading Bird Colony Locations.............................................................................8 Distributional Relationships.........................................................................................9 Calculation of Variables...............................................................................................9 Multivariate Analyses.................................................................................................11 3 RESULTS...................................................................................................................17 Distribution of Colony Sites.......................................................................................17 Bootstrap of Foraging Habitat Calculation.................................................................17 Multivariate Analyses.................................................................................................18 4 DISCUSSION.............................................................................................................38 Effects of Site Availability.........................................................................................38 Model Performance....................................................................................................38 Responses to Hydrological Variability.......................................................................39 Management Implications..........................................................................................40 LITERATURE CITED......................................................................................................42 BIOGRAPHICAL SKETCH.............................................................................................46 vi

PAGE 7

LIST OF TABLES Table page 1-1 Variables and methods of calculation........................................................................4 1-2 Weeks of nest initiation and duration of breeding period for three focal species......4 3-1 Number of colonies inhabited by each species, 1993-2000.....................................19 3-2 Values of FOR calculated in 30 randomly-chosen cells using constant proportions, compared to approximate 95% confidence intervals from a bootstrap analysis involving 1000 iterations where proportions were allowed to randomly vary..........................................................................................................20 3-3 Logistic regression models tested for Great Blue Herons and associated statistical values........................................................................................................21 3-4 Logistic regression models tested for Great Egrets and associated statistical values........................................................................................................................21 3-5 Logistic regression models tested for Tricolored Herons and associated statistical values........................................................................................................22 3-6 Best models for each species and associated measures of classification performance..............................................................................................................23 3-7 Relative importance of variables to each species.....................................................23 vii

PAGE 8

LIST OF FIGURES Figure page 2-1 Southeastern Florida showing the location of WCA3 within the larger Everglades ecosystem..............................................................................................14 2-2 Everglades Landscape Model cells in WCA3..........................................................15 2-3 Algorithm for determining which models to test for each species...........................16 3-1 Distribution of ardeid colonies in WCA3................................................................24 3-2 Relationship of Great Blue Heron colonies relative to all available sites in WCA3 as shown by linearized Ripleys K (l[d]) graphs.........................................26 3-3 Relationship of Great Egret colonies relative to all available sites in WCA3 as shown by linearized Ripleys K (l[d]) graphs..........................................................30 3-4 Relationship of Tricolored Heron colonies relative to all available sites in WCA3 as shown by linearized Ripleys K (l[d]) graphs.........................................34 viii

PAGE 9

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE RELATIVE INFLUENCES OF PREDATION AND PREY AVAILABILITY ON ARDEID BREEDING COLONY SITE SELECTION By Matthew John Bokach May 2005 Chair: Peter Frederick Major Department: Natural Resources and Environment Nest predation and prey availability are two of the most important factors affecting breeding success in long-legged wading birds (Ciconiiformes). I investigated whether the breeding colony site selection of three species of herons and egrets (family Ardeidae) in the Florida Everglades was influenced by environmental characteristics that mediated these two factors. These characteristics were 1) likelihood that sites remained inundated throughout the breeding period; 2) amount of foraging habitat around sites; 3) average weekly proportional decline in water depths around sites throughout the breeding period; 4) spatial variation in water depths around sites at the time of nesting. Variables were calculated within a geographic information system using both raster and vector inputs. I used measures derived from the Akaike information criterion to select the best logistic regression model and to evaluate the relative importance of these four variables for colony site selection in each species. ix

PAGE 10

Amount of foraging habitat and likelihood of remaining inundated were the most important variables influencing colony site selection by all three species. Great Blue Herons (Ardea herodius) also selected sites with high rates of average weekly proportional declines in water depth, but it is likely this variable was a proxy for deep water for this species rather than a reflection of prey availability. Overall, these species seemed to favor stable (rather than variable) hydrological conditions. This might also indicate that their colony site selection is based on conditions at the time of nesting rather than an attempt to predict future conditions. These results confirm the importance of managing the Everglades to maximize the extent of slough habitats, if the goal is to increase breeding populations of wading birds therein. x

PAGE 11

CHAPTER 1 INTRODUCTION Two of the most important factors affecting the breeding success of long-legged wading birds (Ciconiiformes) are egg and chick predation, and the availability of adequate food to raise chicks to fledging (Taylor & Michael 1971; Frederick & Collopy 1989a; Frederick & Spalding 1994; Frederick 2002). Although wading birds cannot control the behaviors of predators or their prey, it is likely that selection has favored the recognition of breeding colony sites with beneficial environmental characteristics that mediate or constrain those behaviors. For example, wading birds nearly always nest on islands or in trees and/or shrubs that are inundated at their base, a pattern that has most often been interpreted as a strategy for deterring mammalian predators (Rodgers 1987; Bancroft et al. 1988; Frederick & Collopy 1989b; Smith & Collopy 1995). However, whether this pattern represents actual selection for this characteristic has never been quantitatively studied. With regard to prey availability, although previous studies (Gibbs et al. 1987; Gibbs 1991; Gibbs & Kinkel 1997; Baxter & Fairweather 1998; Bancroft et al. 2002) have shown that wading birds select colony sites that maximize the amount of wetland habitats within a reasonable foraging range, other environmental characteristics (e.g., water depth) are better determinants of wading birds ability to capture prey. If wading bird colony site selection represents an attempt to deter predators and/or maximize prey availability, then used sites should be measurably different from available unused sites with relation to environmental characteristics that affect these two factors. I 1

PAGE 12

2 identified one characteristic that deters predators from reaching nests and three characteristics that are likely to influence prey availability around sites (Table 1-1) and make the following predictions. Prediction 1: Used sites will have a greater likelihood of remaining inundated throughout the breeding period than unused sites. This characteristic seems to be a good measure of a sites ability to deter mammalian predators (Frederick & Collopy 1989b; Smith & Collopy 1995), which are generally the most destructive in their effects on wading bird colonies (Rodgers 1987; Post 1990; Smith & Collopy 1995). Prediction 2: Used sites will be surrounded by more open or sparsely vegetated habitats than unused sites. Not all wetland habitats are appropriate foraging grounds for wading birds. In particular, they are known to avoid dense vegetation (such as monospecific stands of sawgrass [Cladium jamaicense] or cattail [Typha latifolia]), which interferes with their visual or tactile hunting techniques, provides prey animals with more hiding places, and could serve as a hiding place for predators of the birds themselves (Hoffman et al. 1994; Smith et al. 1995; Surdick 1998). Prediction 3: Used sites will be located in areas that experience greater proportional declines in water depth throughout the breeding season than unused sites. Wading birds are generally limited to foraging in water that is shallower than the length of their bills or legs (Custer & Osborn 1978; Powell 1987). They are often attracted to areas of declining water depth where prey have been concentrated into pools or depressions that remain inundated late in the season (Kushlan 1976; Bancroft 1989; Frederick & Collopy 1989a; Smith 1995; Gawlik 2002; but see Frederick & Spalding 1994 for a critique of this idea).

PAGE 13

3 Prediction 4: Variability in water depths at the time of nesting will be greater around used sites than around unused sites. Water levels are dynamic and unpredictable in most wetlands. Heterogeneous topography around a colony site could therefore provide a foraging advantage to wading birds as this would provide the most diversity of potential foraging sites at almost any time or water condition (Kahl 1964; Bancroft et al. 2002). I used logistic regression models to assess the relative influence of these four variables on the colony site selection of three wading bird species in the ciconiiform family Ardeidae (Table 1-2) over a period of 8 years in the Florida Everglades. Measures derived from the models Akaike information criteria (AIC) (Akaike 1973; Burnham & Anderson 2002) were used to choose the best model and determine the relative influence of the variables on the colony site selection for each species.

PAGE 14

4 Table 1-1. Variables and methods of calculation. Variable Abbreviation Method of calculation Affecting predator deterrence Likelihood that a cell will remain inundated for the duration of the breeding period INUN Used GREATERTHAN function to output number of weeks over duration of a species' breeding period that depths remained above 0; divided by number of weeks to yield values between 0.0 and 1.0 Affecting prey availability Amount of foraging vegetation around sites FOR See text for details Average rate of proportional decline in depths around a site for the duration of a species' breeding period FWI Calculated proportional decline in water depths for every week during a species' breeding period (rises in depth expressed as 0); averaged these values over all weeks; used the FOCALMEAN function to average these cell averages over 3x3 cell neighborhoods, yielding values between 0.0 and 0.5 Spatial variation in water depths around a site at week of nest initiation WV Used FOCALSTD func tion to calculate standard deviation of depths in 3x3 cell neighborhoods All calculations carried out in ESRI Arc/INFO workstation v.8.3 except FOR, which was calculated in ESRI ArcMap v.8.3. Weeks of nest initiation and duration of breeding periods are listed in Table 1-2. Table 1-2. Weeks of nest init iation and duration of breeding period for three focal species Species Week of nest initiationa Duration Source for duration Great Blue Heron ( Ardea herodius ) 9th 14 weeks Butler 1992 Great Egret ( Ardea alba ) 9th 10 weeks McCrimmon et al. 2001 Tricolored Heron ( Egretta tricolor ) 12th 11 weeks Frederick 1997 aSource is Frederick (pers. comm.)

PAGE 15

CHAPTER 2 METHODS Study Area I studied ardeid colony site selection within Water Conservation Area (WCA) 3, a 2,350-km2 human-made impoundment in the central Everglades (Figure 2-1). Most of the wading birds that breed within the Everglades ecosystem have selected colony sites within this impoundment since the early 1970s (Ogden 1994; Frederick & Ogden 2001). The annual hydrology of WCA3 is characterized by a decline in water depths during the dry season between November and April, and most annual rainfall occurring during a subtropical wet season between June and November. The northern end of the impoundment is shallow and quick to dry, while the southern end is almost permanently inundated. This same gradient exists to a lesser extent from west (where flow of water into the adjoining Big Cypress National Preserve is unimpeded) to east (the enclosed sub-impoundment WCA3B created by the L-67 canals, Figure 2-1). Wading birds therefore have a wide range of hydrological conditions within which to select sites in WCA3. The vegetation of WCA3 is characterized by open wet prairie communities in deeper areas interspersed between narrow ridges covered in cattail and/or sawgrass (Gunderson 1994). The patches of woody or shrubby vegetation that serve as potential nesting sites for wading birds are scattered throughout. Large areas in the northern end of the impoundment are dominated by an almost complete monoculture of cattails. 5

PAGE 16

6 Data Sources Water Depth Grids I derived hydrological variables using estimated water depths from the Everglades Landscape Model (ELM; Fitz et al. 2004), which simulates the flows and stages of water across the entire Everglades ecosystem at a resolution of 1 km2. Model output was available for the period 1993, so these 8 years comprised the duration of my study. Each output layer is an Arc/Info grid (Environmental Systems Research Institute 2002) whose values are the estimated weekly average above-ground depth of water in each cell. The model generally performs well both spatially and temporally when compared to actual depths measured throughout the landscape (Fitz et al. 2004). The largest errors occur in cells containing canals, as these contain the most intracellular variation in water depths. ELM 1 km2 cells, rather than discrete colony sites, were my units of analysis as these were the coarsest in resolution of all the data. Vegetation Maps I used a polygon vector geographic information system (GIS) layer showing the distribution of vegetation in the central and southern Everglades and based on aerial infrared photography acquired in 1995 over the entire area of WCA3, Everglades National Park, and Big Cypress National Preserve. The latter two areas were included in my analysis since they are within the typical foraging range of birds nesting near the boundaries of WCA3 (Bancroft et al. 1994). The distinctive reflectance signatures of different types of vegetation within the aerial photography were used to digitize polygons representing discrete patches. These could contain up to three specific vegetative classes, assigned hierarchically. After comparing the polygon layer to 1999 United States Geological Survey (USGS) digital orthophoto quarter quads (DOQQs), I was satisfied

PAGE 17

7 that this layer was an accurate depiction of the vegetation in WCA3 for the entire time period of this study. Since all reports of nesting by ardeid wading birds are on trees or shrubs in WCA3, I defined potential colony sites as polygons that contained at least one class of woody vegetation. Among the classifications used in the vegetation map these included: all variants of forest, all variants of shrublands, Melaleuca, Brazilian pepper, disturbed fish camp sites, spoil areas, and artificial deer islands. I aggregated contiguous polygons that contained at least one of these vegetation types into single discrete patches, which I hereafter refer to interchangeably as patches of woody vegetation or potential colony sites. The locations of wading bird colonies (see below) revealed that a small proportion (~1.1%) of the potential colony sites did not occur within a reasonable distance from a patch of woody vegetation. In October 2002 I visited all such sites in WCA3B and in WCA3A north of Highway I-75 (Figure 2-1), and determined that in most cases either some woody vegetation had not been included in an existing polygon, or that a single tree or small patch of shrubs was present and should have itself been digitized as a polygon. In these cases, I therefore either added a woody vegetation classification to the second or third tier of the existing polygon, or digitized the patch myself using the 1999 USGS DOQQs. Water model cells that either contained no potential colony sites, or contained only the clipped edges of sites whose bulk were in an adjacent cell, were excluded from further analysis since wading birds could not have nested in them. At the scale at which I

PAGE 18

8 calculated the variables (see below), this removed less than 3% of the total area of WCA3 from analysis (Figure 2-2). Wading Bird Colony Locations Locations, species composition, and size of wading bird colonies were determined using both aerial and ground survey techniques (Frederick et al. 1996). Aerial surveys were designed such that the entire ground surface of the study area was visible along at least one east-west transect. These were spaced 2.9 km apart, flown at 245 m altitude, and were conducted monthly between January and June of each year. Dark-colored species are often not visible in aerial surveys (Frederick et al. 1996), and for these species ground-based airboat searches were conducted. All tree islands in the study area were approached and searched for signs of nesting in the middle to late part of the nesting season (April through May). Colony site coordinates were determined by handheld GPS receivers, and populations of every species present were estimated by counting nests and/or number of adults that flushed from the site. In cases where a colony was visited more than once per season, I used the peak count of nests at that colony for a given species. I created point vector GIS layers representing the colony locations for each of the three focal species for every year. Given that point coordinates reflected the location of the airboat or airplane at the time of data collection, rather than the actual location of the colony, I used a spatial join operation to shift points to the centroid of the nearest patch of woody vegetation. In general, these shifts were not more than 100 m for points collected on the ground, or 300 m for points collected during aerial surveys. I then overlaid the water model grid and classified all cells as either used or unused by each of the species for every year.

PAGE 19

9 Distributional Relationships To rule out the possibility that the distribution of these species colonies was simply a reflection of the distribution of available sites, I calculated Ripleys K statistic for each species-year pattern of colony sites as well as for the parent pattern of all available sites, at 500 m intervals from 500 to 20,000 m. This statistic is a scale-dependent measure of spatial dispersion within a defined area (Fotheringham et al. 2000). However, it can also be used to assess whether two patterns within the same area are spatially similar, especially when one is a subset of the other. Differences between two patterns K statistics (i.e., the linearized Ripleys K, or l[d]) indicate that they are spatially dissimilar, with positive values of l(d) indicating a dispersed pattern relative to the parent pattern, and negative values of l(d) indicating a clustered pattern relative to the parent pattern. Since there is no significance test to determine how different from 0 an l(d) value must be to indicate that the patterns are different, I used bootstrapping to create sample-size-dependent 95% confidence intervals against which to compare each species-year point pattern (Fotheringham et al. 2000). Calculation of Variables All variables were calculated at the scale of a 3x3 neighborhood of cells (i.e., 3 km x 3 km) centered on the cell in question except INUN, which was calculated at the scale of the cell itself. I chose these small scales to maximize the spatial independence between samples, since larger neighborhoods would have increased the distance required between two cells before their neighborhoods did not overlap. The calculation of the three hydrological variables is described in Table 1-1. To calculate FOR I first identified all polygons in the vegetation map that contained any of the foraging habitat types. Within the maps classification system these included:

PAGE 20

10 all variations of savanna; spike rush and maidencane-spike rush prairies; all variations of non-graminoid emergent marsh; open water; mixed mangrove scrub; and, buttonwood and saw palmetto scrubs. Vegetation classes were hierarchically assigned to polygons qualitatively (e.g., a large area of wet prairie might contain some scattered shrubs), yet I needed a quantitative estimate of the areal extent of each class within polygons. Since vegetation classes could be assigned to any of the tiers within a polygon, with adjoining polygons often having the same classes in reversed order, the least biased method of estimating these extents was to allocate the same proportion to a given tier in every polygon. In polygons containing two classes of vegetation, I allocated 70% of the area to the dominant vegetation type and 30% to the secondary; in three-class polygons, I allocated 50, 30, and 20% of the area to the three tiers respectively. After allocating a polygons area between its vegetation classes, I calculated the total extent of foraging habitat present within the polygons. I intersected these polygons with a square grid polygon layer corresponding to the ELM grid cell boundaries to assign the vegetation polygons to the cell they inhabited, and summed the areas of foraging vegetation present within every cell. I recalculated the area of foraging vegetation in polygons that had been split by cell boundaries by multiplying their new area by the same proportion of foraging vegetation calculated for the parent polygon. Finally, I calculated the 3x3 neighborhood sums for every cell in WCA3. To ensure that the proportions I chose did not bias the ultimate values of FOR, I randomly chose 30 cells and performed a bootstrap analysis whereby the proportions allocated to vegetation classes within their polygons were allowed to randomly vary. Since the vegetation classes were originally assigned hierarchically, I constrained the

PAGE 21

11 proportions such that the dominant class proportion was always greater than the secondary, and this was always greater than the tertiary. I calculated FOR for each cell 1000 times based on these randomly-determined proportions and used these values to create an approximate 95% confidence interval for the real value of FOR in each cell. I then determined how many of the values of FOR calculated in the 30 cells using the proportions described above fell within these confidence intervals. Multivariate Analyses I used the Number Cruncher Statistical Systems software (NCSS, Hintze 2001) to evaluate logistic regression models where the response variable was the use (presence of at least one colony) or non-use of a water model cell by a species. Model inputs consisted of the values for all cells used by a given species in all 8 years combined with an equal number of randomly-selected unused cells from each year. I bootstrapped these randomly-selected sets for every species-year to ensure that their mean values for all four variables were within the 95% confidence interval suggested by the bootstrap. Results of the logistic regression analyses were interpreted within the information-theoretic approach of Burnham and Anderson (2002). Multiple models were evaluated and the Akaike information criterion (AIC; Akaike 1973) calculated for each one thus: AIC = -2(log-likelihood ratio) + 2K (2-1) where K was the number of parameters estimated in the model. The model with the lowest AIC was the best model among all those evaluated. The last term in Eq. 2-1 therefore represented a penalty against a models AIC value for every variable added to the model.

PAGE 22

12 The performance of a given model relative to the best one in the set was expressed using three values derived from the AIC values. The AIC difference (i) for model i was calculated: i = AICi AICmin (2-2) This was converted to Akaike weights (wi): wi = Rrri12121)exp(/)exp( (2-3) where the denominator in Eq. 2-3 was the sum of all models i values. The wi for model i was a measure of the likelihood (often expressed as a percent) that it is the best model. An analogous quantity is the evidence ratio for model i: evidence ratioi = wi / wj (2-4) where wj was the wi value for the best model in the set. The evidence ratio was a measure of how much better the best model was than model i. Because they relied on the sum of all models i values, the wi and evidence ratios were recalculated every time a new model entered the set. I followed the algorithm shown in Figure 2-3 to determine the models tested for each species. Upon identifying the best model for each species, I report three measures of its classification performance (Fielding & Bell 1997): 1) the number of used and unused cells correctly classified; 2) the Kappa statistic (), which measures the extent to which a models predictions are correct above the proportion expected simply by chance; and, 3) the area under the models ROC curve, which measures the likelihood that the value assigned by the final model to a randomly-selected used cell will be greater than that of a randomly-selected unused cell. I also summarized the relative importance of the

PAGE 23

13 variables tested for each species by adding the wi values for all models in which a given variable is included (Burnham & Anderson 2002: 167-168).

PAGE 24

14 Figure 2-1. Southeastern Florida showing the location of WCA3 within the larger Everglades ecosystem. The dotted areas along the east coast show the extent of urban landscapes and their proximity to what is left of the Everglades.

PAGE 25

15 Figure 2-2. Everglades Landscape Model cells in WCA3. Black dots indicate cells that contained potential colony sites and were included in multivariate analysis. The darker shaded area indicates all cells that were within a 3x3 cell neighborhood of analyzed cells and whose characteristics were therefore included in the calculation of variables. Less than 3% of the area of WCA3, represented by the unshaded areas, was excluded from analysis.

