FORAGING QUALITY OF FLOODED AGRICULTURAL FIELDS WITHIN THE EVERGLADES AGRICULTURAL AREA FOR WADING BIRDS (CICONIIFORMES)

Transcription

1 FORAGING QUALITY OF FLOODED AGRICULTURAL FIELDS WITHIN THE EVERGLADES AGRICULTURAL AREA FOR WADING BIRDS (CICONIIFORMES) By GRANT CASHION SIZEMORE 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

2 2009 Grant Cashion Sizemore 2

3 To my family and friends 3

4 ACKNOWLEDGEMENTS I would like to thank my committee, statistical consultants, and friends and family. From my committee, I thank my advisor, Marty Main, for his belief in me as a scientist and in the project. I want to thank Elise Pearlstine for her dedication to early mornings crammed into a tiny plane. I would like to thank Richard Raid for his positive outlook, exemplary work ethic, and generous vegetable donations. Lastly, I want to thank Peter Frederick for providing me with fair, firm, and incredibly helpful advice. His knowledge about wading birds and the Everglades is truly overwhelming. I want to thank James Collee and Meghan Brennan from IFAS Statistics for their helpful consultations. I am also very grateful to my family and friends who helped me to survive the many non-academic ordeals of graduate school, especially while located at my remote field site. Most importantly, I want to thank my parents for their loving support and willingness to help out in any way possible. 4

5 TABLE OF CONTENTS page ACKNOWLEDGEMENTS... 4 LIST OF TABLES... 6 LIST OF FIGURES... 8 ABSTRACT CHAPTER 1 INTRODUCTION METHODS Study Area Wading Bird Surveys Foraging Observations Habitat Characteristics Prey Sampling Data Analyses RESULTS Wading Bird Surveys: Regional Wading Bird Surveys: Local Foraging Success in Flooded Agricultural Fields Prey Consumption in Flooded Agricultural Fields Prey in Flooded Agricultural Fields Energetics of Foraging in Flooded Agricultural Fields DISCUSSION LIST OF REFERENCES BIOGRAPHICAL SKETCH

6 LIST OF TABLES Table page 2-1 Percent coverage of the Everglades Agricultural Area by various land uses Flooding timeline for agricultural fields within the Everglades Agricultural Area during January-June Abundance of wading birds observed in aerial surveys (2008, 2009) Density of wading birds (individuals/ha) observed in the Everglades Agricultural Area during the 2008 and 2009 breeding seasons A comparison of wading bird abundance per day in flooded agricultural fields within the Everglades Agricultural Area Paired difference test (t-test) results for the mean abundance of wading birds observed in flooded rice fields Results of paired difference tests for the mean abundance of wading birds observed in ground surveys for flooded agricultural fields Environmental variables and their effects on wading bird abundance for ground surveys during the spring of 2008 and Wading bird species and densities (individuals*ha -1 ) observed in flooded agricultural fields during April-June, P-values for paired difference tests comparing the mean wading bird species richness observed in flooded rice fields by month Results of paired difference tests comparing mean wading bird species richness observed in flooded agricultural fields by month*habitat Foraging success rates for Great Egrets and Little Blue Herons in flooded agricultural fields within the Everglades Agricultural Area Paired difference test results for mean water depth by month*habitat Paired difference (t-test) results for mean water depth by month*habitat Paired difference (t-test) results for mean vegetation height by month*habitat Paired difference (t-test) results for mean water depth by month*habitat Summary of prey consumption for focal birds in flooded agricultural fields. Numbers of prey are given as a proportion of the total identified captures

7 3-16 Prey sampling results for spring breeding seasons 2008 and Summary of fish captured during 2008 and 2009 prey sampling Comparison of captures/min for Great Egrets and Little Blue Herons Comparison of fish densities within Florida Comparison of wading bird densities in South Florida for 2008 and

8 LIST OF FIGURES Figure page 2-1 Map of South Florida Total wading bird abundance for aerial surveys during the 2008 and 2009 breeding seasons Density of wading birds in the Everglades Agricultural Area during the spring breeding seasons of 2008 and The number of wading birds observed in fields, ditches, and canals in aerial surveys within the Everglades Agricultural Area during the spring breeding seasons of 2008 and Wading bird densities in fields, ditches, and canals as observed in aerial surveys within the Everglades Agricultural Area during the spring breeding seasons of 2008 and Wading birds observed in the Everglades Agricultural Area and previous months precipitation values for the Everglades, Wading birds observed in the Everglades Agricultural Area and previous months precipitation values for the Everglades, Mean Everglades water depth from March-June in comparison to wading bird abundance in the Everglades Agricultural Area in 2008 based on over 2,000 stations located in Everglades National Park Mean Everglades water depth from January-June in comparison to wading bird abundance in the Everglades Agricultural Area in 2009 based on over 2,000 stations located in Everglades National Park The number of wading birds observed in scans of flooded agricultural fields during the spring breeding seasons in 2008 and A comparison of abundance by month for wading birds in flooded rice and fallow fields of the Everglades Agricultural Area Wading bird species richness in flooded agricultural fields within the Everglades Agricultural Area by month Water depth and its relationship with month*habitat Water temperature and its relation to month*habitat Vegetation height and its relation to month*habitat

9 3-15. Vegetation density and its relationship with month*habitat Captured prey sizes as a proportion of total identified captures

10 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 FORAGING QUALITY OF FLOODED AGRICULTURAL FIELDS WITHIN THE EVERGLADES AGRICULTURAL AREA FOR WADING BIRDS (CICONIIFORMES) Chair: Martin B. Main Co-Chair: Richard N. Raid Major: Wildlife Ecology and Conservation By Grant Cashion Sizemore December 2009 Agricultural wetlands have been shown to be an important foraging resource for wading birds around the world where natural wetlands have been lost or degraded. Cultivation in the Everglades Agricultural Area (EAA), located immediately north of the remaining Everglades wetlands, utilizes both flooded rice and fallow fields and may be an especially important resource due to the altered hydrology of the Everglades and associated effects on breeding wading birds. I hypothesized that flooded agricultural fields within the EAA provide a foraging resource for wading birds during the breeding and immediate post-breeding seasons. I also hypothesized that foraging within these fields would allow wading birds to meet their daily energetic requirements. I conducted surveys at a regional and local scale to quantify wading bird use of the EAA and flooded agricultural fields, measured habitat characteristics of flooded fields, conducted focal observations of Great Egret and Little Blue Heron foraging, and sampled prey densities. Wading bird abundance in the EAA was positively correlated to poor hydrological conditions in the Everglades by drought and the onset of the wet season in late May and June. Vegetation was a discouraging factor to wading bird foraging in rice fields, 10

11 and flooded fallow fields were the preferred habitat in June. No difference existed between specific foraging sites within a field and general field characteristics. Prey densities increased each month, and June rice fields averaged greater density of prey than flooded fallow fields. Although Little Blue Herons foraged sufficiently to meet their daily energy requirements, Great Egrets did not. Wading birds used the EAA as a foraging resource during the breeding and immediate post-breeding seasons, and the timing of use suggests that the EAA is an important transitional foraging resource at a time when Everglades foraging resources are unsuitable. However foraging in flooded agricultural fields did not meet the energy requirements of Great Egrets, suggesting that the region functions as a substandard foraging resource for this species during this time period. Further research is necessary to understand how the EAA might be important throughout the summer and whether potential pollutants may be inadvertently ingested by foraging in an agricultural landscape. 11