PAGE 26

16 Test saturated model (all 4 variables ) and all 3-term models. Best model has 3 terms Test all 2-term models nested within best model Figure 2-3. Algorithm for determining which models to test for each species. At every step the current best model competed with lower-term models comprised of its component variables. Once the best single-order model was identified, models that added a single second-order interaction term were tested. Test all 1-term models nested within best model Test all models comprised of best model + single second-order interaction term between component variables. BEST MODEL Best model has 4 terms Best model has 2 terms Best model has 3 terms Best model has 2 terms Best model has 1 term

PAGE 27

CHAPTER 3 RESULTS Distribution of Colony Sites Table 3-1 summarizes the number of colonies and water model cells inhabited by each species in each year, and Figure 3-1 presents the annual maps of these distributions. Figures 3-2 through 3-4 show the l(d) graphs for each species-year combination, along with their corresponding approximate 95% confidence intervals from the bootstrap simulation. Great Blue Heron colonies were extremely clustered relative to the parent pattern of all potential colony sites in all years except 1998, when the l(d) values were closer to the lower bootstrap limit. Great Egret colonies were dispersed relative to the parent pattern at all scales in 1993, 1994, and 1995; at small scales in 1996, and at large scales in 1998. Their colony patterns were mostly indistinguishable from the parent pattern at practically all scales in 1997, 1999, and 2000. Tricolored Heron colony patterns showed the most variability: they were highly dispersed relative to the parent pattern at medium to large scales in 1993; highly clustered at all scales in 1996; slightly clustered at small scales in 1994 and medium scales in 1999 and 2000; and, otherwise mostly indistinguishable from the parent pattern in 1995, 1997 and 1998. Bootstrap of Foraging Habitat Calculation Table 3-2 lists the values of FOR calculated for 30 randomly-chosen cells using the constant proportions described above and the upper and lower values of the approximate 95% confidence interval from the bootstrap analysis. Four of the 30 cells contained only polygons that did not vary in their values of FOR. In only one of the remaining cells was 17

PAGE 28

18 the value of FOR calculated using the constant proportions outside the 95% confidence interval from the bootstrap; however, this cells value was within the 99% confidence interval. This is strong evidence that the constant proportions used to calculate this variable did not bias the analysis. Multivariate Analyses I tested 11 models for Great Blue and Tricolored Herons, and 12 for Great Egrets (Tables 3-3 through 3-5). A single best model (wi > 98%) was found for the first two of these species, while the best two models for Great Egrets were nearly indistinguishable (wi values of 29.20% and 23.18%, respectively). Table 3-6 lists the best model(s) for each species and its associated measures of classification performance. Best models correctly classified between 68.8% (Great Egrets, model 1) and 78.9% (Great Blue Herons) of the used sites, and between 36.7% (Great Egrets, model 1) and 56.7% (Great Blue Herons) of the unused sites. Kappa statistics were routinely low, ranging between 0.055 (Great Egrets, model 1) and 0.357 (Great Blue Herons). The areas under the ROC curves also indicated relatively poor model performance; these ranged between 0.524 (Great Egrets, model 1) and 0.745 (Great Blue Herons). The amount of foraging habitat (FOR) around potential sites was always the most important variable, with all wi values greater than 98% (Table 3-7). It was followed by the likelihood that a site would remain inundated (INUN), with wi values ranging between 83.88 and 99.99%. As predicted, all three species responded to both of these variables positively, i.e., used cells had higher values overall than unused cells. The pattern for the remaining variables was less consistent across species, with Great Blue Herons responding strongly and positively to FWI, and Great Egrets responding weakly and negatively to WV (Table 3-7).

PAGE 29

19 Table 3-1. Number of colonies inhabited by each species, 1993-2000. Number of water model cells inhabited in parentheses. Great Blue Heron Great Egret Tricolored Heron 1993 156 (127) 32 (32) 11 (11) 1994 200 (153) 45 (44) 46 (43) 1995 298 (229) 39 (39) 40 (39) 1996 167 (136) 45 (43) 52 (42) 1997 101 (90) 34 (33) 8 (8) 1998 110 (97) 44 (41) 23 (23) 1999 258 (206) 77 (71) 66 (63) 2000 309 (235) 71 (62) 57 (53)

PAGE 30

20 Table 3-2. Values of FOR calculated in 30 randomly-chosen cells using constant proportions, compared to approximate 95% confidence intervals from a bootstrap analysis involving 1000 iterations where proportions were allowed to randomly vary. Lower limit of 95% confidence interval Value of FOR calculated using constant proportions Upper limit of 95% confidence interval Observed within confidence interval? 75 75 75 Constant value cell 251 339 644 Yes 750 750 750 Constant value cell 118 887 1,155 Yes 149 1,742 2,389 Yes 4,321 4,321 4,321 Constant value cell 5,925 5,925 5,925 Constant value cell 9,763 10,296 12,057 Yes 1,526 20,539 29,209 Yes 24,992 25,858 26,380 Yes 20,804 35,765 42,190 Yes 74,514 80,693 84,326 Yes 101,367 103,625 119,108 Yes 146,646 149,072 150,791 Yes 172,177 218,953 244,864 Yes 163,585 256,693 339,316 Yes 281,378 286,925 305,647 Yes 28,983 289,828 328,472 Yes 337,236 352,078 356,356 Yes 115,366 360,059 418,467 Yes 77,850 376,847 504,702 Yes 92,661 382,298 501,854 Yes 373,516 445,018 635,874 Yes 220,282 470,162 569,560 Yes 413,803 477,840 597,608 Yes 271,067 504,240 587,148 Yes 307,407 507,071 567,364 Yes 385,418 532,264 559,474 Yes 504,530 607,378 604,063 No, 99% confidence interval 531,142 638,535 646,606 Yes

PAGE 31

21 Table 3-3. Logistic regression models tested for Great Blue Herons and associated statistical values. Italicized models are the best from their subset, while the best overall model is shown in boldface. Model Log-Likelihood K AIC AIC wia Evidence Ratiosb Rank FOR+WV+FWI+INUN -1492.38 5 2994.76 16.15 3 FOR+WV+FWI -1495.94 2999.88 4 21.28 9 FOR+FWI+INUN -1493.21 2994.42 4 15.81 0.04% 2 FOR+WV+INUN -1495.21 2998.42 4 19.81 6 WV+FWI+INUN -1721.50 3451.00 4 472.40 10 FOR+FWI -1496.75 2999.51 3 20.90 8 FOR+INUN -1496.40 2998.81 3 20.21 7 FWI+INUN -1745.57 3497.14 3 518.54 11 FOR+FWI+INUN+FOR*FWI -1492.98 5 2995.96 17.36 0.02% 4 FOR+FWI+INUN+FOR*INUN -1493.15 5 2996.29 17.69 0.01% 5 FOR+FWI+INUN+FWI*INUN -1484.30 5 2978.60 0.00 99.89% 1.00 1 a indicates a wi value < 0.01%; b indicates an evidence ratio > 100 Table 3-4. Logistic regression models tested for Great Egrets and associated statistical values. Italicized models are the best from their subset, while the best overall models are shown in boldface. Model Log-Likelihood K AIC AIC wi Evidence Ratiosa Rank FOR+WV+FWI+INUN -498.86 5 1007.73 4.22 3.54% 8.25 8 FOR+WV+FWI -499.97 4 1007.94 4.44 3.17% 9.20 9 FOR+FWI+INUN -499.59 4 1007.17 3.67 4.67% 6.26 7 FOR+WV+INUN -498.86 4 1005.73 2.22 9.62% 3.03 4 WV+FWI+INUN -502.60 4 1013.20 9.69 0.23% 12 FOR+WV -500.48 1006.96 3 5.18% 3.46 5.63 6 FOR+INUN -499.61 1005.21 3 12.45% 1.71 2.35 3 WV+INUN -502.69 1011.38 3 0.57% 7.88 51.35 10 FOR -501.08 1006.16 2 2.65 7.76% 3.76 5 INUN -503.99 1011.99 2 8.48 0.42% 69.43 11 FOR+INUN+FOR*INUN -497.75 4 1003.51 0.00 29.20% 1.00 1 FOR+WV+INUN+FOR*INUN -496.98 5 1003.97 0.46 23.18% 1.26 2 b indicates an evidence ratio > 100

PAGE 32

22 Table 3-5. Logistic regression models tested for Tricolored Herons and associated statistical values. Italicized models are the best from their subset, while the best overall model is shown in boldface. Model Log-Likelihood K AIC AIC wia Evidence Ratiosb Rank FOR+WV+FWI+INUN -363.45 5 736.90 14.14 0.08% 8 FOR+WV+FWI -364.24 4 736.48 13.72 0.10% 7 FOR+FWI+INUN -363.83 4 735.66 12.90 0.16% 6 FOR+WV+INUN -363.50 4 734.99 12.23 0.22% 4 WV+FWI+INUN -387.66 4 783.33 60.57 10 FOR+WV -364.70 735.40 3 12.64 0.18% 5 FOR+INUN -363.92 733.84 3 11.07 0.39% 2 WV+INUN -387.67 781.33 3 58.57 9 FOR -365.35 734.70 2 11.94 0.25% 3 INUN -390.82 785.64 2 62.88 11 FOR+INUN+FOR*INUN -357.38 4 722.76 0.00 98.62% 1.00 1 a indicates a wi value < 0.01%; b indicates an evidence ratio > 100

PAGE 33

23 Table 3-6. Best models for each species and associated measures of classification performance. Great Blue Herons Model FOR + FWI + INUN FWI*INUN n, % Correct (Used) 1005, 78.9% n, % Correct (Unused) 722, 56.7% Statistic 0.357 Area Under ROC Curve 0.745 Great Egrets Model 1 FOR + INUN + FOR*INUN n, % Correct (Used) 251, 68.8% n, % Correct (Unused) 134, 36.7% Statistic 0.055 Area Under ROC Curve 0.524 Model 2 FOR WV + INUN + FOR*INUN n, % Correct (Used) 274, 75.1% n, % Correct (Unused) 152, 41.6% Statistic 0.167 Area Under ROC Curve 0.559 Tricolored Herons Model FOR + INUN + FOR*INUN n, % Correct (Used) 218, 77.3% n, % Correct (Unused) 137, 48.6% Statistic 0.259 Area Under ROC Curve 0.654 Table 3-7. Relative importance of variables to each species. Great Blue Heron Great Egret Tricolored Heron FOR 100.00% 98.78% 100.00% WV 0.04% 45.50% 0.58% FWI 99.99% 11.61% 0.34% INUN 99.99% 83.88% 99.47% FOR*FWI 0.02% FOR*INUN 0.01% 52.38% 98.62% FWI*INUN 99.89%

PAGE 34

24 Figure 3-1. Distribution of ardeid colonies in WCA3. A) 1993. B) 1994. C) 1995. D) 1996. E) 1997. F) 1998. G) 1999. H) 2000. Triangles indicate Great Blue Heron colonies, stars indicate Great Egret colonies, and circles indicate Tricolored Heron colonies.

PAGE 35

25 Figure 3-1. Continued

PAGE 36

26 A-1-0.8-0.6-0.4-0.200.20.40.605000100001500020000Scale (m)l(d) B-1.2-1-0.8-0.6-0.4-0.200.20.40.605000100001500020000Scale (m)l(d) Figure 3-2. Relationship of Great Blue Heron colonies relative to all available sites in WCA3 as shown by linearized Ripleys K (l[d]) graphs. A) 1993. B) 1994. C) 1995. D) 1996. E) 1997. F) 1998. G) 1999. H) 2000. Sample sizes for each year listed in Table 3-1. Dashed lines: 95% confidence intervals from Monte Carlo simulations of 999 identically-sized, randomly-selected samples from each year; solid line: l(d) graph for used sites.

PAGE 37

27 C-0.8-0.6-0.4-0.200.20.405000100001500020000Scale (m)l(d) D-1.2-1-0.8-0.6-0.4-0.200.20.40.605000100001500020000Scale (m)l(d) Figure 3-2. Continued

PAGE 38

28 E-2-1.5-1-0.500.5105000100001500020000Scale (m)l(d) F-1.4-1.2-1-0.8-0.6-0.4-0.200.20.40.60.805000100001500020000Scale (m)l(d) Figure 3-2. Continued

PAGE 39

29 G-0.8-0.6-0.4-0.200.20.405000100001500020000Scale (m)l(d) H-0.6-0.5-0.4-0.3-0.2-0.100.10.20.30.405000100001500020000Scale (m)l(d) Figure 3-2. Continued

PAGE 40

30 A-2-1.5-1-0.500.511.5205000100001500020000Scale (m)l(d) B-2-1.5-1-0.500.511.5205000100001500020000Scale (m)l(d) Figure 3-3. Relationship of Great Egret colonies relative to all available sites in WCA3 as shown by linearized Ripleys K (l[d]) graphs. A) 1993. B) 1994. C) 1995. D) 1996. E) 1997. F) 1998. G) 1999. H) 2000. Sample sizes for each year listed in Table 3-1. Dashed lines: 95% confidence intervals from Monte Carlo simulations of 999 identically-sized, randomly-selected samples from each year; solid line: l(d) graph for used sites.

PAGE 41

31 C-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) D-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) Figure 3-3. Continued

PAGE 42

32 E-2-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) F-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) Figure 3-3. Continued

PAGE 43

33 G-1.5-1-0.500.5105000100001500020000Scale (m)l(d) H-1.5-1-0.500.5105000100001500020000Scale (m)l(d) Figure 3-3. Continued

PAGE 44

34 A-2-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) B-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) Figure 3-4. Relationship of Tricolored Heron colonies relative to all available sites in WCA3 as shown by linearized Ripleys K (l[d]) graphs. A) 1993. B) 1994. C) 1995. D) 1996. E) 1997. F) 1998. G) 1999. H) 2000. Sample sizes for each year listed in Table 3-1. Dashed lines: 95% confidence intervals from Monte Carlo simulations of 999 identically-sized, randomly-selected samples from each year; solid line: l(d) graph for used sites.

PAGE 45

35 C-2.5-1.5-0.50.51.52.505000100001500020000Scale (m)l(d) D-2-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) Figure 3-4. Continued

PAGE 46

36 E-2.5-1.5-0.50.51.52.53.54.505000100001500020000Scale (m)l(d) F-2-1.5-1-0.500.511.522.505000100001500020000Scale (m)l(d) Figure 3-4. Continued

PAGE 47

37 G-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) H-1.5-1-0.500.511.505000100001500020000Scale (m)l(d) Figure 3-4. Continued

PAGE 48

CHAPTER 4 DISCUSSION Effects of Site Availability The spatial patterns of colony sites used by these three species during the study period were in general dissimilar from the overall pattern of available sites (Figures 3-2 through 3-4). This was always true for Great Blue Herons, and was true at most scales for Great Egrets and Tricolored Herons in 4 and 5 of the 8 years, respectively. These two species exhibited both clustered and dispersed patterns, which likely reflects responses by the birds to different hydrological conditions in each year. These results indicate that site selection by these three species was probably independent of overall site availability or distribution. Model Performance Although there was very strong evidence supporting the best models for Great Blue and Tricolored Herons, relative to the total set of models I tested, model classification performance was poor overall. None of the best models achieved a Kappa statistic of 0.4, which is often considered to be the minimum value for good performance (Fielding & Bell 1997), and only the best model for Great Blue Herons had an area under its ROC curve of at least 0.7. Models always did a better job of classifying used than unused cells (Table 3-6), which suggests that there were more appropriate sites available than were used (Fielding & Bell 1997). A major limitation of this analysis was the coarse spatial resolution of the hydrological data, which prevented me from including variables that operate at the scale of the sites themselves such as vegetation type or patch size. Finally, 38

PAGE 49

39 simply classifying cells as used or unused ignored the enormous variation in the number of Great Egret pairs nesting at sites within WCA3, which is probably why the models for this species performed the worst. Despite these limitations, these results clearly show that the amount of foraging habitat around sites, and the likelihood that they would remain inundated throughout the breeding season, were important influences on the colony site selection for these three species. This is the first study to quantitatively establish the importance of deterring mammalian predators as an influence on colony site selection in this group. These results also establish that the distribution of foraging habitats is as important to wading birds in an enormous wetland-dominated landscape such as the Everglades, as it is in more typical terrestrial landscapes where wetlands are part of a habitat mosaic (Gibbs et al. 1987; Gibbs 1991; Gibbs & Kinkel 1997). Responses to Hydrological Variability In general, neither the average weekly proportional decline in water depths (FWI) nor the spatial variation in water depths (WV) were important influences on colony site selection in these three species. WV was weakly selected against by Great Egrets, and not included in the best models for either of the other two species. Only Great Blue Herons selected cells with large values of FWI. However, areas that began the breeding season with deeper water were likely to have the largest values of for this variable, since they could experience proportional declines for much longer before going dry. Thus it is unclear whether FWI actually represented prey availability for Great Blue Herons or was instead a proxy for deep water. Great Blues also selected against cells that had both rapid declines in water depths and a high likelihood of remaining inundated (i.e., the INUN*FWI interaction term). These could represent either permanently shallow areas

PAGE 50

40 (such as on the west side of WCA3, where water flow westward into Big Cypress National Preserve is unimpeded), and/or areas near canals, both of which were mostly avoided by this species (Figure 3-1). These results suggest that ardeid wading birds are selecting colony sites that are characterized by stable, rather than variable, water depths. Another possible interpretation is that colony site selection by these species does not reflect an attempt to predict future conditions (i.e., a bet hedging strategy), but is rather simply a response to conditions at the time of nesting (Bancroft et al. 1994). These three species are known to be less dependent on concentrated patches of prey created by declining water depths than tactile/social foragers such as Wood Storks (Mycteria americana), White Ibises (Eudocimus albus), and Snowy Egrets (Egretta thula) (Kahl 1964; Gawlik 2002), so it is also possible that they have not experienced any selective pressure to recognize or respond to this environmental characteristic. Management Implications Previous authors have speculated that the size of wading bird breeding populations in the Everglades is dependent on the extent of open slough habitat available to them (Bancroft et al. 1994; Bancroft et al. 2002), and the results of this study establish that this is the most important environmental characteristic determining where the birds nest. Hydroperiod and fire are the two largest factors controlling the distribution of wetland vegetation (Gunderson 1994), so water management can play an important role in maximizing the extent of open slough habitat within the Everglades. Water depths should be maintained at a deep enough level to prevent most breeding sites from going dry, but shallow enough for wading birds to forage. Using these guidelines could make a major

PAGE 51

41 contribution to re-establishing historical breeding populations of wading birds in the Everglades.

PAGE 52

LITERATURE CITED Akaike, H. 1973. Information theory as an extension of the maximum likelihood principle. Pages 267-281 in B. N. Petrov and F. Csaki, editors. Second International Symposium on Information Theory. Akademiai Kiado, Budapest. Bancroft, G. T. 1989. Status and conservation of wading birds in the Everglades. American Birds 43:1258. Bancroft, G. T., D. E. Gawlik, and K. Rutchey. 2002. Distribution of wading birds relative to vegetation and water depths in the northern Everglades of Florida, USA. Waterbirds 25:263. Bancroft, G. T., J. C. Ogden, and B. W. Patty. 1988. Wading bird colony formation and turnover relative to rainfall in the Corkscrew Swamp area of Florida during 1982 through 1985. Wilson Bulletin 100:50. Bancroft, G. T., A. M. Strong, R. J. Sawicki, W. Hoffman, and S. D. Jewell. 1994. Relationships among wading bird foraging patterns, colony locations, and hydrology in the Everglades. Pages 615 in S. Davis and J. Ogden, editors. Everglades: the ecosystem and its restoration. St. Lucie Press, Delray Beach, Florida, USA. Baxter, G. S., and P. G. Fairweather. 1998. Does available foraging area, location or colony character control the size of multispecies egret colonies? Wildlife Research 25:23. Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Second edition. Springer-Verlag, New York, New York, USA. Butler, R. W. 1992. Great Blue Heron (Ardeas herodias). No. 25 in A. Poole and F. Gill, editors. The Birds of North America. The Academy of Natural Sciences, Philadelphia, Pennsylvania, USA; The American Ornithologists Union, Washington, D.C., USA. Custer, T. W., and R. G. Osborn. 1978. Feeding-site description of three heron species near Beaufort, North Carolina. Pages 355 in A. Sprunt IV, J. C. Ogden, and S. Winckler, editors. Wading Birds, Research Report No. 7. National Audubon Society, New York, New York, USA. 42

PAGE 53

43 Environmental Systems Research Institute, 2002. Arc/Info Workstation. Version 8.3. Redlands, California, USA. Fielding, A. H., and J. F. Bell. 1997. A review of methods for the assessment of prediction errors in conservation presence/absence models. Environmental Conservation 24:38. Fitz, H. C., J. Godin, B. Trimble, and N. Wang. 2004. Everglades Landscape Model documentation: ELM v2.2. South Florida Water Management District, West Palm Beach, Florida, USA. Fotheringham, A. S., C. Brunsdon, and M. Charlton. 2000. Quantitative geography: perspectives on spatial data analysis. Sage Publications, London, UK. Frederick, P. C. 1997. Tricolored heron (Egretta tricolor). No. 306 in A. Poole and F. Gill, editors. The Birds of North America. The Academy of Natural Sciences, Philadelphia, Pennsylvania, USA; The American Ornithologists Union, Washington, D.C., USA. Frederick, P. C. 2002. Wading birds in the marine environment. Pages 617 in E. A. Schreiber and J. Burger, editors. Biology of marine birds. CRC Press, Boca Raton, Florida, USA. Frederick, P. C., and M. W. Collopy. 1989a. Nesting success of five ciconiiform species in relation to water conditions in the Florida Everglades. Auk 106:625. Frederick, P. C., and M. W. Collopy. 1989b. The role of predation in determining reproductive success of colonially nesting wading birds in the Florida Everglades. Condor 91:860. Frederick, P. C., and J. C. Ogden. 2001. Pulsed breeding of long-legged wading birds and the importance of infrequent severe drought conditions in the Florida Everglades. Wetlands 21:484. Frederick, P.C., and M. G. Spalding. 1994. Factors affecting reproductive success of wading birds (Ciconiiformes) in the Everglades ecosystem. Pages 659 in S. Davis and J. Ogden, editors. Everglades: the ecosystem and its restoration. St. Lucie Press, Delray Beach, Florida, USA. Frederick, P.C., T. Towles, R. J. Sawicki, and G. T. Bancroft. 1996. Comparison of aerial and ground techniques for discovery and census of wading bird (Ciconiiformes) nesting colonies. Condor 98: 837. Gawlik, D. E. 2002. The effects of prey availability on the numerical response of wading birds. Ecological Monographs 72:329. Gibbs, J. P. 1991. Spatial relationships between nesting colonies and foraging areas of Great Blue Herons. Auk 108:764.