12 CHAPTER 1 INTRODUCTION Agricultural wetlands consist of agricultural fields or impoundments with standing water during all or a portion of the year. These man-made wetlands have the potential to play an important ecological role as habitat for waterbirds (e.g., wintering grounds, foraging habitat, nesting sites), especially in areas where natural wetlands have been lost or degraded (Fasola 1986; Elphick and Oring 1998; Czech and Parsons 2002). Among farmed agricultural wetlands, rice cultivation has been reported to be the most similar to natural wetlands (Martínez-Vilalta 1996). Rice fields require flooding and draining and may be considered temporary wetlands (Lawler 2001). As such, rice fields have the potential to function as surrogate habitat for wetland species (Fasola and Ruiz 1996), especially in areas where natural wetlands may be scarce or degraded. Because rice is the world s most widespread crop with 40% of the world s population depending on it as a primary food source (Forés and Comín 1992), rice has the potential to perform a valuable function in the conservation of wetland-dependent species over a broad geographical range. Flooded fallow fields may also serve as important resources for waterbirds. Flooded fields have been shown to attract waterbirds (Day and Colwell 1998; Elphick and Oring 1998), and certain species may actually prefer flooded fallow fields over rice fields (Maeda 2001). Fujioka et al. (2001) reported that species richness was greater among waterbirds in flooded fallow fields than in rice fields. Pearlstine et al. (2004) found that waterbirds in the Everglades Agricultural Area (EAA) utilized flooded fallow fields and that within the EAA the abundance and diversity of waterbirds in flooded fallow fields were second only to rice fields. Both rice fields and flooded fallow fields, 12

13 therefore, may have a role to play in providing wetland habitat and mitigating the effects of wetland loss or degradation. Wading birds (herons, egrets, storks, ibises, bitterns, and spoonbills) are one of the most visible wildlife components of wetlands (Hafner 1997) and are characterized by long legs, a carnivorous diet, and an association with water (Kushlan 1981a). Despite their charismatic appearance, ability to live near humans, and value as a source of income in the form of ecotourism (Kushlan 1997; Main and Vavrina 2001), many wading birds are of conservation concern. Contamination by pollutants, killings due to a conflict of interest with aquaculture, and the loss and alteration of habitat remain serious considerations for the conservation of wading birds (Kushlan 1997). Within the Greater Everglades Ecosystem, the loss and alteration of habitat is often cited as the primary cause for recent declines among wading birds (National Audubon Society 1992; Ogden 1994; Light and Dineen 1994; Kushlan 1997). These declines, estimated between the 1930s and 1990s, range from 70-93% (Robertson and Kushlan 1974; Ogden 1994; Crozier and Gawlik 2003). Around the world, wading birds have been reported to utilize agricultural wetlands to meet their resource needs (Main and Vavrina 2001; Czech and Parsons 2002). In the Mediterranean region, the distribution and size of heronries depends on the configuration of nearby rice fields (Fasola and Barbieri 1978; Fasola et al. 1996), and rice fields provide foraging habitat that supports between % of all herons during the breeding season (Fasola et al. 1996). Despite the reported use of rice fields by wading birds, rice field habitats may not be of the same quality as natural wetlands and may not be preferred habitat by wading birds (Fasola et al. 1996; Campos and 13

14 Lekuona 2001; Richardson and Taylor 2003). While many studies report that rice fields may not be equivalent to natural or seminatural wetlands as a foraging resource, for species richness, or for species abundance (Fasola et al. 1996; Kazantzidis and Goutner 1996; Elphick 2000; Ma et al. 2004), some studies suggest that rice fields are actually better foraging habitat (Hafner et al. 1986, 1987; Kazantzidis and Goutner 1996). The EAA may be an important resource for wading birds in South Florida because of its proximity to the Everglades, large size, and incorporation of both flooded rice and fallow fields. The role of the EAA as foraging habitat for wading birds in South Florida may be especially important during the breeding and immediate post-breeding seasons. Due to the high energy associated with feeding offspring, successful foraging in high-quality habitat is crucial. Starvation is one of the leading causes of death among wading bird nestlings (Owen 1960; Jenni 1969; Moser 1986; Frederick and Spalding 1994), and Spalding and Forrester (1991) found emaciation to be the cause of 54% of nestling mortality in nesting ardeids (herons, egrets, and bitterns). Although prey abundance is important, prey availability is the major factor influencing foraging success. Disruptions in the normal, annual drydown in the Everglades have the potential to hinder foraging success by altering the rate of drying and influencing prey availability. These disruptions can be natural (e.g., heavy rainfall or drought) or anthropogenic (e.g., diverting water to cities). Drying is an important phenomenon in South Florida s wetlands because it limits the number of deep water pools and concentrates prey resources beneficial to foraging wading birds (Kushlan 1976). 14

15 Altered hydrological regimes in South Florida affect reproduction in wading birds. Disruptions of the drydown in the Everglades can affect nesting success because of the importance of prey availability on successful breeding. The initiation of nesting among wading birds is triggered by the drydown phase of the hydrological cycle (Kahl 1964; Kushlan et al. 1975). In response to altered hydrological patterns, wading birds have shifted the initiation of nesting to later in the season, which has caused nesting failure as a result of not being able to complete nesting before the start of the wet season (Kushlan 1986; Custer et al. 1996). Rising water levels in the wet season limit foraging resources. Wood Storks (Mycteria americana) have also begun to shift their winter foraging grounds to water conservation areas (WCAs), north of Everglades National Park, where water conditions are more stable (Bancroft et al. 1992). The EAA has a different hydrological regime than natural wetlands. The presence of canals, irrigation ditches, and flooded agricultural fields means that water is present in the region even in the dry season. Agricultural fields are flooded during spring, toward the end of the dry season. These fields, in particular, may serve as a valuable foraging refuge for wading birds during times when foraging resources are limited elsewhere in South Florida. Although relationships between wading birds and agricultural wetlands have been studied around the world, there exists little information on how wading birds use the extensive agricultural area bordering the Everglades known as the EAA. This region occurs immediately north of the remaining Everglades wetlands and includes both rice and flooded fallow fields as annual components of the agricultural landscape (Izuno and Bottcher 1994). In an effort to understand how wading birds use the 15

16 agricultural wetlands of the EAA and evaluate their importance as a foraging resource during the breeding and immediate post-breeding seasons, I studied wading bird usage patterns and the foraging behavior of select wading birds within flooded rice and fallow fields. The objectives of this study were to provide information on the use of flooded agricultural fields within the EAA as wading bird foraging habitat at a regional and local scale, the quality of foraging habitat in relation to daily caloric needs, and to advance our understanding of wading bird ecology and the interactions between human-modified landscapes and wildlife in South Florida. I hypothesized that flooded agricultural fields within the EAA provide a foraging resource for wading birds during the breeding and immediate post-breeding season. I also hypothesized that this foraging habitat provides wading birds with sufficient prey to support daily caloric needs. I predict that the presence of wading birds will be influenced by water availability in the agricultural fields within the EAA and hydrological changes within the Everglades. To test this prediction, I surveyed wading birds at a regional and local scale and measured foraging success of two representative species within flooded rice and fallow fields. I also predict that foraging within the EAA will meet daily caloric needs. To test this prediction, I obtained measurements of foraging capture rates and calculated calories acquired for each capture to identify the total number of hours of foraging necessary to meet daily energy requirements. 16

17 CHAPTER 2 METHODS Study Area My study area was the entirety of the 283,000-ha region of southern Florida known as the EAA (Izuno and Bottcher 1994). The EAA exists in a subtropical climate with an average annual precipitation of 133 cm (Abtew and Khanal 1994). The EAA was at one time part of the expanse of wetlands encompassed by the Everglades (Small 1929) but has been drained and now consists of a mosaic of agricultural fields, canals, irrigation ditches, and impoundments (Table 2-1). The EAA extends, roughly, from the southern edge of Lake Okeechobee in the north to the Stormwater Treatment Areas (STAs) and WCAs 1, 2, and 3 along the east and south (Figure 2-1). The region is characterized by highly organic muck soils (Snyder 1994). Muck soils are highly productive and agriculture thrives in the EAA. The predominant crop is sugarcane, which constitutes 76% of all farming in the region (Izuno and Bottcher 1994; Snyder and Davidson 1994). Rice is an important crop in the EAA as well. Approximately 9,600 ha of rice are grown as a rotational crop for sugarcane annually, comprising about 5% of the cultivated landscape (Jones et al. 1994; Izuno and Bottcher 1994). Flooded fallow fields, which are used to cover vegetable fields following harvest, were calculated to cover 2,016 ha (1% of the cultivated landscape) during the time of my study. Table 2-1. Percent coverage of the Everglades Agricultural Area by various land uses. Habitat Type Total Area (ha) % Land Use of EAA Field 161,500 57% Ditch 6,000 2% Canal 6,500 2% Other 109,000 39% 17