PAGE 54

44 Gibbs, J. P., and L. K. Kinkel. 1997. Determinants of the size and location of Great Blue Heron colonies. Colonial Waterbirds 20:1. Gibbs, J. P., S. Woodward, M. L. Hunter, and A. E. Hutchinson. 1987. Determinants of Great Blue Heron colony distribution in coastal Maine. Auk 104:38. Gunderson, L. H. 1994. Vegetation of the Everglades: determinants of community composition. Pages 323 in S. Davis and J. Ogden, editors. Everglades: the ecosystem and its restoration. St. Lucie Press, Delray Beach, Florida, USA. Hintze, J. 2001. NCSS and PASS. Number Cruncher Statistical Systems. Kaysville, Utah, USA. Hoffman, W., G. T. Bancroft, and R. J. Sawicki. 1994. Foraging habitat of wading birds in the Water Conservation Areas of the Everglades. Pages 585 in S. Davis and J. Ogden, editors. Everglades: the ecosystem and its restoration. St. Lucie Press, Delray Beach, Florida, USA. Kahl, M. P., Jr. 1964. Food ecology of the Wood Stork (Mycteria americana) in Florida. Ecological Monographs 34:97. Kushlan, J. A. 1976. Wading bird predation in a seasonally fluctuating pond. Auk 93:464. Kushlan, J. A. 1987. External threats and internal management: the hydrologic regulation of the Everglades, Florida, USA. Environmental Management 11:109. McCrimmon, D. A., Jr., J. C. Ogden, and G. T. Bancroft. 2001. Great Egret (Ardea alba). No. 570 in A. Poole and F. Gill, editors. The Birds of North America. The Academy of Natural Sciences, Philadelphia, Pennsylvania, USA; The American Ornithologists Union, Washington, D.C., USA. Ogden, J. C. 1994. A comparison of wading bird nesting colony dynamics (1931 and 1974) as an indication of ecosystem conditions in the southern Everglades. Pages 533 in S. Davis and J. Ogden, editors. Everglades: the ecosystem and its restoration. St. Lucie Press, Delray Beach, Florida, USA. Post, W. 1990. Nest survival in a large ibis-heron colony during a three-year decline to extinction. Colonial Waterbirds 13:50. Powell, G. V. N. 1987. Habitat use by wading birds in a subtropical estuary: implications of hydrography. Auk 104:740. Rodgers, J. A. 1987. On the antipredator advantages of coloniality: a word of caution. Wilson Bulletin 99:269. Rutchey, K., L. Vilchek, and M. Love, in review. Development of a vegetation map for Water Conservation Area 3.

PAGE 55

45 Smith, J. P. 1995. Foraging flights and habitat use of nesting wading birds (Ciconiiformes) at Lake Okeechobee, Florida. Colonial Waterbirds 18:139. Smith, J. P., and Collopy, M. W. 1995. Colony turnover, nest success and productivity, and causes of nest failure among wading birds (Ciconiiformes) at Lake Okeechobee, Florida (1989). Pages 287 in N. G. Aumen and R. G. Wetzel, editors. Ecological studies on the littoral and pelagic systems of Lake Okeechobee, Florida (USA). E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart, Germany. Smith, J. P., J. R. Richardson, and M. W. Collopy. 1995. Foraging habitat selection among wading birds (Ciconiiformes) at Lake Okeechobee, Florida, in relation to hydrology and vegetative cover. Pages 247 in N. G. Aumen and R. G. Wetzel, editors. Ecological studies on the littoral and pelagic systems of Lake Okeechobee, Florida (USA). E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart, Germany. Surdick, J. A., Jr. 1998. Biotic and abiotic indicators of foraging site selection and foraging success of four ciconiiform species in the freshwater Everglades of Florida. Thesis. University of Florida, Gainesville, Florida, USA. Taylor, R. J., and E. D. Michael. 1971. Predation on an inland heronry in eastern Texas. Wilson Bulletin 83:172.

PAGE 56

BIOGRAPHICAL SKETCH Matthew John Bokach completed his Bachelor of Science degree with a double major in biology and chemistry at Adrian College (Adrian, Michigan) in 1994. He fled the sciences for 2 years in a masters program in student affairs administration at Michigan State University, then returned to them as a U.S. Peace Corps Volunteer at Lukosi Government School in Hwange District, Zimbabwe. While teaching general science at Lukosi, Matthew fell in love with the southern African avifauna and spent an additional 2 years in Zimbabwe, mostly so he could continue watching them. He returned to the United States in 2000 and began a masters degree in interdisciplinary ecology at the University of Florida in 2002. He plans to move to the San Francisco Bay area, pursue a career in GIS application development, and eventually flee the sciences once again to pursue a career in popular music. 46