18 Figure 2-1. Map of South Florida. The Everglades Agricultural Area (EAA), located north of the Water Conservation Areas (WCAs), is solid gray and outlined in black. East-west lines represent aerial transects flown in 2008 and The hydrological regime within the EAA is different from the rest of South Florida. Fields within the EAA begin to become flooded in the spring (Table 2-2). Rice fields are flooded as early as mid-march but typically not until April. Flooded fallow fields become flooded after the final vegetable harvest, which may be as early as late May, but are not flooded on a large scale until June. Precipitation can influence the timing of flooding. Higher amounts of precipitation may lead to earlier flooding, and less precipitation may lead to later flooding (personal communication, M. Ulloa, Florida Crystals). Table 2-2. Flooding timeline for agricultural fields within the Everglades Agricultural Area during January-June. January February March April May June None None Flooding Flooding Flooded Flooded Rice begins (late) Flooded Fallow None None None None Small scale flooding begins (late) Flooded 18

19 Wading Bird Surveys I conducted aerial surveys throughout the wading bird breeding season (March- June, 2008 and January- June, 2009) to estimate wading bird use of the EAA (i.e., abundance and density) at a regional scale. Aerial surveys were flown once a month using a fixed-wing aircraft (Cessna 172) at an altitude of m between 0800 and 1200 h. Transects were spaced 4 km apart to maximize overall coverage and minimize the risk of counting any individual more than once. Each survey included 7 transects and covered approximately 895 km 2 of observed area, resulting in 32% coverage of the entire area of the EAA (Figure 2-1). Two observers identified and recorded the habitat (except March 2008) and activity of all wading birds within 1.5 km of either side of the aircraft. I selected the flight area for known agricultural wetlands. These surveys allowed for a large expanse of ground to be covered relatively quickly but limited the accuracy with which certain species could be seen. Species with white or pink plumage (e.g., Great Egret [Ardea alba], Roseate Spoonbill [Platalea ajaja]) were particularly visible from the air, but dark-plumaged species (e.g., Little Blue Heron [Egretta caerulea]) were less visible. I also conducted ground surveys to measure species richness, community composition, and density of wading birds. I visited flooded rice fields and flooded fallow fields in the EAA at least five days each week between April-June in 2008 and Fields were selected opportunistically based on the presence of wading birds, which reflected habitat use. Each morning between 0700 h and 1200 h, I visited a single flooded rice or fallow field (approximately 15 ha) where wading birds were present. I visually surveyed and identified to species all wading birds in half a field (8 ha) once every hour. Unlike aerial surveys, these ground counts were well suited for observing 19

20 and quantifying less conspicuous species. Although most birds at a field were observed with reliability, visibility decreased with increasing vegetation height which obscured smaller species (e.g., Green Heron [Butorides virescens], Least Bittern [Ixobrychus exillis]). Despite reduced visibility, I am confident that nearly all wading birds were observed on the basis of the duration of observation of a field. For each survey, I recorded the number of each species present, date, location, habitat (flooded rice field or flooded fallow field), and time. Observations did not occur during periods of heavy rainfall. Field observations were always conducted between the same hours. In this way, I was able to control for biases associated with time of day. I chose to start with early morning observations because birds tend to forage most actively at sunrise and sunset (Kushlan 1978). Consistency in time of observation was also important because the abundance of most aquatic animals varies during the day (Kersten et al. 1991) and could, therefore, affect the foraging behavior of wading birds. Sampling in the early morning was likely to best represent ideal foraging conditions since fish may have to come to the surface at this time due to low dissolved oxygen (Kushlan 1974; Kersten et al. 1991). Foraging Observations I conducted focal observations for Great Egrets and Little Blue Herons as two representative wading bird species. These two species were chosen because they are highly visible within the EAA (Townsend et al. 2006) and are representative of different size classes of wading birds. The Little Blue Heron represents medium-sized wading birds while the Great Egret represents larger species (Kushlan 1978). Despite their size 20