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101130_AAAAFA INGEST_TIME 2010-11-30T23:46:22Z PACKAGE UFE0009586_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 458757 DFID F20101130_AADBER ORIGIN DEPOSITOR PATH bokach_m_Page_35.jp2 GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
6b9ef16bb297c845447347a2859be03b
SHA-1
edc1ccb79b51ea83a870950ba809cf525e869564
1053954 F20101130_AADAXU bokach_m_Page_15.tif
2612c81e144a369563f5a5f7f6a21d1a
3a7c54558d6bb5bd3d62e9c629d93ee28792d042
35298 F20101130_AADBJO bokach_m_Page_37.QC.jpg
a8c7982b98c60ca29a2470a56b5bd0a4
1702e510d14b09332f28520d4178274768397b91
99025 F20101130_AADBES bokach_m_Page_04.jp2
71ee0b142ffd185dd5ea287ce444d6a4
feeeb090964b7d9c2762890838c4c0d141e2c0eb
8423998 F20101130_AADAXV bokach_m_Page_38.tif
42783f4945d0452609229b6546c7582d
e9626c563870f18af5b0dcaa18ad340eb1ebb8c7
6632 F20101130_AADBJP bokach_m_Page_29thm.jpg
240ba4fdfcb31f1865e3c4d22d2d38b1
73c0e4eee266cc8fb1deaa4bde2f0dff00a68c8b
24591 F20101130_AADBET bokach_m_Page_32.QC.jpg
0a38098e5a2c7ce49bd55a5b1790ced3
2b35bce12044aa5841f6288532d7e8571a7a0fcc
1824 F20101130_AADAXW bokach_m_Page_07.txt
e7f512af96b2b550425fc399dbd399ca
404058e711bbd1d4cedda4198189abbad5e0d540
10741 F20101130_AADBJQ bokach_m_Page_03.QC.jpg
b7843db819f70035ba5b752745417cbf
accb145f9cefe906537b0661b0857401d9c81a7a
F20101130_AADBEU bokach_m_Page_28.tif
603ae65bc67c300954e2299c2c2ec5f3
f239f21388384809f3af5f91d8fb093f1e86287e
492674 F20101130_AADAXX bokach_m_Page_26.jp2
a06a8a003fcfdd91d4fd522b5e97364a
254ca8d26caa6f1416525adf7b371e370d3a0836
23640 F20101130_AADBJR bokach_m_Page_16thm.jpg
0a6ed7d0684c7c549364d419868e4935
bcf164350db3a5be2e9ec4e59c8b8045d12ece05
15202 F20101130_AADBEV bokach_m_Page_03.jp2
037aa65fe2e5fd9b5694f86c57e64a2b
b678afc7d4513ac34ee7e68ffa38bdfc08d1bc70
17228 F20101130_AADAXY bokach_m_Page_38thm.jpg
17d98ed766fa259c4dceabf7c95a0dc2
9281dd076ce79385c85ab7de4b30da113ae953e6
35814 F20101130_AADBJS bokach_m_Page_45.QC.jpg
1f75ab5d973b751b2823992c3fba2c0c
8922913ecbabaa9301e06a089e5169a4b1a8e239
42996 F20101130_AADBEW bokach_m_Page_33.QC.jpg
8b3859199fb54bc295762c094a2631c3
8b43715a42749aeb111d531bfc995d3d05e3c317
86001 F20101130_AADAXZ bokach_m_Page_24.QC.jpg
75fd050ca278c1cf909b197ea8b36342
2bcf68f578972e64dc661d3b58a9525eabd464e1
13059 F20101130_AADBJT bokach_m_Page_13thm.jpg
75594285cd3bc2aa1920c547814e7409
2947513e0e29d6b19eaefdc84e2635ca1a6208f4
18591 F20101130_AADBEX bokach_m_Page_44.pro
8c3fea76024fc1aea8ff3659bce631b8
369ea91c05dfdcb6c6600871d1c28bf5f3b79ff8
3773 F20101130_AADBJU bokach_m_Page_51thm.jpg
507fd1e5c76ddeefc47244bd4dab11e1
7393f06a0082cf517916bc87853d0948853ddf36
F20101130_AADBEY bokach_m_Page_02.tif
5ea7f7b955c85c7073c9064160d4d021
76605d8bb6dc87fdaabb26882243bcc01498b03a
81506 F20101130_AADBJV bokach_m_Page_49.QC.jpg
8d21b131a14db9b41d320ac31eaf7e65
3b3b6bdd9ae5f7d219e53e427298c0253637f795
146059 F20101130_AADBCA bokach_m_Page_31.jpg
769b05e913e2122a532ce7442cef5de3
3942473b753092040ad56411b0226097adf648cc
131814 F20101130_AADBEZ bokach_m_Page_33.jpg
9df508fd916442292451a3846e8a5684
d0f850efdab9750526a71987a0ef7cd126085ab7
43035 F20101130_AADBJW bokach_m_Page_06thm.jpg
bc8febff4a2a1bcf961d9f20945bc289
73b0cc4789618abfe76755e25a1fb0dcd720d7f3
48805 F20101130_AADBCB bokach_m_Page_50.pro
f582988b8467ee8594b903c027520a09
c1ad3f92a0c601ce43103d8909cc1d4134b95f25
24924 F20101130_AADBJX bokach_m_Page_19thm.jpg
9dba77d102cc27ca1ceab627fb15d3f7
7edb390be7ae167681c49f2ed8bcdd49fa91da8c
F20101130_AADBCC bokach_m_Page_19.tif
0c488db622886628a9f8f54b41934340
2484c3a29c94b1228af726fc71c181ec05cc450c
25271604 F20101130_AADBHA bokach_m_Page_35.tif
0bc00376a0867d4366308db2241708ed
1fdf744b394eedf8b3257afa29df85d056df8aa7
33312 F20101130_AADBJY bokach_m_Page_47.QC.jpg
8e19d8eb82687ce3fc4f079e24b95e37
01ca5c5a235371bf41af3d0daa2cc8cb921c264c
1825 F20101130_AADBCD bokach_m_Page_04.txt
0989df1f30cb7ac37e7aa4001ee28fb3
69a2425e7fa16527f5410b8aea529740242ab0f0
F20101130_AADBHB bokach_m_Page_39.tif
1184e8c23b165ba112ebc23c13532756
c43196ca62950875a8605f65225768995b152734
79196 F20101130_AADBJZ bokach_m_Page_28.QC.jpg
0fd9aff055e810e72a9478d97a2b773d
7599347a1af631ed47c42ab1cbf21b950603746a
2147 F20101130_AADBCE bokach_m_Page_52.txt
0f4f768f7d7ff61943f54d2231cbe40d
8bfe0f9ceb7afa941ef1dc8059d2e0e47083e077
F20101130_AADBHC bokach_m_Page_40.tif
3dbbbdb4857f5a9c01e187b3f08ac71d
825382de024cc56e68ebb2243e3a4f631ec7d33b
F20101130_AADBHD bokach_m_Page_41.tif
430f9a16ac2fb66aea2e4921a03894c8
d57787488e262f76915de146e077b2b4baf1e2ba
258698 F20101130_AADBCF bokach_m_Page_41.jp2
fc52df14c88ce3c5addda5a395bd160c
dde8b854ea639a5e1c74e388f719225b25087847
F20101130_AADBHE bokach_m_Page_45.tif
614e8e3ed0eb5368939a7016c908bad5
0ab01cf6f3b6f4d833664a8782427238040bdeb8
F20101130_AADBCG bokach_m_Page_01.tif
e8fab423dd539ee0277eb7467b7647dc
7872bcd1642dcc1477bc300f17b9068fb0ab06eb
73130 F20101130_AADAVK bokach_m_Page_04.QC.jpg
49bd154b6fd6da337feb4a151afc94eb
810e8dcb1a1a9743a3b0393813ee28cd74c2d9eb
F20101130_AADBHF bokach_m_Page_47.tif
cdb04a517068411eb4df00be7cf2e448
37d1df62645d585f8763d4828bd0b4400fbfbc0c
23878 F20101130_AADBCH bokach_m_Page_56.pro
33e53f77e96102f3e3506ca7efead85a
a83728118c00cc8238b1c79ed6db761b250187a3
F20101130_AADBHG bokach_m_Page_49.tif
c29a90de56734e63731f1c0fa34226dc
3644556363952ae15fdf09368abec7946e1bc8b3
49111 F20101130_AADBCI bokach_m_Page_18.pro
841f94d764b6e4c8aa7c0bdc24a066a6
820dae5d154a9477cd752aeb8a5ccb287e0d5e1f
290957 F20101130_AADAVL bokach_m_Page_25.jpg
5ad42d0387e16e83009168d4567e5f77
51f1579207f835ef94f2453e8032977fc0adba4f
F20101130_AADBHH bokach_m_Page_55.tif
ba5ab1f5f5a4a1176f244938ef9a6793
3ac1ade2b7319f8441b8a148850925f2c9fedb83
69320 F20101130_AADBCJ bokach_m_Page_52.QC.jpg
e6b3b64e8e2175d67e0ba32d6b556379
f1d81384540be2e2e16d9a156aca99236fb23664
19834 F20101130_AADAVM bokach_m_Page_01.QC.jpg
41055b1383d50ba48bc9ba4c2cbed500
91377693fae5d94a88100b873115528c5f97ed32
F20101130_AADBHI bokach_m_Page_56.tif
b246573b74207df182a7aac78f225f0c
b454303b16f9d9f56e352ca298cc05be4fb18806
195066 F20101130_AADBCK bokach_m_Page_04.jpg
939ee7faf4e656a817cd9b054843c6b8
86ba3c899d904a982ea962a64b91c1be7fefd6d4
125288 F20101130_AADAVN bokach_m_Page_36.jpg
311c840605ce8136d6d5fa65a71d4efa
69f2df41ef8486afa931c9463359b15dc5ab5179
1278 F20101130_AADBHJ bokach_m_Page_02.pro
b96e302e06e4cf819b5c4c98e40dbeba
38788eb28360afc4be7f225c5b7ee5a5180353d4
8687 F20101130_AADBCL bokach_m_Page_01.pro
94c90520d4adb97c7d84b368a137e5d3
1c2536458206d67017fa9ca31efec225b85cd1f8
52985 F20101130_AADAVO bokach_m_Page_31.pro
1bb6ee6fa91c24352a235880fd6d06d5
82a28abfa92e6163db679f7154c98c5164682aa0
F20101130_AADBCM bokach_m_Page_53.tif
e05de2125f4a7a6afd518e6cf751f70f
10dae82f7a497573b5383070d4cd61bea7e77e1c
110009 F20101130_AADAVP bokach_m_Page_26.jpg
dde89b576382853af1522c637b28a8bc
56cf665fec0d503ad2305f5940fc72d54e17a30b
45632 F20101130_AADBHK bokach_m_Page_04.pro
40511af8a8ac69245c9f3638a56a6d8b
a2e6bedb88af582e55692076a16a652029504211
1420 F20101130_AADBCN bokach_m_Page_55.txt
6b489bb80ebcbd01c241c235a0de5a99
4e8ac6bb5632fee6deedf8530793b90439ee577d
F20101130_AADAVQ bokach_m_Page_12.tif
13a4f05e7ac25ffcd562d01ef0ed6c95
d6801bf0946e1c075c07abeb93655bd45adbfc29
79034 F20101130_AADBHL bokach_m_Page_06.pro
0553a7beb0f2393b4df439124816461c
bdaf85f57f3f16e662acfe93b6fcbcf09e3fa547
265126 F20101130_AADBCO bokach_m_Page_37.jp2
132b8a1b341125694f6c4637a4676ab4
53f5e7d449d8221df438206755078a0ba248283c
62378 F20101130_AADAVR bokach_m_Page_53.pro
53b393f63811a21a19513a6639d9c3da
905924645d5f201b4ee0761e0cb955ae9f30781b
44113 F20101130_AADBHM bokach_m_Page_07.pro
c9622b3cbf25a6089e018f0228559125
90710350b6bd88da304ae6d1d63ce1eefb9e0226
201796 F20101130_AADBCP bokach_m_Page_07.jpg
6bc45d9310031c595b8960bfbe5b93e0
ce93bcaafef605258ea83847fca09c1ac8993991
F20101130_AADAVS bokach_m_Page_24.tif
95de05e23e56fb0e6485d2dc33163579
fe3cdccd5835274427742fc958f42a8c052ad0dc
48008 F20101130_AADBHN bokach_m_Page_12.pro
c9e78d16f1f85d08e3ba809a83be701f
3a1df6b1e39ba1fdd4556c6a99e2cca5cf2bb17d
83903 F20101130_AADBCQ bokach_m_Page_53.QC.jpg
9d27dcf9937f3b949a5a1ac1b2a67802
fba09c220bdcb8f854772faac09446ced1f6a413
2516 F20101130_AADAVT bokach_m_Page_54.txt
a93b0e78392d8188dc54ab595ac1d555
a6830be96b75270eee20365eca975d48c6a05f92
38516 F20101130_AADBHO bokach_m_Page_15.pro
68ff3733e9367063fcd69bc7ab71706a
d769cbbf87e6f6f5ebad9741d82da7687cf5153c
24790 F20101130_AADBCR bokach_m_Page_49thm.jpg
f06b5055d92dde499a94dad3d3be9a95
823ffca700459c158a0afbc1777916646dea80d8
38390 F20101130_AADAVU bokach_m_Page_08thm.jpg
d55e755cf0554456fd606d33e8676ff9
5778454b7b9e0c7dea5dfb3d7d4f0ab0050c1e52
45278 F20101130_AADBHP bokach_m_Page_17.pro
88aca9af024d689520f61c14a28dfb63
882d506968094caa7bcb9b33aa751d2add497604
44174 F20101130_AADBCS bokach_m_Page_24thm.jpg
f89e517324c13900523cadeb62d0e954
cb760000aada64ef53e144f7553311fc43d34831
1054428 F20101130_AADAVV bokach_m_Page_32.tif
eca19aa63d6add25fee7fb45855909bb
3e8e77c2ff3dd017c7139de2d3845b9630b4ad55
51277 F20101130_AADBHQ bokach_m_Page_19.pro
6643b65ab2f77643646e6cab4a767e3a
55eb11033fb92768c0cd56dbfe4413bca516522d
21514 F20101130_AADBCT bokach_m_Page_36thm.jpg
736bfbb78c16147ef37e01aedc39273d
38597e41f58312a9250f436390714aa27a59632f
2634 F20101130_AADAVW bokach_m_Page_30.txt
998336a0a7462b0717a9fd386bf05d17
8a74c8dc82c5a86340eb12ae5111ed87edbae924
44423 F20101130_AADBHR bokach_m_Page_21.pro
8a3d21e61398a6213eee434063013cb0
181b6b2be3c146d0c6dd3d36fb0f7b1c47f237cb
111102 F20101130_AADBCU bokach_m_Page_49.jp2
5b48ead150f3f8a9c7ed170e56dbbf76
282053d53abdfde3110e101c80ffb2d9a82fbbde
17820 F20101130_AADAVX bokach_m_Page_39thm.jpg
d5fff3e5ae3514cc535e2fdadbf7fc45
a48f0e32d28f22f3715377232682c63217fbd0a8
4182 F20101130_AADBHS bokach_m_Page_23.pro
f8061975ce669a98f1541fbfab2acf36
b9e90419efc86f3a4d96b12081a87f5db5ed97d1
679 F20101130_AADBCV bokach_m_Page_26.txt
6e3460b4b2962aa4d195f061694aea14
06fdbef0abfef7266440367cb9c0b85b2a9d78f9
924 F20101130_AADAVY bokach_m_Page_25.txt
1af3d1050c863b557c81bb11569f27ba
e8311de8fafcb68ccaa1de95308ff6769d308806
9678 F20101130_AADBHT bokach_m_Page_24.pro
70c2549667f1f0a264ec75075ca65e29
b4e782710498ac78288611ad442fe7cd0b367fed
F20101130_AADBCW bokach_m_Page_30.tif
82f5dd67c774ca56aaad63199055d19a
60e51b7a9400eafee0b8d9d7ebdc3f8b0394de28
19913 F20101130_AADBHU bokach_m_Page_25.pro
943e2603b2896382690dce18b44f1b72
29eef1c45a1880b07c7dfd7246177db205987ed9
28412 F20101130_AADBCX bokach_m_Page_29.jp2
c525cfe5eccc06a261ed3797773ca3d3
af63daa676a45fe2bf42c1c950d18b4eea0e0aac
214224 F20101130_AADAVZ bokach_m_Page_52.jpg
6d0c81743a184782211d2ee460f6a7b8
1941d61f4a7fd288d296d2fd76ef54253638c3b1
16269 F20101130_AADBHV bokach_m_Page_26.pro
ca7b3f3333c85afedc715527e6189bd2
74dccd2523aeeece8f58c3a305209b7aa56a0ac0
30395 F20101130_AADBAA bokach_m_Page_03.jpg
1439251844231feefe19cf02c107e4ac
1b86201ac0f5982a77a4d8593773e8264177b6cd
52511 F20101130_AADBCY bokach_m_Page_52.pro
fbc2d9ab0d610c60d00e9e9014d7b2f2
49316801c73a67bfe0b6dd541af745f00ceebdd0
49486 F20101130_AADBHW bokach_m_Page_28.pro
89f01a15a976ddd97f764f0a50429836
c70affa53d4e20a01394111d3e0259ab8d20deaf
22106 F20101130_AADAYA bokach_m_Page_10.pro
e2fa3903bc376611cb60e3254f7a5b33
1f6143bc60e12cee257b205d13f021cc9669096f
46641 F20101130_AADBAB bokach_m_Page_14.pro
8b08d6af48146f60700f251e5ab3bc9a
4a52483e4d2887a8230e72078b6f0358b44215ec
596211 F20101130_AADBCZ bokach_m_Page_34.jp2
052d74b62dfa5c18c45fa2dd765ae71e
741d5ed7cb260e6851f9b53b248288e26bef99c3
12669 F20101130_AADBHX bokach_m_Page_29.pro
1c78b5497f6aa0ab2516de48473cf79f
1e177785ba9b90b5e606bb26c4d20da71497805e
24502 F20101130_AADAYB bokach_m_Page_13.pro
75fc5b1f764d2cf689d38ad0b80953ac
5182364f8f969bc91201a1d31b62a88e03d6de51
32238 F20101130_AADBAC bokach_m_Page_43.QC.jpg
44a8f49790e6425eac05cd756906626f
f68935b744f2b19b21b3990939556832e54883d0
46593 F20101130_AADBHY bokach_m_Page_30.pro
dec16ce6121395d126ffbbf5731ecd30
6ee82070640c75a3f2bbd7e44443da368f6f279d
43892 F20101130_AADAYC bokach_m_Page_27.pro
3e5722817debe83019e7762b0a129e8e
441e9fa8bc9c891000a1f37ba82c8d95d59c0f6c
338 F20101130_AADBFA bokach_m_Page_43.txt
8da9761ffc00b70e07bc96d95c71eaf2
84f58a472ad465c9c56c6e8f5bdbd4b8bf62b932
21591 F20101130_AADBHZ bokach_m_Page_32.pro
5ad2ab54dc07341da65d26f955f22a65
b8bcf5fe3c18eda1a0132bf6bed5ac7724360cfc
188800 F20101130_AADAYD bokach_m_Page_30.jpg
92d6832d687e188471ac90a339dcdd05
a18f38070f3b57edc055571eacd2671817c339d0
22721 F20101130_AADBAD bokach_m_Page_04thm.jpg
f5a31cb9c437b0f610ed45ee0136cc2c
3986fa37f8fc0f4a921cb2bcf15e84034d683f0d
81570 F20101130_AADBFB bokach_m_Page_37.jpg
0a5c36f89f57d9e2b0529763c7c2b007
c13923365c4daf29237ee67ebf0633daa28d25ed
112976 F20101130_AADAYE bokach_m_Page_20.jp2
57ef7d6bc83c72a5668b9c54c3ae16ca
0475bcc21851466613e9782061310f6d942c5c98
171289 F20101130_AADBAE bokach_m_Page_15.jpg
1e2408b47fe857ddae54128c9f0cbb7f
8413fffbda48e506d9fd9be2f21c451d4aa163ca
162 F20101130_AADBFC bokach_m_Page_51.txt
bc9109154b798852dea6e646123ad33a
da26c8458a770bdaaa136aa93a9c803a3acfc5de
882 F20101130_AADAYF bokach_m_Page_40.txt
3ee9c609bf4f33ac317b3bff4a224ba6
f0e06f2b01169f5c3ac2fe6ad9e4a0f9a6362613
F20101130_AADBAF bokach_m_Page_33.tif
97bc9afd985e03d32a1145bba05f5924
4f56d576bc9c7e135a6b48d17af4955b9c149eb2
66061 F20101130_AADBKA bokach_m_Page_48.QC.jpg
b6be8d89f73d406c647d631eb7c8d2b2
01f70a322411869a3893683a9bad1e50240e378e
203561 F20101130_AADBFD bokach_m_Page_50.jpg
0631da580cba8a4803456052f6984d18
0cd0f06fe8751d1c9da9837388d4cd039d5aff93
17483 F20101130_AADAYG bokach_m_Page_43thm.jpg
8611687bf521dc8fad5ce490c53dc659
efc28ea41c2677ebbce94a8c5e81851c18fe6822
73414 F20101130_AADBAG bokach_m_Page_55.jp2
cb6c2032d679ebfb3a47b4300a869ebb
a73b8745e546cf1bf9710bcbf53217083e5ec788
19392 F20101130_AADBKB bokach_m_Page_29.QC.jpg
d6ad8fdf3e62bbc89f28f816968c37c6
6755fff7757af99e355cc03bede92e510af06c53
6196 F20101130_AADBFE bokach_m_Page_02.QC.jpg
7a1b0b506fb70f382cbe76b082873fdf
0e530928f185e6253e9091ec5b4233177785897e
25539 F20101130_AADBAH bokach_m_Page_20thm.jpg
04bbae61d6a093560add8f8ea94c643e
bcf59562e38d3b0b775928b6ce686ed2d96f00af
17972 F20101130_AADBKC bokach_m_Page_46thm.jpg
2c7f24b71c0271821aca3ac4fe2bf858
c6dc5f2e2a5bd6ceef7e2600e103df3a8b4092cd
F20101130_AADBFF bokach_m_Page_27.tif
354929fdb37640debe2a1f93cdfc1200
5df3e48b2a99f83021ed8c8f220faf7f5b2026a2
119 F20101130_AADAYH bokach_m_Page_02.txt
39624aee752718f203e7133a83b2a904
5f4bb24bbd096330ad152f7314e82c473f07f048
56271 F20101130_AADBAI bokach_m_Page_13.jp2
23e96141ad040c3ddfddfd887f908ccd
bd1caae60cd060e1773b20b39ec951ca18546fbe
61926 F20101130_AADBKD bokach_m_Page_15.QC.jpg
74108a19d67e106b7f7c689b37bb3bd4
a990ea3396676eb904196485509d66f2616a8e92
55826 F20101130_AADBFG bokach_m_Page_09.QC.jpg
2603d50a6258fd2d9e87ba019339c996
f1963497ded288781526d4046902a0d67106c704
F20101130_AADAYI bokach_m_Page_51.tif
7725e5825c24508c5c950db94189f055
36763d5fc7f3a30383f788f3927d26113ddb0d69
1051952 F20101130_AADBAJ bokach_m_Page_24.jp2
a259dc34f42c4b0fc02d2796c9abc7ee
fa35298e8f9324d72d5941cf78d7181a503091af
18145 F20101130_AADBKE bokach_m_Page_30thm.jpg
a6dd731a21edf68745b4223185f4d86f
6753c20a03dadfea0d416f608ceec21e79dc03b2
69247 F20101130_AADBFH bokach_m_Page_35.QC.jpg
a0df77effae07d9e7d4450b9a1c92d34
fd3ea728f64a0e159dd7c04e823c667214a3ceeb
995 F20101130_AADAYJ bokach_m_Page_56.txt
5e6e3df19889d41972e8ceb9e85f01e1
f117e49334513831b92c870ef9e0fd2adac716e3
99751 F20101130_AADBAK bokach_m_Page_25.QC.jpg
dcff933181e56e3114aa3dc5135fe964
efa202219988ad46797adbb75010d4b6ad49bd3f
86496 F20101130_AADBKF UFE0009586_00001.xml FULL
0fea4d675a0bec96e68c8db205d014da
9af4d5f0b30807d83a9246a1db859dfbf2309a51
215096 F20101130_AADAYK bokach_m_Page_20.jpg
b1a16ac3df52cd820aeacf9867f2447e
8b557188e1f2252f4034292a7bcd3faf77f89e17
6885 F20101130_AADBAL bokach_m_Page_41.pro
f829b2351e4819f6a63f5daf86216f28
19875b6d3b000a4bc31c96be9971cc374396179a
7533 F20101130_AADBKG bokach_m_Page_01thm.jpg
24b76d5d2b0de89b1d8c526e76ffb952
dd638dc358ef2894ea86fa873ab69d99130bfbba
66990 F20101130_AADBFI UFE0009586_00001.mets
2b53c3bb37da904196878394f9a9c236
1ac48ad78378b9f9836388a637a510f13f61f518
206226 F20101130_AADAYL bokach_m_Page_12.jpg
59776bc0f9a67545fa386916e8aa5d66
7b61d714ee3b02fed009ea9bc85096e212e6ad97
F20101130_AADBAM bokach_m_Page_46.tif
9075a0c116293d40014be96776d4ec6b
7cd5d62c066fccde9c7d560bf86e00003593c5fb
25367 F20101130_AADBKH bokach_m_Page_05.QC.jpg
bdd63d6b04955ecb309e22c5e4352820
fea25a975164eae0ed68fa4d6c085c33c707cc83
21576 F20101130_AADAYM bokach_m_Page_27thm.jpg
7182bef9a8101d1655f40875d0cf0da7
48f792beebc5a9889634f2159dbb68cf5c58392c
231908 F20101130_AADBAN bokach_m_Page_06.jpg
e8e05be85a9dd23fd53299dcc5bfe7d6
7d0ce479fa29fa6c2252c66fb955d24742e12536
24473 F20101130_AADBKI bokach_m_Page_12thm.jpg
d9ae273ed7ff49c49dc44de4cfec5d01
79f26b942a054a1fd07675f728a1ed67a951c86c
36673 F20101130_AADAYN bokach_m_Page_09.pro
267825b808296a6966365d1d0e24dd81
9096d2b0671ed05cecf3cabe22a722ebd3cfbb95
19307 F20101130_AADBKJ bokach_m_Page_15thm.jpg
4254572c33ab73f4ea1c549944870fb0
a33cc32db2bb8b89a6eac4fc1dfc794ea5d619f5
63499 F20101130_AADBFL bokach_m_Page_05.jpg
6c4805e4f462e5445e38026b7d388e60
1272dce98e0567c86be1638a740b31f22a1711d8
F20101130_AADAYO bokach_m_Page_54.tif
00bf2eef0fffa450741c2153e1580dd8
ad3bf2f59dc1016692a5a9a1b01ded7c4964a94e
12263 F20101130_AADBAO bokach_m_Page_23.jp2
3915156f067f78d94279fb5ebfade9e6
cde37c24c68c106992a82ff131d61ecfeb8adfb1
78258 F20101130_AADBKK bokach_m_Page_19.QC.jpg
7f6786f4f03dea55763e916b954376c7
cbf9e7022a5a0a4a13f633912b7d30569d74185a
153700 F20101130_AADBFM bokach_m_Page_08.jpg
933c88ded401c93b4402bc9a1cbaddab
d36a10b4d0e6df988bb238f6ace440aa423ae23a
8749 F20101130_AADAYP bokach_m_Page_32thm.jpg
0407c66895cbabd62a027d72fe2cd8d0
9b1c6949f1c7e4fa6712fa14ce4d591c8c3554fc
36033 F20101130_AADBAP bokach_m_Page_10.QC.jpg
75e9c9a0696db1ab93203537cfa8d985
4f9d16b1d7c2c65ca3af49c9a2b090daa11264f9
22099 F20101130_AADBKL bokach_m_Page_21thm.jpg
1be8ca719bf1d9f19f7f86195ddd5509
f7c10fefc9622e228b07cf179ab70e99a558d23c
205520 F20101130_AADBFN bokach_m_Page_16.jpg
39fee016d162ce492a3c2d195f0821d4
8f430328f78024904b37a179409e115beb9cf7f2
24916 F20101130_AADAYQ bokach_m_Page_01.jp2
5da0fa0722602c8eeac358620c875957
3ed93ec89fdf7b36960b229bc6159fc13012024b
F20101130_AADBAQ bokach_m_Page_31.tif
125a89ecf6d4418a96538464cb4242a3
21cde1215e11e38644f53ef58f57864ea282d3e9
24962 F20101130_AADBKM bokach_m_Page_28thm.jpg
a865ad4537b0787ba9b31875ca77cb9f
3cbbd1efb80753535a95ed55d06988c802bbc09e
190257 F20101130_AADBFO bokach_m_Page_17.jpg
9f500e3f30e41ec900a998b77e8683b4
6ccfe8f411a42a00fb793d9beafa19e451d305c2
F20101130_AADAYR bokach_m_Page_50.tif
7e0ccff50a52d761199f32b1f5312f21
2821d2781dca485b56faf41e260379e75f9a01b2
1952 F20101130_AADBAR bokach_m_Page_50.txt
e429fafece6ec3ec639968396f37a032
548bcb1ef2d5f16db80ec6b3e04f62d968a6b999
204258 F20101130_AADBFP bokach_m_Page_18.jpg
b9468a5c8c19f70cdfb1853d30e4685a
b8943b039c01384b485c5ebf949a1f048a5d116e
65824 F20101130_AADAYS bokach_m_Page_33.jp2
3c1007dfda0c2761afcb7080ab7954a0
e36435b5130f4ac532b6a5c6407ffdffd54e0da2
2935 F20101130_AADBAS bokach_m_Page_51.pro
2862fddfbc63c7041187e7d2817ca1e1
72c9a5d3cd055d669752b3f1b5e48a04ee8f64a5
35740 F20101130_AADBKN bokach_m_Page_39.QC.jpg
4064a0c8dc296c32f7969b7a772cc1a7
6daf0af0656098398529230cc0c097d09de393df
212952 F20101130_AADBFQ bokach_m_Page_19.jpg
94d88758fda9fc35b4494c361fcfb537
3fbbdec5f36472f1b9805a5714e1cf8bb480ec4d
108851 F20101130_AADAYT bokach_m_Page_13.jpg
5bde30cdba381a719f637f9cd5f18324
1188e7e1240df8bbc4de4b0c0b8be0a3d5d39fb0
110279 F20101130_AADBAT bokach_m_Page_52.jp2
60808f7c1321157dd30497150ff38be0
022f5e1a3434509716444898bdf295088deec4c4
49024 F20101130_AADBKO bokach_m_Page_40.QC.jpg
0b9af6046460d3cae172826ba71b9c66
86bea850b48bfea01426a4d22f9e314dfdec2ed8
170512 F20101130_AADBFR bokach_m_Page_22.jpg
58cfaba6589c1c5072609912389e5ed6
808b0cf37c67351c9fe1d1b4e0bd1641982f18eb
27081 F20101130_AADAYU bokach_m_Page_33.pro
9fdc459c5cf6fc14ebc26d22441f9c84
ca80b2014d350f0c33e93ccb9ff528978ab1b56c
13714 F20101130_AADBAU bokach_m_Page_05.pro
b7733f9d5ebaaa38c036df158b7a3160
6301d96b3dea2096c1d3dbd0e55aeb61b5ca47dd
8154 F20101130_AADBKP bokach_m_Page_51.QC.jpg
8cd80ddc5067c0590986ad16f890357f
2ec7164ab58678e7a3a39127ca7157a1c6c31963
184781 F20101130_AADBFS bokach_m_Page_27.jpg
032574ac9a76803e89101c550bd8f8c9
4c571e7e0527f6604b282af09d196f1f105a8681
44785 F20101130_AADAYV bokach_m_Page_11.pro
0b66bb6b5be655884d9293dae5036478
00819e09a659c2b8abf0ef8faa4e541c3f8e762b
1673657 F20101130_AADBAV bokach_m.pdf
d68dcbf6392575baeab8f16896a83393
a81e66f2709cf6c9f19f72f45e4463e02cf6a609
46142 F20101130_AADBKQ bokach_m_Page_55.QC.jpg
bd408979e633ea10defdbd4dea65e2c2
861143e2c6c3a22bd9c275316ba1f01e7590bba5
164441 F20101130_AADBFT bokach_m_Page_34.jpg
b27c5cd8ab5b8215d19a3b03184206de
3146c9ab2b1023f42885da71366fc60c5653f6a5
5576 F20101130_AADAYW bokach_m_Page_42.pro
bee39535287d6265c7a4808acdb4ff33
bb4c290aa8b7e9319155dca1051ec5015494dab2
423 F20101130_AADBAW bokach_m_Page_24.txt
b98328f0ba1bdb5638b35fd130cf0b87
f3d3d6a1046b6e5102a2fbfc9165c9d7f078d526
77944 F20101130_AADBFU bokach_m_Page_38.jpg
ca1a90065af3898b089afd599baaa2c8
77fc14bb9350142faa6aecc8ca8fa1455b711062
8318 F20101130_AADAYX bokach_m_Page_05thm.jpg
6cca37bb80f9cb684b48895f7ac513bf
e85d57880e134b8dc21d94a7d5997b2b62ae5285
7431 F20101130_AADBAX bokach_m_Page_38.pro
163c6d4aa415779c768ade5cfef0a612
79e3e8f59cddd4cea7d28d011c8e82b0c5f22f4d
77000 F20101130_AADBFV bokach_m_Page_45.jpg
d93358914f783b5f2f6f17f684b11735
c28feb3b895cf84218da50454fa8eb422de11164
2497 F20101130_AADAYY bokach_m_Page_31.txt
3ab8f13bc9382359d49effc28eb01667
eeab6619a0f7c3e1230e4ac5daced374512a044a
F20101130_AADBAY bokach_m_Page_17.tif
c5b2ea65cf2b842d01835c55fb24a4d3
9a951767c2a9d255219e341a571647f16ace3513
80778 F20101130_AADBFW bokach_m_Page_46.jpg
f5c1578b43f16679c28ec6e65820e244
53828a78f093f85d76bcf1f025263b8face72cfd
F20101130_AADAWA bokach_m_Page_43.tif
1cc466d94442625d6c11bd87ea7a61d3
f2e4b934eb9c41212a30e1f37d28ec677df061ad
29352 F20101130_AADAYZ bokach_m_Page_08.pro
bd8a90f174d46566a33868dcfca2f8c6
56d29f765e9cfcb0a607b039547946965c6533dc
500718 F20101130_AADBAZ bokach_m_Page_40.jp2
7a8552a6e9c07e6b5883a1bcf45d7597
f3c61c24642754d76244bba17e1e693543d447b1
75833 F20101130_AADBFX bokach_m_Page_47.jpg
4789d276084540441b88b549cfebe253
78b2b6ea26a926ae88b5d13c88ef9af9a03cb65f
52887 F20101130_AADAWB bokach_m_Page_32.jp2
0092899a99afb5d40891b8179b1f5226
0245d710f553014c3504e53861ab97530be51363
188310 F20101130_AADBDA bokach_m_Page_11.jpg
206e4a14c93245ebdb01fda5d0769889
255ed96c3d80132a56798fc1b247c35b24cfb228
179798 F20101130_AADBFY bokach_m_Page_48.jpg
d6603598c35376ddc970935bf34780c3
880cdaeb2e1a912f82748a240ebdb7c19c64f45e
1051953 F20101130_AADAWC bokach_m_Page_25.jp2
5b12a8a277318de8906e56d895e2ede5
95bba19b27e03c5c6355c5da10933f12e8cea9db
262998 F20101130_AADBDB bokach_m_Page_54.jpg
35be55909c83a47d30725eb192f1811e
fa34f4e1038b82a1ae20c3f3767adc04484623c8
147757 F20101130_AADBFZ bokach_m_Page_55.jpg
7fcc8740d5fb542401d1da2a366a12ee
7da2a56233e8ebdbad35e53b1b1942b0f36f253f
1979 F20101130_AADAWD bokach_m_Page_28.txt
55ccc1e4367ec2b4ab7c6b5608ceb99d
7780c93efe0f3ca44729d780cb600ed6953ea3c6
F20101130_AADBDC bokach_m_Page_09.tif
720077a7c00560529b71ebe91bc39e05
ca66e3d08bab8ea01340fcc1c650d636697882e8
15139 F20101130_AADAWE bokach_m_Page_31thm.jpg
6f79300c7a3f402a919a492ab329ad7a
4bc8f3fbea9ed13cd0881ab42edb06c0a479ade6
69819 F20101130_AADBDD bokach_m_Page_21.QC.jpg
d6ef3505971aa69a47ad6c219a148b13
fbe5c2ca8021384af9fb51f64f27e01e55d5f075
483 F20101130_AADAWF bokach_m_Page_38.txt
c7db5da913d4a3b012928db24f8e1d32
ab9244095ff72d487205298aaca187eeaf100487
4384 F20101130_AADBIA bokach_m_Page_35.pro
1e0a57655bee9d8460395dcf73f32753
b985e65c9e0c95e7a51352cdcacae24eabd240e9
49040 F20101130_AADBDE bokach_m_Page_31.QC.jpg
d6004e0ab3f8d9d2ac5aac19b7288d9b
082fd7ece2367b492fc65f10d098218d8550320d
21652 F20101130_AADAWG bokach_m_Page_11thm.jpg
18b0cead1643cf42aebf4f15183df012
54a8b2d103314ccc4e99323bc9b65978cb90c213
7842 F20101130_AADBIB bokach_m_Page_37.pro
f4175a6136e8f59995615d147874fc71
69c443142cb5b37bce16451fffcdd66ffbcf431d
21427 F20101130_AADBDF bokach_m_Page_52thm.jpg
241f2c5206c6b432ee09494b812a6072
01f968bcd45703ad7ca3183675ba989d7e225ff2
88872 F20101130_AADAWH bokach_m_Page_30.jp2
6966a8ac8016394b3f03b6c505e89999
fe51079de3b6e12e70f1555dbf03e873f5e44494
8797 F20101130_AADBIC bokach_m_Page_46.pro
430dfd4252536390371a430fea2af3ec
b14dc329eabbdb398438f93445f3208585b44a5a
472 F20101130_AADAWI bokach_m_Page_35.txt
ccfbd87cc8e4339fb24fc2365313f629
80843f39b1ffb1532f463bb5bb442dfabb1d2789
6273 F20101130_AADBID bokach_m_Page_47.pro
e5b776e95eb61a91b5e4067c94947c3c
62073a0e416469f16dcc9ed77f932636b391c696
18193 F20101130_AADBDG bokach_m_Page_36.pro
ebae8225fe64762f6a20442401d69601
96b092f643536307b901d286eff37b3b762ae825
F20101130_AADAWJ bokach_m_Page_07.tif
6d8d221925cee4059b4fe732850699fa
a5ab6bf378b7eae585674dc000348fd94271dd4b
468 F20101130_AADBIE bokach_m_Page_01.txt
44e2761ecd1efd858d0d6ceea8b2f610
813629aaac6ae2ef4713f1ac5f13845fca427c44
88663 F20101130_AADBDH bokach_m_Page_06.QC.jpg
c632c10e73f631f967df896b6419a355
e5db885bdaa757d97e25bcdd946722f94bd4ade3
56056 F20101130_AADAWK bokach_m_Page_29.jpg
747c9111911bce74dac588d535eed075
adbd6817f13845f890d7cdf29a55d3d9b975427e
278 F20101130_AADBIF bokach_m_Page_03.txt
2889a6292d862374ac81246395baa095
7dc6e6e567345cad0aa9a13e343ad9f7ad30ad83
294 F20101130_AADBDI bokach_m_Page_42.txt
9873b3d5b9150bb6390bcfb1b88eae46
710da49b75bb26d2da28211aa4b74dca452e50d8
215903 F20101130_AADAWL bokach_m_Page_24.jpg
e2396fd00854e8e0791f1c1562aad188
4a7afb9492a920d56e8a5890adfc075ed82955dc
3160 F20101130_AADBIG bokach_m_Page_06.txt
be88c92e1cc1ae94d1b7b72c15c7b226
8d372f82745e759d30f615c59f2df1bb087b9817
21708 F20101130_AADBDJ bokach_m_Page_22thm.jpg
692b2a4e6400ef334e57ad2c599fbfd2
9acbbfa3abaf6744110983ec6c5d01f103dd2acd
9566 F20101130_AADAWM bokach_m_Page_51.jp2
c82331d8c5fe65849047f9ceed2f3d1f
0b08b29776105b25b2ae6afdbc45cabca9eab7e6
1224 F20101130_AADBIH bokach_m_Page_08.txt
99a5f5a71077fab1ae65a162c52f422d
58a3f4c84e4f3d091246bcf33c2dbf2b74b763b8
1051967 F20101130_AADAWN bokach_m_Page_08.jp2
8233e2dd5d90d5f91aa07ee3c73bff7f
8eae53efde2108fd1315ced30d38b1bec4a60a31
1846 F20101130_AADBII bokach_m_Page_11.txt
82c59634931825d07bb50d24dffdc487
ac48e0a690229940d8c25daa8cd11878668de1f1
42448 F20101130_AADBDK bokach_m_Page_48.pro
b7c95dcd26d17d34ec988acd6d0d4c88
272c93e666eb8ebda8a05a5995a9287185d96f66
13765 F20101130_AADAWO bokach_m_Page_55thm.jpg
df651656c9dc78cdfd3ab99a0e58355b
1b38de71ae443fec9175cc9bbc46727ab419b8b2
2034 F20101130_AADBIJ bokach_m_Page_14.txt
f4a416147bf318ceb11ab774e0914d85
9ee959cea7478bb735855accad56c471cdce3bfe
F20101130_AADBDL bokach_m_Page_42.tif
767a7bb2e25505ecffb1ce4ee9d2a564
6ffb471a54dffe6e30978d3ebfc886e7248d831c
856 F20101130_AADAWP bokach_m_Page_36.txt
af5e501dc977a22c24a19385f573d5b3
deb4cc2c35e98741a0d973d6666427c56401e701
1640 F20101130_AADBIK bokach_m_Page_15.txt
2a76f6e75b760c46cb0af40034395118
f3aabca9c11f3be951896436447f523f91e6998b
244801 F20101130_AADBDM bokach_m_Page_43.jp2
910c3660e4c4b14a17d9cf0da4c4b085
93c8eabaa8296198253b6f1aff5e972e22ff4112
F20101130_AADAWQ bokach_m_Page_52.tif
7a37169b0427bce94456c2e3bf67d070
e6fd05fbd902b03ba0ef033524a8eae10a955f39
1755 F20101130_AADBDN bokach_m_Page_22.txt
50bafea7e22072d90741c7869e3f5f8f
955c071fba1913f14c145b78ac982c5f8e6e73b7
11223 F20101130_AADAWR bokach_m_Page_23.QC.jpg
2ac02dcf1a16f5002ebfaf5456342db0
c5acf16229e66f47a0fdb31608690bbffa0511db
2084 F20101130_AADBIL bokach_m_Page_19.txt
0bfff3b29e269d6f176aaa9e807f9045
c962853c137b4b31839911027063ef9f95653285
260725 F20101130_AADBDO bokach_m_Page_45.jp2
aefa383f07a7e2bac8cda82dd7f24794
b5f175eee11e8da93449be532ececa5d8d28e85d
104992 F20101130_AADAWS bokach_m_Page_12.jp2
e49750a2da302a4b742f0a0a2ee013ca
6ac46d1111b097171313bc4ecde0aff74e2de308
1803 F20101130_AADBIM bokach_m_Page_21.txt
d6becee977e2edea563c151b002334bc
96f45b3cb16c0d0d8d3023e9df6af83a49f11307
41040 F20101130_AADBDP bokach_m_Page_56.QC.jpg
654100d6743a8c73ec4a78f687358010
3aac2c2121266038e57fd916cd910ea259cb169a
1815 F20101130_AADAWT bokach_m_Page_48.txt
b4decb109f30bdc0d976f63c2f906eda
bf639b8c0b3c126cf8de83af0b1d2959f1e7538d
208 F20101130_AADBIN bokach_m_Page_23.txt
baf662c1cb25cf7c9fb37923d706e2d8
25e84cf7d600c388374bfa560c4831df55fa7688
F20101130_AADBDQ bokach_m_Page_37.tif
d489825af7bd1c094612322654f55ec9
a3f1b72f08b44b68c51c10574e207982cc278ee6
21354 F20101130_AADAWU bokach_m_Page_48thm.jpg
49afea29218eeeb46ce3f14d32831793
08369f38e95abb7c70bc54403c747e0eb65dc644
1858 F20101130_AADBIO bokach_m_Page_27.txt
4e3cb776c84a9339b7a2407668ec379f
60d3d4030ebeacdf6906260864121eed6a62a66f
164347 F20101130_AADBDR bokach_m_Page_09.jpg
7f62e91f9672da67e3e10a2b0d3b07a9
aa4515972d7746bf0744dce4a1058cbbe0d6b802
F20101130_AADAWV bokach_m_Page_13.tif
1a06c8f977d39d28311ad589b507fe9e
e04fc2b632e6cd10be7a9fd1fbe815f9bcfdadf4
618 F20101130_AADBIP bokach_m_Page_29.txt
c82a23ee4f8e9fd0c04b6f216032f734
9d1631a50a57638f758273fb375a575e5acc44fc
884 F20101130_AADBDS bokach_m_Page_10.txt
fb6e96321a258817536cc9331d27c226
6e6ee36928d9e286779751f765be5e445133a39b
19883 F20101130_AADAWW bokach_m_Page_14thm.jpg
b001dff5343e56f9d21d07bcc3a5ba3d
7441e710c70f89983cadb7bec35ed406c816e9bd
1307 F20101130_AADBIQ bokach_m_Page_33.txt
2ca25531754636b92e2493cc004a1361
e63ad2d9d42d2e026523ca6179b2a37f408ee93f
122204 F20101130_AADBDT bokach_m_Page_44.jpg
59c9b823001582d65ebfde4ad59f5277
ae2aba661a94b011bd6917b35dfea39bb06fd7b7
68456 F20101130_AADAWX bokach_m_Page_22.QC.jpg
1fd839eb4bce1263d0c13d649ecd8648
2f822cd633d9a987e257439bdad8164054af7885
893 F20101130_AADBIR bokach_m_Page_34.txt
2d82a9464c2863aff3a6a0e72f706b7a
25f4b38824c3e0280c00b145c040a7b65a386e81
508809 F20101130_AADBDU bokach_m_Page_36.jp2
fc9022f4de8ca3ba744b181087eee0a6
41b344d4111925f3957a796210b88049e69720ec
F20101130_AADAWY bokach_m_Page_21.tif
60d95c307b133337cdcfa8b6afce0cf6
61ad9b458cf7e3bd5abe3901fc3072fa19d12791
457 F20101130_AADBIS bokach_m_Page_37.txt
f4a642a4503161bb461eeb8abb906065
252c659a893067ab0e44de8795f0815cbf957093
942 F20101130_AADBDV bokach_m_Page_32.txt
febc5d33bcbbc33dfab4a0324039b7b5
327fb088bdb9105ba5a26a33d338c86b83c84820
123357 F20101130_AADAWZ bokach_m_Page_40.jpg
ad560133987ecae1fb4398b9e82438ac
de2ba0883df5b3f89e1040e9bdf6227528151f9c
373 F20101130_AADBIT bokach_m_Page_41.txt
751dcd93d2b88fcba10b1156e37bc4d9
9a2ffb8a5c275012a4f6769eec917b17265ee329
108703 F20101130_AADBDW bokach_m_Page_19.jp2
8599488d110f88edc7df7f0fdd46e0e3
e1ee67832bfb117fdc4a835e52f7b06543d7c108
510 F20101130_AADBIU bokach_m_Page_45.txt
ac1a82fc96e5b8663610cc6578298111
43bbc152a9a1701fbcb31035279d8ebb6fd51ef4
25635 F20101130_AADBDX bokach_m_Page_23.jpg
76b9003c41e6c1ea46ed47e152f6b3d8
55c64dfce4905ffe6d2f933d6cdc00da646225d0
552 F20101130_AADBIV bokach_m_Page_46.txt
9ca1fae912533e5c691908612ccd09e7
93cd42139f32f0382a0901f04ebcd1654fbdb672
51930 F20101130_AADBBA bokach_m_Page_49.pro
b5e09ce80d1610257112247832564baa
1240e09b38728c7b14245c8c232eea035b00006e
1902 F20101130_AADBDY bokach_m_Page_12.txt
751cf465e008a533e98cfc2b50fb6f69
ad076f77463c7dbe8032ee2952f8e9bae94e13b5
2528 F20101130_AADBIW bokach_m_Page_53.txt
c2bfe1853bdaf767a50643f75ddba145
044925f68676dfff64a68e5524ba9b15eea60be0
987 F20101130_AADAZA bokach_m_Page_13.txt
5593bb798b57ab014e4eee24c5ab6389
05d6c51213a8dba1ee6aa21bf0e86fdee6ed5e10
33113 F20101130_AADBBB bokach_m_Page_05.jp2
7cfd8af8ba856042b867b7df17a5b31c
24faad4c2106e7b49931d3fcfa00252d4f6a7338
1051985 F20101130_AADBDZ bokach_m_Page_07.jp2
8425b87ee98fe890193376354572d1f2
001ce9d6d4b0af3b82dcc0168dcd3fab51bd1e92
77685 F20101130_AADBIX bokach_m_Page_18.QC.jpg
ef62eea7eb2753262d698c5504241e55
f7e12f76945a2e747c8cb458ef579a6672b47bd8
F20101130_AADAZB bokach_m_Page_34.tif
d628c7670bb390f48e13875d54a47754
e5d271efbf796c06e5c84c64a1ba99536d076e7a
34367 F20101130_AADBBC bokach_m_Page_41.QC.jpg
62e00dbd4e61929291dc019425fddda5
546495d5dd45b859d2da9ada2325930721d3b1f0
6035 F20101130_AADBGA bokach_m_Page_02.jp2
b5a3194befb3040d465d0d71edb523fb
0c420de737263ae196ff5eb2413a46440b88376e
44833 F20101130_AADBIY bokach_m_Page_26.QC.jpg
fc39fc6a588529faa864f40f08c409f5
5a62b2905c964806a621cc23c10447adef033ed3
48038 F20101130_AADAZC bokach_m_Page_44.QC.jpg
8e43b4ce1e7ff4815228ed499256a578
b674732c000b10184f80ef873325cee96a8eb1e4
F20101130_AADBBD bokach_m_Page_18.tif
24b5191269e493477e69e4913df784f7
e2977eae655416163e3eb31b6dd1ba25ed80bc02
82560 F20101130_AADBGB bokach_m_Page_09.jp2
1f9130219bc0786ade718883a617e67a
cee7f6831008b153095fb58d05af2f22c2bcf68c
44793 F20101130_AADBIZ bokach_m_Page_25thm.jpg
7d9a2d8dc9474d7b976870fddeed4108
22c6b2efd7ebb632962295b3e5346d5b85c76dec
57312 F20101130_AADAZD bokach_m_Page_01.jpg
a1202b3724d22b31d151ef29cc2fb161
d90829d2876fee3151ef6b04cb3a7df180757c97
98175 F20101130_AADBGC bokach_m_Page_11.jp2
f01f03c4a3b01c032821681954da7ceb
cce39d19400178a526a48f35fdb4278bbd919c7b
6986 F20101130_AADAZE bokach_m_Page_39.pro
1e693a5b3629ed9d247416c48e3d53ba
4a3dae2a99e18366ec045a23f1b6df17e071eac5
212230 F20101130_AADBBE bokach_m_Page_28.jpg
7217cb0d3752add8d94b47525dafae04
aac4086a00a151ccaebe023c89fadc45cc66af52
97930 F20101130_AADBGD bokach_m_Page_14.jp2
46300708b9b8ffaa5a417eb1cb246827
d2295b7771df5e8293fd1ebd5c8af6ca81ee8204
39040 F20101130_AADAZF bokach_m_Page_35thm.jpg
cae59e5d6bd0a1ddd593bb98b5b11b0c
a3853541a5af5c47a32259d1ad0a847c7515b1a1
F20101130_AADBBF bokach_m_Page_44.tif
96a1eb33c654d2c655ba282b4a56f05c
a2ad7ec11b712419ab453896a5d7381f5bcf9ff9
87048 F20101130_AADBGE bokach_m_Page_15.jp2
421f0c2fc1c304c5d48f2752c1f5ac52
ef332e42d37338666243397516dd33293b7d3a8a
98762 F20101130_AADAZG bokach_m_Page_17.jp2
1c0f751d528e89b370b625574216192f
9ac99eaf0da77519b65b959b7d689bdaa124a232
42035 F20101130_AADBBG bokach_m_Page_22.pro
45e262e7fa521c30cc37f68dd7e10089
a71b7ecf27dfcd924941a58e2ec9afb208347706
106849 F20101130_AADBGF bokach_m_Page_16.jp2
c2a735314a5f593ca822bd3c29c57e88
03a233ec18ac74534dc635ef48690a0e5c66651c
76300 F20101130_AADAZH bokach_m_Page_12.QC.jpg
e9e741b5a1c1a09e40c565960847dde1
97d0815dea0f5e53e089d5376826dfeb63440ec7
42080 F20101130_AADBBH bokach_m_Page_13.QC.jpg
d9b234027e0c94b2aa9393e1b7612e22
5fa24b877d60dbc987b626120ecc8e46d71984a2
106174 F20101130_AADBGG bokach_m_Page_18.jp2
2fb2b9c71f12a0ab2da1864ab22c9917
1e67d86dc25697cef6cd585791743fae86c02cc5
69602 F20101130_AADAZI bokach_m_Page_32.jpg
d5a6c44d768b5b1777c3194329c1c470
038f755089ca788d0b195d66ec6adbedac320b8a
18116 F20101130_AADBBI bokach_m_Page_40.pro
c5d080d445249bde933f2150f1d157a1
38b489550fc42601e15db0089d144f396c485a4c
81011 F20101130_AADAZJ bokach_m_Page_20.QC.jpg
2ef104db4cfe3014595d5f070f4ccbf3
21097b473f9ccc72f78ae81fd8a7af912ec7d3f2
F20101130_AADBBJ bokach_m_Page_08.tif
38cdafb56ef18938b3310f62e16320d2
75fbbad3a719639b2665954ade2df100d03cf0c6
109865 F20101130_AADBGH bokach_m_Page_28.jp2
7f7f237071e13e05fb7835af2224563a
e6384254a2241546aa241986c129f019184c56b7
13488 F20101130_AADAZK bokach_m_Page_33thm.jpg
b81d426008c49173c037633266088111
8b63e3b43f77a6a92553368e84e0bc9230112c37
83148 F20101130_AADBBK bokach_m_Page_07.QC.jpg
47e7de639e23497c4758976c119dfa70
2c726df05296fa2da7a9bc1f7ceb7c1a61c5e6dd
257914 F20101130_AADBGI bokach_m_Page_38.jp2
9946cb43dfc8fef95fa6d25cf56b7cab
9fa383298e166d3816418baed8ceae165eed91fa
137346 F20101130_AADAZL bokach_m_Page_35.jpg
4bbdce51ce11a003888dec4d15973b27
18bb09d11adc9f4f0037ce050e3e7804d271f34a
5880 F20101130_AADBBL bokach_m_Page_03.pro
792ccaa67371e10de1e0bf01d3dda351
12e217f036677cb4c9258f38d1983aa6ee9dfcda
188685 F20101130_AADAZM bokach_m_Page_21.jpg
c6c51ee5360f56a325f65d8342bed2b3
a6bd14247a66e02cb8ded487584ac7aa2f6b3bec
73633 F20101130_AADBBM bokach_m_Page_43.jpg
254eb38afd0353883194fc4a715da338
b4ca3fef62e5e822a2384a09dd97177a933f53b9
264591 F20101130_AADBGJ bokach_m_Page_39.jp2
c4c77cab70f9429cc6057fb42bbda2d0
d05feb94c2832e453911de3704a43ce59ebfc37f
74967 F20101130_AADAZN bokach_m_Page_16.QC.jpg
3914d4694a29f4787f5e3171e4a3e47e
2a10adf757db75439254d4c9eb869e3f7c596e93
17167 F20101130_AADBBN bokach_m_Page_41thm.jpg
b592b76cac34590272329bab69a5222b
feb7cbefecbe1a229bd8cce60981b3cd7328884c
261600 F20101130_AADBGK bokach_m_Page_42.jp2
59a30793017cca7fafe5ea0ca46dbba5
5b9de9423ed132741620a03f3bd692ab3a76c421
400 F20101130_AADAZO bokach_m_Page_47.txt
4b59f1df2b25b5d78b94c3e17c9b86e1
5d3a68c856b0f8d9c7a6bfe6bdf82cd6731a83e0
81229 F20101130_AADBBO bokach_m_Page_39.jpg
fc47e7f63bdfa58d41d511a21f0dbab8
db62ed070f2694ae970f0b4ea3e95bb2cc92e3cf
272533 F20101130_AADBGL bokach_m_Page_46.jp2
7f35d2bc2b58b46d9c729677181ba87c
2d1e63d7e099afea65a64ea90ba1a0c7dbec8274
876 F20101130_AADAZP bokach_m_Page_44.txt
8855a98566b4e9dac29fa3c53f4a25e0
6fb132ed9100981f16a6462406bb72620cf4584b
19309 F20101130_AADBBP bokach_m_Page_26thm.jpg
4d83ca7c8ff059ec9763f138f665d86f
4bb172293774e6134ce712cbefc547d8641931fe
249727 F20101130_AADBGM bokach_m_Page_47.jp2
afbfe824dc9bd99d7009cab6439cd73b
85ca4c6d8c83422a95fafa16cb8d6a144a6d8b53
192333 F20101130_AADAZQ bokach_m_Page_14.jpg
5e548e1af46d280968e3170e38a1a788
7571eca698c2448cdbdbc37e626aaa44027a8000
50067 F20101130_AADBBQ bokach_m_Page_10.jp2
961f4b8b40a50fb0f62a4029bd828226
b35c66e90826a3d1f4f3d32e89a87c875441095e
92055 F20101130_AADBGN bokach_m_Page_48.jp2
64b0f6d4ea4855eb4d4891271be325d2
f432ee409b8f4c42a215d6e33041e863af5bd156
1051974 F20101130_AADAZR bokach_m_Page_06.jp2
ca5d3bd84eccbe22cfa20af040a128d3
f910a6f5bcbc46ba8a5ab50fe02ca5d087361218
554 F20101130_AADBBR bokach_m_Page_05.txt
74a625c86e76edd02131d9893ead538b
4394259bc5540a69ff2a5115264e1584abc43c31
107427 F20101130_AADBGO bokach_m_Page_50.jp2
51a8e082730483ddb7abd294192855be
9288ed60cc203f6ccccf9c7f8fe17f2282193e01
4495 F20101130_AADAZS bokach_m_Page_23thm.jpg
b04bdf9da3ca84957b4fdffb85f4b28d
0bb6429a34275ad356f784281f019547c374f97b
18267 F20101130_AADBBS bokach_m_Page_09thm.jpg
5362082dbd0543375a41d8c26a15b2ae
83d8174a7fbb7f93e4b1c819ee1e6597df80a1fc
130247 F20101130_AADBGP bokach_m_Page_53.jp2
2dac30f77b7d00393b8c24c602479e2e
a293a7b1c62bd62655f6e8476b630189e91b751f
94975 F20101130_AADAZT bokach_m_Page_27.jp2
6fe0d5bf4e52128aa8f76e9f3579b2bd
0d09cb13ad542ce0a9e709325a1a014080fd3a62
8876 F20101130_AADBBT bokach_m_Page_45.pro
b2c234b31f3cabcece124cf02eaf3281
8c3afcfd114b2f4d62a83d6dbd6ef00adeb42a35
129395 F20101130_AADBGQ bokach_m_Page_54.jp2
08a8deb88167334c99bf10139dec8348
64250e1b77b2b545e77a8a8a41530f3efd17ba51
1948 F20101130_AADAZU bokach_m_Page_16.txt
ec120b7061c1b5f7c318b0f0997b36f3
d89c07afbd1331a39c40cb01dbf2e29176d8a04e
F20101130_AADBBU bokach_m_Page_26.tif
55e8797903edebd1029f440ca157f10f
d3a8ee814a04727fd8fb54da1ae0b468aaa18e33
55326 F20101130_AADBGR bokach_m_Page_56.jp2
ef3ef211e2b87cbd31e0a3f60bb194f5
cb709d435a4791c0565366d753bdb3b3205373ba
12185 F20101130_AADAZV bokach_m_Page_34.pro
1197352ca79dfbed47af4aa80e66d708
4d3fbdf76b26078a69b8b9dbe6d6cbfb5f6b3dfa
67722 F20101130_AADBBV bokach_m_Page_17.QC.jpg
77d55910c5ba132981d0b3ef8604d565
b55f1013181217b603709cb50720bcf0ce95edac
F20101130_AADBGS bokach_m_Page_03.tif
1c0525cdf7f4315d9c0b278435896808
736b48b1d0a454b9359e69a60548a4d0104ce2a7
1935 F20101130_AADAZW bokach_m_Page_18.txt
75b7d8f60107789558e389e22b1bb456
6f660295bd389f85240bd06bdf2b1d5c41ffa88a
97850 F20101130_AADBBW bokach_m_Page_21.jp2
8d2f36397db3083f6343df53968827da
bbac53fb84f37e4abb4b46694d1c9c2dbdc73add
F20101130_AADBGT bokach_m_Page_06.tif
79d679f6a892c9333df6200c9f60ae50
691ef1f9158ffdfa0dcc0d239b37483959f79650
2041 F20101130_AADAZX bokach_m_Page_20.txt
5a04330a6484b6207619cf00755c3d60
4c6e14df0838aa6dda8fae33b119d82d66aa1d1a
26130 F20101130_AADBBX bokach_m_Page_53thm.jpg
008bd1b73ebd90fac62672201b13e4c2
b623a4d060eae014a731049f63939a669c5a3f9f
F20101130_AADBGU bokach_m_Page_11.tif
aae714cc3b63aa1928ad2f0dabc71230
55ae191117d7eaec64fc7219c4e5b9f7992d4926
17713 F20101130_AADAZY bokach_m_Page_37thm.jpg
2e0d51296347418015c6d423081b1ac7
e5759a6922ae20ab03863bc18c36d9b93663fe76
502685 F20101130_AADBBY bokach_m_Page_44.jp2
4fb788cf3a80fd39d7061342c4b49e77
264d23e55b34478f148246efae5c1a96999fa786
F20101130_AADBGV bokach_m_Page_14.tif
eeabd6337ad4330dbbe296f64105e62c
fcec4ab7db5159646482c6a1b6610b352526235f
15430 F20101130_AADAZZ bokach_m_Page_02.jpg
5ae70607de6a066ab42486de8f8cc1ca
3a6df111a1c572afdf9def74a8f99f693107f302
107216 F20101130_AADBBZ bokach_m_Page_31.jp2
bb404e00f89d7023166d121a669e8562
05d5a0e59549e5cf2b46eb169f206f5ee11aa2e2
F20101130_AADBGW bokach_m_Page_20.tif
306f83deb2a21e0c3e2a803223f131bb
07f65a5e1d742007ad6a2dabca5372657da4b3c9
55349 F20101130_AADAXA bokach_m_Page_30.QC.jpg
b16d3e16e9ac0a92f6051abb70b4ce74
278d4d15d5c7a5cfe0ba384a60ba69446113bdef
F20101130_AADBGX bokach_m_Page_23.tif
fbbf21a9c28f227e5cc80a33e6bb462c
3c1613cbd48de5fdeacb8f8d0009fb4c5f1b9cdb
F20101130_AADAXB bokach_m_Page_36.tif
de5f9c5ea3abe21ed14642456b61e1e2
ecbdaf49ec04e29b5bd49472215876a70eebd28f
34420 F20101130_AADBEA bokach_m_Page_55.pro
93cec6e0a55e9c5062bb9240535429a2
b406fbbf65bc9253669e04b0be900647cf711de9
F20101130_AADBGY bokach_m_Page_25.tif
7a9dfb57c5f5feff01bb76d5e44aa2b1
0e40b0af51306d7d65993ef2c890c9fe2ab37600
F20101130_AADAXC bokach_m_Page_22.tif
3ecf522ab7db102eea30ea2e1cf4bb2e
60399c1b54c2de1fad9a259134501f4a7678831a
89469 F20101130_AADBEB bokach_m_Page_22.jp2
b1339bde1980e2aaaa378df557b16812
59442be7061133b8a1545611e38b6f73153c37c7
F20101130_AADBGZ bokach_m_Page_29.tif
15eb33afb3ae76d55739f5764f2f6418
faa60fbdb38a3721dc328fa36dc1360ac722db01
24300 F20101130_AADAXD bokach_m_Page_18thm.jpg
60df3c2401f3ba50d27265597ff795a8
f10fad773ff4378a8beb49f81d19edf23091471e
217418 F20101130_AADBEC bokach_m_Page_49.jpg
7f0c14c57ff2c850fc63cbcac01bd725
6462c36e383ffd4e2bdc7f52d805d4db0e9bddf5
5164 F20101130_AADAXE bokach_m_Page_43.pro
88c5fe75c47105746250705fc124c2aa
dd92721b6fcafa8d1f42ab124a01063af334c37e
21999 F20101130_AADBJA bokach_m_Page_44thm.jpg
57300605fbb9a8cd7b08f0045f8b95ac
fc99bfa20f5f115cd538e10814884819919eb886
1797 F20101130_AADBED bokach_m_Page_17.txt
7eb5bd24934bf0c21591990d4a25ccf2
1cfe9c392d75cde49793384b6fdc046eae62e847
48607 F20101130_AADAXF bokach_m_Page_16.pro
bb675f1026a060a2038a4ac13834e3fe
08badad6f1557692efd6e79d00a197c5a9eaf172
64013 F20101130_AADBJB bokach_m_Page_14.QC.jpg
982e49dc8a77cd33fd22ba274ef1d016
f8f2c60849456f1f786bd6cbdb742ebba52005e7
51867 F20101130_AADBEE bokach_m_Page_20.pro
dff1387ba944468c0109a1f771c066ab
41b1cb4c4443a58f756d6a552021521000bee901
F20101130_AADAXG bokach_m_Page_48.tif
2f3f40d188ba271ef9b53a5a16a1888b
2c74d72c01c6506a80e8aa7cf1061aa4aea49521
42470 F20101130_AADBJC bokach_m_Page_07thm.jpg
bb4a9852474529133e93993f27823017
8b838713513f2c75a4d1f7aea7aeda09f98afafa
34220 F20101130_AADBEF bokach_m_Page_38.QC.jpg
ad130e7a300ff69fb3301e65dc852ba2
1ec8ebe44ffda0cf991ab7197a23e2f7934c0fd4
35039 F20101130_AADAXH bokach_m_Page_46.QC.jpg
450ff45c29f1d9b68ed15572ae52701c
5cca6348f3d8de127c8effd89f84c15de99ab918
F20101130_AADBEG bokach_m_Page_04.tif
cf2b1d3dc4941d83f8cca359ebcbffe9
dd1e1841035d0c490ece68c99dec18701278a07f
67894 F20101130_AADAXI bokach_m_Page_27.QC.jpg
0c9a8df8e8dd92d3953b7a954bb7a02e
104fbf16bc4f0bc790a45629fc76a0383064c52a
48120 F20101130_AADBJD bokach_m_Page_36.QC.jpg
f4cfdf88265ee43b37836db47ae54015
a03d258aa68c896b0c15902c3480aa7c0db28aec
1633 F20101130_AADAXJ bokach_m_Page_09.txt
90cc21ba57ed72f46cc267e319a7d8ae
9b4ed0846288aaa17344d2aa10a98f5258dc3671
69009 F20101130_AADBJE bokach_m_Page_08.QC.jpg
5a78e0bcc92853e40801c6a96302e91c
5a7944ec7e970930144fd3b5e6421c98fbc7dfd0
2062 F20101130_AADBEH bokach_m_Page_49.txt
3449fa3c457c1404e2764b233fab2e42
926cf89c1216a55a87046381ad25a795b2fb6b67
265264 F20101130_AADAXK bokach_m_Page_53.jpg
e0aeee093ca3b0112945b7510607184d
78e0163e0f76a16e8c528a460baf3040904ad520
23730 F20101130_AADBJF bokach_m_Page_50thm.jpg
e05d82ed66e3c44169dfcb4bd33fa2bd
92e77ea8cded51e32f287d9c3681e070e1572c07
12560 F20101130_AADBEI bokach_m_Page_10thm.jpg
0a4b9b6fddf9c5bf50c7b2ba400ee022
e157e00ab388b621922beac2d99516757a2dbdb2
77700 F20101130_AADAXL bokach_m_Page_41.jpg
7fb71d78b1dbda3703877b8e70e5627d
d9623a19425523a5c4b3a95a7c87d5183c9296b7
13637 F20101130_AADBJG bokach_m_Page_56thm.jpg
e926c36b0743831c92f8cf8ea99ff5e6
f1ddb3ad624e247b051e8ba84c445c7c326ee43b
109424 F20101130_AADBEJ bokach_m_Page_56.jpg
81a1fa0e8ff3b5d91587f9c411697244
f23ff74f32e7b6d5e72e926ce6f08ef800db03ed
F20101130_AADAXM bokach_m_Page_16.tif
61b4a5e0eb48c89e2f3fa3fb27d2fedc
bae25563458dd311c03f4cd107002acc1836c3c2
78464 F20101130_AADBJH bokach_m_Page_34.QC.jpg
0cfa3fc0c01f7684eb50b0950a53b761
77dee168e34b2f3c3da241da00f15e5d271ea410
70892 F20101130_AADBEK bokach_m_Page_11.QC.jpg
7ce4f095f8f06562c202b7eb063b44b2
0990cd3970aba0dd73d3d897b222993ee8b1758f
78026 F20101130_AADAXN bokach_m_Page_42.jpg
720300d19223596e831e3c30d8be8631
39c1cc322bbd44109248234d62f3301f67a0ff6a
22835 F20101130_AADBJI bokach_m_Page_17thm.jpg
1578e97c5a296c310ab7713b9168fb5d
a9554d70154cc19fb870664456aa7a2ae01c7efc
25932 F20101130_AADBEL bokach_m_Page_54thm.jpg
0e7780b81a92211e41ddb4a3e4926f7f
0163e361d1120b704f73eda6d2e7aacf381ab5c9
4573 F20101130_AADAXO bokach_m_Page_03thm.jpg
3d8e967b4182775720ac879d45743818
45169c45f8cb069dd7750ff361fc0f2112253ae3
17943 F20101130_AADBJJ bokach_m_Page_42thm.jpg
4223db4b34cfb10d79b241feff0d5d29
d1ef61b552cd2188d9e5e097d128295ab5a98421
34889 F20101130_AADBEM bokach_m_Page_42.QC.jpg
48475f2144e6b723d923df3793f5985b
def154c5f9619d2a6af6cdaecbc34e21d7fae933
F20101130_AADAXP bokach_m_Page_10.tif
72d39f60541b947035bd801464d700d6
31096c3966e5891ed92835fdb899958b4b01bf1c
84176 F20101130_AADBJK bokach_m_Page_54.QC.jpg
3dfe20469eab99f5854443e70c79b1a6
949187edd390ba15fe9228b5fff8ace7a71a00fc
62182 F20101130_AADBEN bokach_m_Page_54.pro
0b31be1525c8ce2f4ee6397ca946dd81
e7905e4f29ff03ac23664f1c847cf3ef5025a34d
19637 F20101130_AADAXQ bokach_m_Page_51.jpg
8da6d50a712452d50d826ec68bd5381d
a8b25b3fc6cee753e7ec7f0b52920a8b1b8023fe
17684 F20101130_AADBJL bokach_m_Page_47thm.jpg
5341bc85136837ce69957736def1428b
1b961a53225858d2e86868ecd863d1ff7fd0f3a0
17577 F20101130_AADBEO bokach_m_Page_45thm.jpg
905e4e8c61eee84b06d70665f67dcb3d
dd1edc2c96564a0391969a0495b61ea593e0b9d5
96886 F20101130_AADAXR bokach_m_Page_10.jpg
efe6c08f7a4117d64ee8c23d16eca6e6
43dc833f382ce589fa67b4f5cf0a42bfe91a41a1
F20101130_AADBEP bokach_m_Page_05.tif
9b805f61f4447a072f7e9e6ae9f55360
fc89a11cb853298405ba5d0619fbf8cdb5fd46f3
21771 F20101130_AADAXS bokach_m_Page_40thm.jpg
6db6317c2950aae212e8c8968a270eb5
a2cd23bef4632b0db95a6edb9d6a962d45ba3dbd
42172 F20101130_AADBJM bokach_m_Page_34thm.jpg
50cbeb227524cc2e750805c20f01bc31
3507a232e0ffaa73ccb95da165f43200b07c5b47
597 F20101130_AADBEQ bokach_m_Page_39.txt
cff7031ae5b06fb5ba9f3a1cd56ebd69
011c1044dc6cf344780c6387e9fcfaa01da0ddc6
3193 F20101130_AADAXT bokach_m_Page_02thm.jpg
c5910b67c654ede28f7f8fb761b29402
f5463e2369ddf280943b45f24f5d74b60c0d7ab8
77146 F20101130_AADBJN bokach_m_Page_50.QC.jpg
c51878c60066066bcab9f26edef9d20a
c67dbe6fb2e127fc6a1c24b5c188ba2809d83554


Permanent Link: http://ufdc.ufl.edu/UFE0009586/00001

Material Information

Title: The Relative Influences of Predation and Prey Availability on Ardeid Breeding Colony Site Selection
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0009586:00001

Permanent Link: http://ufdc.ufl.edu/UFE0009586/00001

Material Information

Title: The Relative Influences of Predation and Prey Availability on Ardeid Breeding Colony Site Selection
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0009586:00001


This item has the following downloads:


Full Text












THE RELATIVE INFLUENCES OF PREDATION AND PREY AVAILABILITY
ON ARDEID BREEDING COLONY SITE SELECTION














By

MATTHEW JOHN BOKACH


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Matthew John Bokach


























This thesis is dedicated to my parents Mark and Cathy, who have
weathered the many twists and turns of my life's last decade with unconditional
love and support, leavened with just a touch of legitimate bemusement.















ACKNOWLEDGMENTS

I could not have conducted this research or even made it through graduate school

without the support and assistance of a humbling number of other people. Dr. Peter

Frederick has been the best advisor I could have asked for. Without his support, patience,

and humor, I would not have finished this degree. My data on the wading bird colony

sites was collected by a long series of his former students and technicians, and this

research is therefore as much a product of their work as mine. Dr. Graeme Cumming and

Dr. Jane Southworth, my other two committee members, gave me invaluable advice

through nearly every stage of the research. My initial explorations into the daunting task

of analyzing these data benefited immensely from Dr. Tim Fik's statistical insights and

encouragement. I gratefully acknowledge Dr. Carl Fitz, Mr. Jason Godin, and Mr. Ken

Rutchey of the South Florida Water Management District; and, Mr. Troy Mullins of the

National Park Service, for providing me with the hydrology and vegetation data used

herein. Lorraine Heisler, Les Vilchek, and Shawn Komlos all provided insights

concerning the development and appropriate uses of these data. Thanks to Theresa

Burcsu and Desiree Price, I was allowed constant access to the computer labs of the

University of Florida Geography Department, even when not enrolled in one of their

classes; I will always think of Turlington Hall as the "home" of this research. Lin

Cassidy, Andres Guhl, and especially Matt Marsik all gave me valuable assistance with

GIS questions. I blame Carl Evans for both complicating and vastly simplifying my life

by introducing me to Matlab. The patience and support of my boyfriend Rodney Brown









was a steadying influence through most of this long process. Finally, I thank my family

for all of their love and support through the years.

I was supported financially for nearly my entire time at the University of Florida by

a research assistantship through the College (now School) of Natural Resources and

Environment. I extend my most sincere thanks to Dr. Steve Humphrey, Cathy Ritchie,

and Meisha Wade for giving me the opportunity to pursue this degree, and for extending

my funding for an extra semester.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ................................................... vii

LIST OF FIGURES ........................................... ............................ viii

A B S T R A C T ........................................................................................................ ........... ix

CHAPTER

1 IN T R O D U C T IO N ................................................. .............................................. .

2 M E T H O D S .......................................................................................................... ... 5

S tu dy A rea ....................................................................... ................................. . 5
D ata Sources ................................................................................. . .. ...............6
W ater D epth G rids ... ... ......................................... ....................... . .......... 6
V egetation M aps... ....................................................................... . .......... 6
W ading Bird Colony Locations.................. ..................................................8...
D istributional R relationships ......................................... ........................ ...............9...
C alculation of V ariables ............. .. ............... ..............................................9... .
M u ltiv ariate A n aly ses .... ... ......................................... ....................... ............... 11

3 R E S U L T S ................................................................................................................. .. 1 7

D distribution of C olony Sites .................................................................. ............... 17
Bootstrap of Foraging H habitat Calculation............................................ ............... 17
M ultiv ariate A naly ses .... ... ......................................... ....................... ............... 18

4 D ISCU SSION ............................................................................ ....... ........ ............... 38

E effects of Site A availability ......................................... ........................ ................ 38
M odel Perform ance ............................. ..... ... ............... 38
R esponses to H ydrological V ariability .................................................. ................ 39
M anagem ent Im plications ................. .............................................................. 40

L IT E R A T U R E C IT E D ................................................... ............................................. 42

BIOGRAPHICAL SKETCH ...................................................... 46















LIST OF TABLES


Table page

1-1 V ariables and m ethods of calculation ................................................... ...............4...

1-2 Weeks of nest initiation and duration of breeding period for three focal species......4

3-1 Number of colonies inhabited by each species, 1993-2000 ...............................19

3-2 Values of FOR calculated in 30 randomly-chosen cells using constant
proportions, compared to approximate 95% confidence intervals from a
bootstrap analysis involving 1000 iterations where proportions were allowed to
ran d o m ly v ary ......................................................................................................... 2 0

3-3 Logistic regression models tested for Great Blue Herons and associated
statistical v alu e s........................................................................................................ 2 1

3-4 Logistic regression models tested for Great Egrets and associated statistical
v alu e s ...................................................................................................... ........ .. 2 1

3-5 Logistic regression models tested for Tricolored Herons and associated
statistical v alu es........................................................................................................ 2 2

3-6 Best models for each species and associated measures of classification
p erfo rm an ce .............................................................................................................. 2 3

3-7 Relative importance of variables to each species................................................23















LIST OF FIGURES


Figure page

2-1 Southeastern Florida showing the location of WCA3 within the larger
Everglades ecosystem .............. ................. ................................................ 14

2-2 Everglades Landscape M odel cells in W CA3 ..................................... ................ 15

2-3 Algorithm for determining which models to test for each species........................ 16

3-1 D distribution of ardeid colonies in W CA 3 ........................................... ................ 24

3-2 Relationship of Great Blue Heron colonies relative to all available sites in
WCA3 as shown by linearized Ripley's K (l[d]) graphs ....................................26

3-3 Relationship of Great Egret colonies relative to all available sites in WCA3 as
shown by linearized Ripley's K (l[d]) graphs..................................... ................ 30

3-4 Relationship of Tricolored Heron colonies relative to all available sites in
WCA3 as shown by linearized Ripley's K (l[d]) graphs ....................................34















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

THE RELATIVE INFLUENCES OF PREDATION AND PREY AVAILABILITY ON
ARDEID BREEDING COLONY SITE SELECTION

By

Matthew John Bokach

May 2005

Chair: Peter Frederick
Major Department: Natural Resources and Environment

Nest predation and prey availability are two of the most important factors affecting

breeding success in long-legged wading birds (Ciconiiformes). I investigated whether the

breeding colony site selection of three species of herons and egrets (family Ardeidae) in

the Florida Everglades was influenced by environmental characteristics that mediated

these two factors. These characteristics were 1) likelihood that sites remained inundated

throughout the breeding period; 2) amount of foraging habitat around sites; 3) average

weekly proportional decline in water depths around sites throughout the breeding period;

4) spatial variation in water depths around sites at the time of nesting. Variables were

calculated within a geographic information system using both raster and vector inputs. I

used measures derived from the Akaike information criterion to select the best logistic

regression model and to evaluate the relative importance of these four variables for

colony site selection in each species.









Amount of foraging habitat and likelihood of remaining inundated were the most

important variables influencing colony site selection by all three species. Great Blue

Herons (Ardea herodius) also selected sites with high rates of average weekly

proportional declines in water depth, but it is likely this variable was a proxy for deep

water for this species rather than a reflection of prey availability. Overall, these species

seemed to favor stable (rather than variable) hydrological conditions. This might also

indicate that their colony site selection is based on conditions at the time of nesting rather

than an attempt to predict future conditions. These results confirm the importance of

managing the Everglades to maximize the extent of slough habitats, if the goal is to

increase breeding populations of wading birds therein.














CHAPTER 1
INTRODUCTION

Two of the most important factors affecting the breeding success of long-legged

wading birds (Ciconiiformes) are egg and chick predation, and the availability of

adequate food to raise chicks to fledging (Taylor & Michael 1971; Frederick & Collopy

1989a; Frederick & Spalding 1994; Frederick 2002). Although wading birds cannot

control the behaviors of predators or their prey, it is likely that selection has favored the

recognition of breeding colony sites with beneficial environmental characteristics that

mediate or constrain those behaviors.

For example, wading birds nearly always nest on islands or in trees and/or shrubs

that are inundated at their base, a pattern that has most often been interpreted as a strategy

for deterring mammalian predators (Rodgers 1987; Bancroft et al. 1988; Frederick &

Collopy 1989b; Smith & Collopy 1995). However, whether this pattern represents actual

selection for this characteristic has never been quantitatively studied. With regard to prey

availability, although previous studies (Gibbs et al. 1987; Gibbs 1991; Gibbs & Kinkel

1997; Baxter & Fairweather 1998; Bancroft et al. 2002) have shown that wading birds

select colony sites that maximize the amount of wetland habitats within a reasonable

foraging range, other environmental characteristics (e.g., water depth) are better

determinants of wading birds' ability to capture prey.

If wading bird colony site selection represents an attempt to deter predators and/or

maximize prey availability, then used sites should be measurably different from available

unused sites with relation to environmental characteristics that affect these two factors. I









identified one characteristic that deters predators from reaching nests and three

characteristics that are likely to influence prey availability around sites (Table 1-1) and

make the following predictions.

Prediction 1: Used sites will have a greater likelihood of remaining inundated

throughout the breeding period than unused sites. This characteristic seems to be a

good measure of a site's ability to deter mammalian predators (Frederick & Collopy

1989b; Smith & Collopy 1995), which are generally the most destructive in their effects

on wading bird colonies (Rodgers 1987; Post 1990; Smith & Collopy 1995).

Prediction 2: Used sites will be surrounded by more open or sparsely

vegetated habitats than unused sites. Not all wetland habitats are appropriate foraging

grounds for wading birds. In particular, they are known to avoid dense vegetation (such

as monospecific stands of sawgrass [Cladiumjamaicense] or cattail [Typha latifolia]),

which interferes with their visual or tactile hunting techniques, provides prey animals

with more hiding places, and could serve as a hiding place for predators of the birds

themselves (Hoffman et al. 1994; Smith et al. 1995; Surdick 1998).

Prediction 3: Used sites will be located in areas that experience greater

proportional declines in water depth throughout the breeding season than unused

sites. Wading birds are generally limited to foraging in water that is shallower than the

length of their bills or legs (Custer & Osborn 1978; Powell 1987). They are often

attracted to areas of declining water depth where prey have been concentrated into pools

or depressions that remain inundated late in the season (Kushlan 1976; Bancroft 1989;

Frederick & Collopy 1989a; Smith 1995; Gawlik 2002; but see Frederick & Spalding

1994 for a critique of this idea).









Prediction 4: Variability in water depths at the time of nesting will be greater

around used sites than around unused sites. Water levels are dynamic and

unpredictable in most wetlands. Heterogeneous topography around a colony site could

therefore provide a foraging advantage to wading birds as this would provide the most

diversity of potential foraging sites at almost any time or water condition (Kahl 1964;

Bancroft et al. 2002).

I used logistic regression models to assess the relative influence of these four

variables on the colony site selection of three wading bird species in the ciconiiform

family Ardeidae (Table 1-2) over a period of 8 years in the Florida Everglades. Measures

derived from the models' Akaike information criteria (AIC) (Akaike 1973; Burnham &

Anderson 2002) were used to choose the best model and determine the relative influence

of the variables on the colony site selection for each species.









Table 1-1. Variables and methods of calculz
Variable Abbreviation
Affecting predator deterrence
Likelihood that a
cell will remain
inundated for
the duration of
the breeding
period
Affecting prey availability
Amount of foraging
vegetation around FOR
sites


Average rate of
proportional
decline in depths
around a site for
the duration of a
species' breeding
period


FWI


Method of calculation

Used GREATERTHAN function to
output number of weeks over duration
of a species' breeding period that
depths remained above 0; divided by
number of weeks to yield values
between 0.0 and 1.0


See text for details


Calculated proportional decline in water
depths for every week during a
species' breeding period (rises in depth
expressed as 0); averaged these values
over all weeks; used the
FOCALMEAN function to average
these cell averages over 3x3 cell
neighborhoods, yielding values
between 0.0 and 0.5


Spatial variation in
water depths Used FOCALSTD function to calculate
around a site at WV standard deviation of depths in 3x3
week of nest cell neighborhoods
initiation
All calculations carried out in ESRI Arc/INFO workstation v.8.3 except FOR, which was
calculated in ESRI ArcMap v.8.3. Weeks of nest initiation and duration of breeding
periods are listed in Table 1-2.

Table 1-2. Weeks of nest initiation and duration of breeding period for three focal species
Week of nest
Species initiationa Duration Source for duration
initiation


Great Blue Heron
(Ardea herodius)
Great Egret 9th
(Ardea alba)
Tricolored Heron
(Egretta tricolor)
aSource is Frederick (pers. comm.)


14 weeks Butler 1992
10 weeks McCrimmon et al. 2001

11 weeks Frederick 1997














CHAPTER 2
METHODS

Study Area

I studied ardeid colony site selection within Water Conservation Area (WCA) 3, a

2,350-km2 human-made impoundment in the central Everglades (Figure 2-1). Most of the

wading birds that breed within the Everglades ecosystem have selected colony sites

within this impoundment since the early 1970s (Ogden 1994; Frederick & Ogden 2001).

The annual hydrology of WCA3 is characterized by a decline in water depths during the

dry season between November and April, and most annual rainfall occurring during a

subtropical wet season between June and November. The northern end of the

impoundment is shallow and quick to dry, while the southern end is almost permanently

inundated. This same gradient exists to a lesser extent from west (where flow of water

into the adjoining Big Cypress National Preserve is unimpeded) to east (the enclosed sub-

impoundment WCA3B created by the L-67 canals, Figure 2-1). Wading birds therefore

have a wide range of hydrological conditions within which to select sites in WCA3.

The vegetation of WCA3 is characterized by open "wet prairie" communities in

deeper areas interspersed between narrow ridges covered in cattail and/or sawgrass

(Gunderson 1994). The patches of woody or shrubby vegetation that serve as potential

nesting sites for wading birds are scattered throughout. Large areas in the northern end of

the impoundment are dominated by an almost complete monoculture of cattails.









Data Sources

Water Depth Grids

I derived hydrological variables using estimated water depths from the Everglades

Landscape Model (ELM; Fitz et al. 2004), which simulates the flows and stages of water

across the entire Everglades ecosystem at a resolution of 1 km2. Model output was

available for the period 1993-2000, so these 8 years comprised the duration of my study.

Each output layer is an Arc/Info grid (Environmental Systems Research Institute 2002)

whose values are the estimated weekly average above-ground depth of water in each cell.

The model generally performs well both spatially and temporally when compared to

actual depths measured throughout the landscape (Fitz et al. 2004). The largest errors

occur in cells containing canals, as these contain the most intracellular variation in water

depths. ELM 1 km2 cells, rather than discrete colony sites, were my units of analysis as

these were the coarsest in resolution of all the data.

Vegetation Maps

I used a polygon vector geographic information system (GIS) layer showing the

distribution of vegetation in the central and southern Everglades and based on aerial

infrared photography acquired in 1995 over the entire area of WCA3, Everglades

National Park, and Big Cypress National Preserve. The latter two areas were included in

my analysis since they are within the typical foraging range of birds nesting near the

boundaries of WCA3 (Bancroft et al. 1994). The distinctive reflectance signatures of

different types of vegetation within the aerial photography were used to digitize polygons

representing discrete patches. These could contain up to three specific vegetative classes,

assigned hierarchically. After comparing the polygon layer to 1999 United States

Geological Survey (USGS) digital orthophoto quarter quads (DOQQs), I was satisfied









that this layer was an accurate depiction of the vegetation in WCA3 for the entire time

period of this study.

Since all reports of nesting by ardeid wading birds are on trees or shrubs in WCA3,

I defined potential colony sites as polygons that contained at least one class of woody

vegetation. Among the classifications used in the vegetation map these included: all

variants of forest, all variants of shrublands, Melaleuca, Brazilian pepper, disturbed fish

camp sites, spoil areas, and artificial "deer islands." I aggregated contiguous polygons

that contained at least one of these vegetation types into single discrete patches, which I

hereafter refer to interchangeably as "patches of woody vegetation" or "potential colony

sites."

The locations of wading bird colonies (see below) revealed that a small proportion

(-1.1%) of the potential colony sites did not occur within a reasonable distance from a

patch of woody vegetation. In October 2002 I visited all such sites in WCA3B and in

WCA3A north of Highway 1-75 (Figure 2-1), and determined that in most cases either

some woody vegetation had not been included in an existing polygon, or that a single tree

or small patch of shrubs was present and should have itself been digitized as a polygon.

In these cases, I therefore either added a woody vegetation classification to the second or

third tier of the existing polygon, or digitized the patch myself using the 1999 USGS

DOQQs.

Water model cells that either contained no potential colony sites, or contained only

the "clipped" edges of sites whose bulk were in an adjacent cell, were excluded from

further analysis since wading birds could not have nested in them. At the scale at which I









calculated the variables (see below), this removed less than 3% of the total area of WCA3

from analysis (Figure 2-2).

Wading Bird Colony Locations

Locations, species composition, and size of wading bird colonies were determined

using both aerial and ground survey techniques (Frederick et al. 1996). Aerial surveys

were designed such that the entire ground surface of the study area was visible along at

least one east-west transect. These were spaced 2.9 km apart, flown at 245 m altitude, and

were conducted monthly between January and June of each year. Dark-colored species

are often not visible in aerial surveys (Frederick et al. 1996), and for these species

ground-based airboat searches were conducted. All tree islands in the study area were

approached and searched for signs of nesting in the middle to late part of the nesting

season (April through May). Colony site coordinates were determined by handheld GPS

receivers, and populations of every species present were estimated by counting nests

and/or number of adults that flushed from the site. In cases where a colony was visited

more than once per season, I used the peak count of nests at that colony for a given

species.

I created point vector GIS layers representing the colony locations for each of the

three focal species for every year. Given that point coordinates reflected the location of

the airboat or airplane at the time of data collection, rather than the actual location of the

colony, I used a spatial join operation to shift points to the centroid of the nearest patch of

woody vegetation. In general, these shifts were not more than 100 m for points collected

on the ground, or 300 m for points collected during aerial surveys. I then overlaid the

water model grid and classified all cells as either used or unused by each of the species

for every year.









Distributional Relationships

To rule out the possibility that the distribution of these species' colonies was

simply a reflection of the distribution of available sites, I calculated Ripley's K statistic

for each species-year pattern of colony sites as well as for the parent pattern of all

available sites, at 500 m intervals from 500 to 20,000 m. This statistic is a

scale-dependent measure of spatial dispersion within a defined area (Fotheringham et al.

2000). However, it can also be used to assess whether two patterns within the same area

are spatially similar, especially when one is a subset of the other. Differences between

two patterns' K statistics (i.e., the linearized Ripley's K, or l[d]) indicate that they are

spatially dissimilar, with positive values of 1(d) indicating a dispersed pattern relative to

the parent pattern, and negative values of 1(d) indicating a clustered pattern relative to the

parent pattern. Since there is no significance test to determine how different from 0 an

1(d) value must be to indicate that the patterns are different, I used bootstrapping to create

sample-size-dependent 95% confidence intervals against which to compare each species-

year point pattern (Fotheringham et al. 2000).

Calculation of Variables

All variables were calculated at the scale of a 3x3 neighborhood of cells (i.e., 3 km

x 3 km) centered on the cell in question except INUN, which was calculated at the scale

of the cell itself. I chose these small scales to maximize the spatial independence

between samples, since larger neighborhoods would have increased the distance required

between two cells before their neighborhoods did not overlap. The calculation of the

three hydrological variables is described in Table 1-1.

To calculate FOR I first identified all polygons in the vegetation map that contained

any of the foraging habitat types. Within the map's classification system these included:









all variations of savanna; spike rush and maidencane-spike rush prairies; all variations of

non-graminoid emergent marsh; open water; mixed mangrove scrub; and, buttonwood

and saw palmetto scrubs.

Vegetation classes were hierarchically assigned to polygons qualitatively (e.g., a

large area of wet prairie might contain some scattered shrubs), yet I needed a quantitative

estimate of the areal extent of each class within polygons. Since vegetation classes could

be assigned to any of the tiers within a polygon, with adjoining polygons often having the

same classes in reversed order, the least biased method of estimating these extents was to

allocate the same proportion to a given tier in every polygon. In polygons containing two

classes of vegetation, I allocated 70% of the area to the dominant vegetation type and

30% to the secondary; in three-class polygons, I allocated 50, 30, and 20% of the area to

the three tiers respectively. After allocating a polygon's area between its vegetation

classes, I calculated the total extent of foraging habitat present within the polygons. I

intersected these polygons with a square grid polygon layer corresponding to the ELM

grid cell boundaries to assign the vegetation polygons to the cell they inhabited, and

summed the areas of foraging vegetation present within every cell. I recalculated the area

of foraging vegetation in polygons that had been split by cell boundaries by multiplying

their new area by the same proportion of foraging vegetation calculated for the parent

polygon. Finally, I calculated the 3x3 neighborhood sums for every cell in WCA3.

To ensure that the proportions I chose did not bias the ultimate values of FOR, I

randomly chose 30 cells and performed a bootstrap analysis whereby the proportions

allocated to vegetation classes within their polygons were allowed to randomly vary.

Since the vegetation classes were originally assigned hierarchically, I constrained the









proportions such that the dominant class proportion was always greater than the

secondary, and this was always greater than the tertiary. I calculated FOR for each cell

1000 times based on these randomly-determined proportions and used these values to

create an approximate 95% confidence interval for the "real" value of FOR in each cell. I

then determined how many of the values of FOR calculated in the 30 cells using the

proportions described above fell within these confidence intervals.

Multivariate Analyses

I used the Number Cruncher Statistical Systems software (NCSS, Hintze 2001) to

evaluate logistic regression models where the response variable was the use (presence of

at least one colony) or non-use of a water model cell by a species. Model inputs consisted

of the values for all cells used by a given species in all 8 years combined with an equal

number of randomly-selected unused cells from each year. I bootstrapped these

randomly-selected sets for every species-year to ensure that their mean values for all four

variables were within the 95% confidence interval suggested by the bootstrap.

Results of the logistic regression analyses were interpreted within the information-

theoretic approach of Burnham and Anderson (2002). Multiple models were evaluated

and the Akaike information criterion (AIC; Akaike 1973) calculated for each one thus:

AIC = -2(log-likelihood ratio) + 2K (2-1)

where K was the number of parameters estimated in the model. The model with the

lowest AIC was the "best" model among all those evaluated. The last term in Eq. 2-1

therefore represented a "penalty" against a model's AIC value for every variable added to

the model.









The performance of a given model relative to the best one in the set was expressed

using three values derived from the AIC values. The AIC difference (A,) for model i was

calculated:

A, = AIC, AICm,, (2-2)

This was converted to "Akaike weights" (w,):

R
w, = exp(-1A,)/ exp(- A,) (2-3)
r=l

where the denominator in Eq. 2-3 was the sum of all models' A, values. The w, for model

i was a measure of the likelihood (often expressed as a percent) that it is the best model.

An analogous quantity is the evidence ratio for model i:

evidence ratio, = w, / wj (2-4)

where w, was the w, value for the best model in the set. The evidence ratio was a measure

of "how much" better the best model was than model i. Because they relied on the sum of

all models' A, values, the w, and evidence ratios were recalculated every time a new

model entered the set.

I followed the algorithm shown in Figure 2-3 to determine the models tested for

each species. Upon identifying the best model for each species, I report three measures of

its classification performance (Fielding & Bell 1997): 1) the number of used and unused

cells correctly classified; 2) the Kappa statistic (K), which measures the extent to which a

model's predictions are correct above the proportion expected simply by chance; and, 3)

the area under the model's ROC curve, which measures the likelihood that the value

assigned by the final model to a randomly-selected "used" cell will be greater than that of

a randomly-selected "unused" cell. I also summarized the relative importance of the






13


variables tested for each species by adding the wi values for all models in which a given

variable is included (Burnham & Anderson 2002: 167-168).




















MIiccosukee
Indian
Reservation


t Palm
11


Iort
Lauderdale


Miami


K lometers
0 10 20 30 40


Figure 2-1. Southeastern Florida showing the location of WCA3 within the larger
Everglades ecosystem. The dotted areas along the east coast show the extent
of urban landscapes and their proximity to what is left of the Everglades.












**.** **- ********** ****o****** *







*094 9 0*
****** *** :*:*::* : *******:
^ ***** ^:******* :* :** ** *** *





9*0*0*9I 0 9099EEEEEEE a
0||||||||||||||||||||||0,E E E
******* 0** *060* *
9999 0*0*0 000 R IM 99


Kilometers
0 5 10 15


Figure 2-2. Everglades Landscape Model cells in WCA3. Black dots indicate cells that
contained potential colony sites and were included in multivariate analysis.
The darker shaded area indicates all cells that were within a 3x3 cell
neighborhood of analyzed cells and whose characteristics were therefore
included in the calculation of variables. Less than 3% of the area of WCA3,
represented by the unshaded areas, was excluded from analysis.











Test saturated model (all 4
variables) and all 3-term models.


Test all 2-term models
nested within best model


Test all 1-term models
nested within best model


BEST MODEL


Test all models comprised of best model
+ single second-order interaction term
between component variables.


Figure 2-3. Algorithm for determining which models to test for each species. At every
step the current best model competed with lower-term models comprised of its
component variables. Once the best single-order model was identified, models
that added a single second-order interaction term were tested.


C'J














CHAPTER 3
RESULTS

Distribution of Colony Sites

Table 3-1 summarizes the number of colonies and water model cells inhabited by

each species in each year, and Figure 3-1 presents the annual maps of these distributions.

Figures 3-2 through 3-4 show the l(d) graphs for each species-year combination, along

with their corresponding approximate 95% confidence intervals from the bootstrap

simulation. Great Blue Heron colonies were extremely clustered relative to the parent

pattern of all potential colony sites in all years except 1998, when the l(d) values were

closer to the lower bootstrap limit. Great Egret colonies were dispersed relative to the

parent pattern at all scales in 1993, 1994, and 1995; at small scales in 1996, and at large

scales in 1998. Their colony patterns were mostly indistinguishable from the parent

pattern at practically all scales in 1997, 1999, and 2000. Tricolored Heron colony patterns

showed the most variability: they were highly dispersed relative to the parent pattern at

medium to large scales in 1993; highly clustered at all scales in 1996; slightly clustered at

small scales in 1994 and medium scales in 1999 and 2000; and, otherwise mostly

indistinguishable from the parent pattern in 1995, 1997 and 1998.

Bootstrap of Foraging Habitat Calculation

Table 3-2 lists the values of FOR calculated for 30 randomly-chosen cells using the

constant proportions described above and the upper and lower values of the approximate

95% confidence interval from the bootstrap analysis. Four of the 30 cells contained only

polygons that did not vary in their values of FOR. In only one of the remaining cells was









the value of FOR calculated using the constant proportions outside the 95% confidence

interval from the bootstrap; however, this cell's value was within the 99% confidence

interval. This is strong evidence that the constant proportions used to calculate this

variable did not bias the analysis.

Multivariate Analyses

I tested 11 models for Great Blue and Tricolored Herons, and 12 for Great Egrets

(Tables 3-3 through 3-5). A single best model (w, > 98%) was found for the first two of

these species, while the best two models for Great Egrets were nearly indistinguishable

(w, values of 29.20% and 23.18%, respectively). Table 3-6 lists the best models) for each

species and its associated measures of classification performance. Best models correctly

classified between 68.8% (Great Egrets, model 1) and 78.9% (Great Blue Herons) of the

used sites, and between 36.7% (Great Egrets, model 1) and 56.7% (Great Blue Herons) of

the unused sites. Kappa statistics were routinely low, ranging between 0.055 (Great

Egrets, model 1) and 0.357 (Great Blue Herons). The areas under the ROC curves also

indicated relatively poor model performance; these ranged between 0.524 (Great Egrets,

model 1) and 0.745 (Great Blue Herons).

The amount of foraging habitat (FOR) around potential sites was always the most

important variable, with all Ew, values greater than 98% (Table 3-7). It was followed by

the likelihood that a site would remain inundated (INUN), with Yw, values ranging

between 83.88 and 99.99%. As predicted, all three species responded to both of these

variables positively, i.e., used cells had higher values overall than unused cells. The

pattern for the remaining variables was less consistent across species, with Great Blue

Herons responding strongly and positively to FWI, and Great Egrets responding weakly

and negatively to WV (Table 3-7).






19

Table 3-1. Number of colonies inhabited by each species, 1993-2000. Number of water
model cells inhabited in parentheses.
Great Blue Heron Great Egret Tricolored Heron
1993 156(127) 32(32) 11 (11)
1994 200 (153) 45(44) 46(43)
1995 298(229) 39(39) 40(39)
1996 167(136) 45(43) 52(42)
1997 101 (90) 34 (33) 8 (8)
1998 110(97) 44(41) 23(23)
1999 258(206) 77(71) 66(63)
2000 309 (235) 71 (62) 57 (53)









Table 3-2. Values of FOR calculated in 30 randomly-chosen cells using constant
proportions, compared to approximate 95% confidence intervals from a
bootstrap analysis involving 1000 iterations where proportions were allowed
to randomly vary.
Lower limit of Value of FOR Upper limit of Observed within
95% confidence calculated using 95% confidence confidence
interval constant proportions interval interval?
75 75 75 Constant value cell
251 339 644 Yes
750 750 750 Constant value cell
118 887 1,155 Yes
149 1,742 2,389 Yes
4,321 4,321 4,321 Constant value cell
5,925 5,925 5,925 Constant value cell
9,763 10,296 12,057 Yes
1,526 20,539 29,209 Yes
24,992 25,858 26,380 Yes
20,804 35,765 42,190 Yes
74,514 80,693 84,326 Yes
101,367 103,625 119,108 Yes
146,646 149,072 150,791 Yes
172,177 218,953 244,864 Yes
163,585 256,693 339,316 Yes
281,378 286,925 305,647 Yes
28,983 289,828 328,472 Yes
337,236 352,078 356,356 Yes
115,366 360,059 418,467 Yes
77,850 376,847 504,702 Yes
92,661 382,298 501,854 Yes
373,516 445,018 635,874 Yes
220,282 470,162 569,560 Yes
413,803 477,840 597,608 Yes
271,067 504,240 587,148 Yes
307,407 507,071 567,364 Yes
385,418 532,264 559,474 Yes

504,530 607,378 604,063 No, 99%i
confidence interval
531,142 638,535 646,606 Yes











Table 3-3. Logistic regression models tested for Great Blue Herons and associated statistical values. Italicized models are the best
from their subset, while the best overall model is shown in boldface.


Model Log-Likelihood
FOR+WV+FWI+INUN -1492.38
FOR+WV+FWI -1495.94
FOR+ FWI+ INUN -1493.21
FOR+WV+INUN -1495.21
WV+FWI+INUN -1721.50
FOR+FWI -1496.75
FOR+INUN -1496.40
FWI+INUN -1745.57
FOR+FWI+INUN+FOR*FWI -1492.98
FOR+FWI+INUN+FOR*INUN -1493.15
FOR+FWI+INUN+FWI*INUN -1484.30
a,"-, indicates a w, value < 0.01%; b "-,, indicates


K AIC AAIC
5 2994.76 16.15
4 2999.88 21.28
4 2994.42 15.81
4 2998.42 19.81
4 3451.00 472.40
3 2999.51 20.90
3 2998.81 20.21
3 3497.14 518.54
5 2995.96 17.36
5 2996.29 17.69
5 2978.60 0.00
an evidence ratio > 100


w1a Evidence Ratiosb


0.04%


0.02%
0.01%
99.89%


1.00


Table 3-4. Logistic regression models tested for Great Egrets and associated statistical values. Italicized models are the best from their
subset, while the best overall models are shown in boldface.


Model Log-Likelihood
FOR+WV+FWI+INUN -498.86
FOR+WV+FWI -499.97
FOR+FWI+INUN -499.59
FOR+ WV+ INUN -498.86
WV+FWI+INUN -502.60
FOR+WV -500.48
FOR +INUN -499.61
WV+INUN -502.69
FOR -501.08
INUN -503.99
FOR+INUN+FOR*INUN -497.75
FOR+WV+INUN+FOR*INUN -496.98
b "-" indicates an evidence ratio > 100


AIC
1007.73
1007.94
1007.17
1005.73
1013.20
1006.96
1005.21
1011.38
1006.16
1011.99
1003.51
1003.97


AAIC
4.22
4.44
3.67
2.22
9.69
3.46
1.71
7.88
2.65
8.48
0.00
0.46


w,
3.54%
3.17%
4.67%
9.62%0
0.23%
5.18%
12.45%o
0.57%
7.76%
0.42%
29.20%
23.18%


Evidence Ratiosa
8.25
9.20
6.26
3.03

5.63
2.35
51.35
3.76
69.43
1.00
1.26


Rank
8
9
7
4
12
6
3
10
5
11
1
2


Rank
3
9
2
6
10












Table 3-5. Logistic regression models tested for Tricolored Herons and associated statistical values. Italicized models are the best from
their subset, while the best overall model is shown in boldface.


Model
FOR+WV+FWI+INUN
FOR+WV+FWI
FOR+FWI+INUN
FOR WV+ INUN
WV+FWI+INUN
FOR+WV
FOR +INUN
WV+INUN
FOR
INUNUN+FOR*INUN
FOR+INUN+FOR*INUN


Log-Likelihood
-363.45
-364.24
-363.83
-363.50
-387.66
-364.70
-363.92
-387.67
-365.35
-390.82
-357.38


AIC
736.90
736.48
735.66
734.99
783.33
735.40
733.84
781.33
734.70
785.64
722.76


A AIC
14.14
13.72
12.90
12.23
60.57
12.64
11.07
58.57
11.94
62.88
0.00


wa
0.08%
0.10%
0.16%
0.22%

0.18%
0.39%

0.25%

98.62%


Evidence Ratiosb


Rank
8
7


1.00


a" indicates a w, value < 0.01%; o "-" indicates an evidence ratio > 100









Table 3-6. Best models for each species and associated measures of classification
performance.
Great Blue Herons


Model
n, % Correct (Used)
n, % Correct (Unused)
K Statistic
Area Under ROC Curve
Great Egrets
Model 1
n, % Correct (Used)
n, % Correct (Unused)
K Statistic
Area Under ROC Curve
Model 2
n, % Correct (Used)
n, % Correct (Unused)
K Statistic
Area Under ROC Curve
Tricolored Herons
Model
n, % Correct (Used)
n, % Correct (Unused)
K Statistic
Area Under ROC Curve


FOR + FWI + INUN FWI*INUN
1005, 78.9%
722, 56.7%
0.357
0.745

FOR + INUN + FOR*INUN
251, 68.8%
134, 36.7%
0.055
0.524
FOR WV + INUN + FOR*INUN
274, 75.1%
152, 41.6%
0.167
0.559

FOR + INUN + FOR*INUN
218, 77.3%
137, 48.6%
0.259
0.654


Table 3-7. Relative importance of variables to each species.


FOR
WV
FWI
INUN
FOR*FWI
FOR*INUN
FWI*INUN


Great Blue Heron
100.00%
0.04%
99.99%
99.99%
0.02%
0.01%
99.89%


Great Egret
98.78%
45.50%
11.61%
83.88%

52.38%


Tricolored Heron
100.00%
0.58%
0.34%
99.47%

98.62%












A B











\


4v
v T v



v D
vv IT T 0
V v T v *




v 7 v v V


















V / **,* *
v v *
T *V



19*. If B u
C D
\ V\









Fig b o a o. B 1 C 1 D








Heron... c ic G Eto, ad c i






T rivcoo Ire Heo coones
| N 5 10 1 |






























*














v*
v v v


T v

v V


Figure 3-1. Continued


F


*\

\




















0 *
S v v
* *











V V i


*





*



*


v
V *
* .


*


10 15














0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
-1


0 5000


0.6
0.4
0.2
0
-0.2
-0.4

-0.6
-0.8
-1
-1.2


5000


Figure 3-2. Relationship of Great Blue Heron colonies relative to all available sites in
WCA3 as shown by linearized Ripley's K (l[d]) graphs. A) 1993. B) 1994. C)
1995. D) 1996. E) 1997. F) 1998. G) 1999. H) 2000. Sample sizes for each
year listed in Table 3-1. Dashed lines: 95% confidence intervals from Monte
Carlo simulations of 999 identically-sized, randomly-selected samples from
each year; solid line: 1(d) graph for used sites.


10000


15000


Scale (m)


20000


-
p
I
I


10000


15000


Scale (m)


20000












0.4
0.2
0
-0.2
-0.4
-0.6
-0.8


5000


0.6
0.4
0.2
0
S-0.2
-0.4
-0.6
-0.8
-1
-1.2


5000


10000
Scale (m)


10000
Scale (m)


Figure 3-2. Continued


-- .~ ~. -

-




2>


15000


20000


-- ------


15000


20000















1

0.5

0

-0.5

-1

-1.5

-2


5000


0.8
0.6
0.4
0.2
0
-0.2
-- -0.4
-0.6
-0.8
-1
-1.2
-1.4


5000


10000

Scale (m)


10000

Scale (m)


Figure 3-2. Continued


-

-
I
I
A
I


15000


20000


-


15000


20000














0.4

0.2

0

-0.2

-0.4

-0.6

-0.8


5000


10000


15000


Scale (m)



H


5000


10000


15000


Scale (m)


Figure 3-2. Continued


-
-

-
I


7


20000


0.4
0.3
0.2
0.1
0
-0.1
- -0.2
-0.3
-0.4
-0.5
-0.6


-
-

-
-
I


20000














2
1.5
1
0.5
S 0
-0.5
-1
-1.5
-2


0 5000 10000 15000


Scale (m)



B


2
1.5
1
0.5
S0
-0.5
-1
-1 -
-1.5
-2


5000


10000


15000


Scale (m)


Figure 3-3. Relationship of Great Egret colonies relative to all available sites in WCA3 as
shown by linearized Ripley's K (l[d]) graphs. A) 1993. B) 1994. C) 1995. D)
1996. E) 1997. F) 1998. G) 1999. H) 2000. Sample sizes for each year listed
in Table 3-1. Dashed lines: 95% confidence intervals from Monte Carlo
simulations of 999 identically-sized, randomly-selected samples from each
year; solid line: 1(d) graph for used sites.


20000


I -
' I
I t
'I


20000














1.5

1

0.5

0

-0.5

-1
-1 -

-1.5


5000


10000


15000


Scale (m)



D


5000


10000


15000


Scale (m)


Figure 3-3. Continued


I # -~
' /


20000


1.5

1-

0.5

0

-0.5

-1 -

-1.5


I
1




S- i

i J
\ t
iO


20000














1.5

1

0.5

0

-0.5

-1

-1.5
-2


5000 10000 15000


Scale (m)



F


0 5000 10000 15000


Scale (m)


Figure 3-3. Continued


- - - B-


20000


20000













1

0.5

0

-0.5

-1

-1.5









1

0.5

0

-0.5

-1

-1.5


0 5000 10000 15000

Scale (m)


0 5000 10000 15000

Scale (m)


Figure 3-3. Continued


20000


20000














1.5

1

0.5
S0

-0.5
-1

-1.5
-2


5000


10000


Scale (m)



B


10000


Scale (m)


Figure 3-4. Relationship of Tricolored Heron colonies relative to all available sites in
WCA3 as shown by linearized Ripley's K (l[d]) graphs. A) 1993. B) 1994. C)
1995. D) 1996. E) 1997. F) 1998. G) 1999. H) 2000. Sample sizes for each
year listed in Table 3-1. Dashed lines: 95% confidence intervals from Monte
Carlo simulations of 999 identically-sized, randomly-selected samples from
each year; solid line: 1(d) graph for used sites.


- -d
% s,
i


15000


20000


1.5

1

0.5

0

-0.5

-1 -

-1.5


.



1 -


*


5000


15000


20000















2.5

1.5

o 0.5
*0
- -0.5

-1.5

-2.5


5000


1.5
1

0.5
- 0-
0

S-0.5

-1

-1.5
-2


10000

Scale (m)



D


5000


10000

Scale (m)


Figure 3-4. Continued


1

i- -- %_.- ..-.---------------..........
J m m

-
I


15000


20000


t
-


J.. .- ... _me i m e eJ


! !

.


15000


20000















4.5

3.5

2.5

. 1.5

^ 0.5

-0.5

-1.5

-2.5


5000


10000


15000


Scale (m)



F


5000


10000


15000


Scale (m)


Figure 3-4. Continued


t
- <.

V --


-t cf"^^
-* -

---.-
~t *. me- e mm--


20000


1~ ~


2.5
2
1.5
1
S0.5
g 0-

-0.5
-1
-1 -
-1.5
-2


20000














1.5

1

0.5

S0

-0.5

-1
-1 -

-1.5


5000


1.5

1

0.5

S0

-0.5
-1
-1 -

-1.5


5000


10000

Scale (m)


10000

Scale (m)


Figure 3-4. Continued


- -


15000


20000


-
-



'5
V


15000


20000














CHAPTER 4
DISCUSSION

Effects of Site Availability

The spatial patterns of colony sites used by these three species during the study

period were in general dissimilar from the overall pattern of available sites (Figures 3-2

through 3-4). This was always true for Great Blue Herons, and was true at most scales for

Great Egrets and Tricolored Herons in 4 and 5 of the 8 years, respectively. These two

species exhibited both clustered and dispersed patterns, which likely reflects responses by

the birds to different hydrological conditions in each year. These results indicate that site

selection by these three species was probably independent of overall site availability or

distribution.

Model Performance

Although there was very strong evidence supporting the best models for Great Blue

and Tricolored Herons, relative to the total set of models I tested, model classification

performance was poor overall. None of the best models achieved a Kappa statistic of 0.4,

which is often considered to be the minimum value for "good" performance (Fielding &

Bell 1997), and only the best model for Great Blue Herons had an area under its ROC

curve of at least 0.7. Models always did a better job of classifying used than unused cells

(Table 3-6), which suggests that there were more appropriate sites available than were

used (Fielding & Bell 1997). A major limitation of this analysis was the coarse spatial

resolution of the hydrological data, which prevented me from including variables that

operate at the scale of the sites themselves such as vegetation type or patch size. Finally,









simply classifying cells as used or unused ignored the enormous variation in the number

of Great Egret pairs nesting at sites within WCA3, which is probably why the models for

this species performed the worst.

Despite these limitations, these results clearly show that the amount of foraging

habitat around sites, and the likelihood that they would remain inundated throughout the

breeding season, were important influences on the colony site selection for these three

species. This is the first study to quantitatively establish the importance of deterring

mammalian predators as an influence on colony site selection in this group. These results

also establish that the distribution of foraging habitats is as important to wading birds in

an enormous wetland-dominated landscape such as the Everglades, as it is in more typical

terrestrial landscapes where wetlands are part of a habitat mosaic (Gibbs et al. 1987;

Gibbs 1991; Gibbs & Kinkel 1997).

Responses to Hydrological Variability

In general, neither the average weekly proportional decline in water depths (FWI)

nor the spatial variation in water depths (WV) were important influences on colony site

selection in these three species. WV was weakly selected against by Great Egrets, and not

included in the best models for either of the other two species. Only Great Blue Herons

selected cells with large values of FWI. However, areas that began the breeding season

with deeper water were likely to have the largest values of for this variable, since they

could experience proportional declines for much longer before going dry. Thus it is

unclear whether FWI actually represented prey availability for Great Blue Herons or was

instead a proxy for deep water. Great Blues also selected against cells that had both rapid

declines in water depths and a high likelihood of remaining inundated (i.e., the

INUN*FWI interaction term). These could represent either permanently shallow areas









(such as on the west side of WCA3, where water flow westward into Big Cypress

National Preserve is unimpeded), and/or areas near canals, both of which were mostly

avoided by this species (Figure 3-1).

These results suggest that ardeid wading birds are selecting colony sites that are

characterized by stable, rather than variable, water depths. Another possible interpretation

is that colony site selection by these species does not reflect an attempt to predict future

conditions (i.e., a "bet hedging" strategy), but is rather simply a response to conditions at

the time of nesting (Bancroft et al. 1994). These three species are known to be less

dependent on concentrated patches of prey created by declining water depths than

tactile/social foragers such as Wood Storks (Mycteria americana), White Ibises

(Eudocimus albus), and Snowy Egrets (Egretta thula) (Kahl 1964; Gawlik 2002), so it is

also possible that they have not experienced any selective pressure to recognize or

respond to this environmental characteristic.

Management Implications

Previous authors have speculated that the size of wading bird breeding populations

in the Everglades is dependent on the extent of open slough habitat available to them

(Bancroft et al. 1994; Bancroft et al. 2002), and the results of this study establish that this

is the most important environmental characteristic determining where the birds nest.

Hydroperiod and fire are the two largest factors controlling the distribution of wetland

vegetation (Gunderson 1994), so water management can play an important role in

maximizing the extent of open slough habitat within the Everglades. Water depths should

be maintained at a deep enough level to prevent most breeding sites from going dry, but

shallow enough for wading birds to forage. Using these guidelines could make a major






41


contribution to re-establishing historical breeding populations of wading birds in the

Everglades.















LITERATURE CITED


Akaike, H. 1973. Information theory as an extension of the maximum likelihood
principle. Pages 267-281 in B. N. Petrov and F. Csaki, editors. Second International
Symposium on Information Theory. Akademiai Kiado, Budapest.

Bancroft, G. T. 1989. Status and conservation of wading birds in the Everglades.
American Birds 43:1258-1265.

Bancroft, G. T., D. E. Gawlik, and K. Rutchey. 2002. Distribution of wading birds
relative to vegetation and water depths in the northern Everglades of Florida, USA.
Waterbirds 25:263-277.

Bancroft, G. T., J. C. Ogden, and B. W. Patty. 1988. Wading bird colony formation and
turnover relative to rainfall in the Corkscrew Swamp area of Florida during 1982
through 1985. Wilson Bulletin 100:50-59.

Bancroft, G. T., A. M. Strong, R. J. Sawicki, W. Hoffman, and S. D. Jewell. 1994.
Relationships among wading bird foraging patterns, colony locations, and
hydrology in the Everglades. Pages 615-657 in S. Davis and J. Ogden, editors.
Everglades: the ecosystem and its restoration. St. Lucie Press, Delray Beach,
Florida, USA.

Baxter, G. S., and P. G. Fairweather. 1998. Does available foraging area, location or
colony character control the size of multispecies egret colonies? Wildlife Research
25:23-32.

Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a
practical information-theoretic approach. Second edition. Springer-Verlag, New
York, New York, USA.

Butler, R. W. 1992. Great Blue Heron (Ardeas herodias). No. 25 in A. Poole and F. Gill,
editors. The Birds of North America. The Academy of Natural Sciences,
Philadelphia, Pennsylvania, USA; The American Ornithologists' Union,
Washington, D.C., USA.

Custer, T. W., and R. G. Osborn. 1978. Feeding-site description of three heron species
near Beaufort, North Carolina. Pages 355-360 in A. Sprunt IV, J. C. Ogden, and S.
Winckler, editors. Wading Birds, Research Report No. 7. National Audubon
Society, New York, New York, USA.









Environmental Systems Research Institute, 2002. Arc/Info Workstation. Version 8.3.
Redlands, California, USA.

Fielding, A. H., and J. F. Bell. 1997. A review of methods for the assessment of
prediction errors in conservation presence/absence models. Environmental
Conservation 24:38-49.

Fitz, H. C., J. Godin, B. Trimble, and N. Wang. 2004. Everglades Landscape Model
documentation: ELM v2.2. South Florida Water Management District, West Palm
Beach, Florida, USA.

Fotheringham, A. S., C. Brunsdon, and M. Charlton. 2000. Quantitative geography:
perspectives on spatial data analysis. Sage Publications, London, UK.

Frederick, P. C. 1997. Tricolored heron (Egretta tricolor). No. 306 in A. Poole and F.
Gill, editors. The Birds of North America. The Academy of Natural Sciences,
Philadelphia, Pennsylvania, USA; The American Ornithologists' Union,
Washington, D.C., USA.

Frederick, P. C. 2002. Wading birds in the marine environment. Pages 617-655 in E. A.
Schreiber and J. Burger, editors. Biology of marine birds. CRC Press, Boca Raton,
Florida, USA.

Frederick, P. C., and M. W. Collopy. 1989a. Nesting success of five ciconiiform species
in relation to water conditions in the Florida Everglades. Auk 106:625-634.

Frederick, P. C., and M. W. Collopy. 1989b. The role of predation in determining
reproductive success of colonially nesting wading birds in the Florida Everglades.
Condor 91:860-867.

Frederick, P. C., and J. C. Ogden. 2001. Pulsed breeding of long-legged wading birds and
the importance of infrequent severe drought conditions in the Florida Everglades.
Wetlands 21:484-491.

Frederick, P.C., and M. G. Spalding. 1994. Factors affecting reproductive success of
wading birds (Ciconiiformes) in the Everglades ecosystem. Pages 659-691 in S.
Davis and J. Ogden, editors. Everglades: the ecosystem and its restoration. St.
Lucie Press, Delray Beach, Florida, USA.

Frederick, P.C., T. Towles, R. J. Sawicki, and G. T. Bancroft. 1996. Comparison of aerial
and ground techniques for discovery and census of wading bird (Ciconiiformes)
nesting colonies. Condor 98: 837-840.

Gawlik, D. E. 2002. The effects of prey availability on the numerical response of wading
birds. Ecological Monographs 72:329-346.

Gibbs, J. P. 1991. Spatial relationships between nesting colonies and foraging areas of
Great Blue Herons. Auk 108:764-770.









Gibbs, J. P., and L. K. Kinkel. 1997. Determinants of the size and location of Great Blue
Heron colonies. Colonial Waterbirds 20:1-7.

Gibbs, J. P., S. Woodward, M. L. Hunter, and A. E. Hutchinson. 1987. Determinants of
Great Blue Heron colony distribution in coastal Maine. Auk 104:38-47.

Gunderson, L. H. 1994. Vegetation of the Everglades: determinants of community
composition. Pages 323-340 in S. Davis and J. Ogden, editors. Everglades: the
ecosystem and its restoration. St. Lucie Press, Delray Beach, Florida, USA.

Hintze, J. 2001. NCSS and PASS. Number Cruncher Statistical Systems. Kaysville, Utah,
USA.

Hoffman, W., G. T. Bancroft, and R. J. Sawicki. 1994. Foraging habitat of wading birds
in the Water Conservation Areas of the Everglades. Pages 585-614 in S. Davis and
J. Ogden, editors. Everglades: the ecosystem and its restoration. St. Lucie Press,
Delray Beach, Florida, USA.

Kahl, M. P., Jr. 1964. Food ecology of the Wood Stork (Mycteria americana) in Florida.
Ecological Monographs 34:97-117.

Kushlan, J. A. 1976. Wading bird predation in a seasonally fluctuating pond. Auk
93:464-476.

Kushlan, J. A. 1987. External threats and internal management: the hydrologic regulation
of the Everglades, Florida, USA. Environmental Management 11:109-119.

McCrimmon, D. A., Jr., J. C. Ogden, and G. T. Bancroft. 2001. Great Egret (Ardea alba).
No. 570 in A. Poole and F. Gill, editors. The Birds of North America. The
Academy of Natural Sciences, Philadelphia, Pennsylvania, USA; The American
Ornithologists' Union, Washington, D.C., USA.

Ogden, J. C. 1994. A comparison of wading bird nesting colony dynamics (1931-1946
and 1974-1989) as an indication of ecosystem conditions in the southern
Everglades. Pages 533-570 in S. Davis and J. Ogden, editors. Everglades: the
ecosystem and its restoration. St. Lucie Press, Delray Beach, Florida, USA.

Post, W. 1990. Nest survival in a large ibis-heron colony during a three-year decline to
extinction. Colonial Waterbirds 13:50-61.

Powell, G. V. N. 1987. Habitat use by wading birds in a subtropical estuary: implications
of hydrography. Auk 104:740-749.

Rodgers, J. A. 1987. On the antipredator advantages of coloniality: a word of caution.
Wilson Bulletin 99:269-270.

Rutchey, K., L. Vilchek, and M. Love, in review. Development of a vegetation map for
Water Conservation Area 3.









Smith, J. P. 1995. Foraging flights and habitat use of nesting wading birds
(Ciconiiformes) at Lake Okeechobee, Florida. Colonial Waterbirds 18:139-158.

Smith, J. P., and Collopy, M. W. 1995. Colony turnover, nest success and productivity,
and causes of nest failure among wading birds (Ciconiiformes) at Lake
Okeechobee, Florida (1989-1992). Pages 287-316 in N. G. Aumen and R. G.
Wetzel, editors. Ecological studies on the littoral and pelagic systems of Lake
Okeechobee, Florida (USA). E. Schweizerbart'sche Verlagsbuchhandlung,
Stuttgart, Germany.

Smith, J. P., J. R. Richardson, and M. W. Collopy. 1995. Foraging habitat selection
among wading birds (Ciconiiformes) at Lake Okeechobee, Florida, in relation to
hydrology and vegetative cover. Pages 247-285 in N. G. Aumen and R. G. Wetzel,
editors. Ecological studies on the littoral and pelagic systems of Lake Okeechobee,
Florida (USA). E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart, Germany.

Surdick, J. A., Jr. 1998. Biotic and abiotic indicators of foraging site selection and
foraging success of four ciconiiform species in the freshwater Everglades of
Florida. Thesis. University of Florida, Gainesville, Florida, USA.

Taylor, R. J., and E. D. Michael. 1971. Predation on an inland heronry in eastern Texas.
Wilson Bulletin 83:172-177.















BIOGRAPHICAL SKETCH

Matthew John Bokach completed his Bachelor of Science degree with a double

major in biology and chemistry at Adrian College (Adrian, Michigan) in 1994. He fled

the sciences for 2 years in a master's program in student affairs administration at

Michigan State University, then returned to them as a U.S. Peace Corps Volunteer at

Lukosi Government School in Hwange District, Zimbabwe. While teaching general

science at Lukosi, Matthew fell in love with the southern African avifauna and spent an

additional 2 years in Zimbabwe, mostly so he could continue watching them. He returned

to the United States in 2000 and began a master's degree in interdisciplinary ecology at

the University of Florida in 2002. He plans to move to the San Francisco Bay area,

pursue a career in GIS application development, and eventually flee the sciences once

again to pursue a career in popular music.