21 differences, both species forage in similar habitats and, unlike tactile foragers (e.g., Wood Storks, White Ibises [Eudocimus albus]), use a visual feeding strategy. Foraging behavior was recorded using focal sampling (Altmann 1974). Each focal sample lasted 10 minutes. Focal samples ended if 1) the 10-min period expired, 2) the bird flew away, 3) the bird was lost from sight, or 4) the bird ceased foraging. I randomly assigned the first species for observation for the first focal sampling period (between arrival and the first hour s survey) and then alternated between species, observing new individuals in each focal sample. Focal individuals were chosen on the basis of active foraging. The first adult individual of the selected species seen making a strike was observed. By attempting to only select Great Egrets with breeding plumes and Little Blue Herons in adult (blue) plumage, I minimized the likelihood of observing juvenile foraging during focal observations. Selecting only adults was important because studies have shown that juvenile wading birds have lower foraging success per unit time than adults (Recher and Recher 1969; Quinney and Smith 1980; Bildstein 1984). During each focal sample, I recorded the number of feeding attempts (i.e., strikes) and the number of successes. A feeding attempt was deemed successful if 1) a prey item was seen captured and swallowed, 2) movements of the gular region consistent with swallowing occurred, or 3) the individual showed evidence of prey transport by head throwing (Gans 1961; Elphick 2000). For successful feeding attempts, I attempted to identify the prey and note the size using a comparison with bill length (Recher and Recher 1969; Cramp and Simmons 1977; Quinney and Smith 1980; Bayer 1985; Campos and Lekuona 2001). I divided the wading bird bill size into fourths 21

22 and estimated prey size as fractions of this denomination. I attempted to observe four individuals (2 of each species) foraging in four separate focal samples per hour. In observing foraging behavior and surveying wading birds, I used 10x50 binoculars and a spotting scope, and all observations were made from a vehicle in order to minimize disturbance. I observed foraging individuals between m from the vehicle. Habitat Characteristics I measured water depth, water temperature, vegetation height, and vegetation density to obtain flooded field characteristics in fields where wading bird ground surveys were conducted. I recorded measurements to describe fields in general and at specific foraging locations within a field. To measure water depth and water temperature, I established a 50 m transect with a measuring tape perpendicular to the edge of the flooded field. I measured water depth to the nearest 0.5 cm using a standard meter stick and water temperature to the nearest 0.1 C. I collected water depth and temperature measurements at 10 m intervals. Upon reaching 50 m from the edge of the field, I turned 90 o and sampled an additional 50 m, obtaining a total of 11 samples of water depth and temperature per plot. I measured vegetation height using a modified technique from Robel et al. (1970). Traveling the same transects as were used for water depth and water temperature, I measured vegetation height every 25 m. At each point, I sank a pole (marked every 10 cm) into the earth and stood back a distance of 2 m. From a height of 1 m, I recorded the lowest visible decimeter marking on the pole. This technique was used to provide a representative measure of the obstructive vegetation from a foraging bird s perspective. I measured vegetation density at the same points as vegetation height by estimating the proportion of vegetative cover within 9 square, grid cells (33 cm 2 ) in a 1-m 2 floating grid (Surdick 1998). 22

23 I also measured environmental variables in relation to specific foraging sites within flooded fields. At the end of all focal sampling and surveys in the morning, I identified a foraging area which was either being used or had been used at some point throughout the morning. I used random sampling to collect measurements within a 30- m radius of the identified foraging area. I collected a single measurement for vegetation density, vegetation height, water temperature, and water depth at three locations within the specified radius. Habitat characteristics were measured at wading bird foraging sites to provide a more precise account of specific foraging site characteristics and for comparison to characteristics measured for fields in general. Prey Sampling Prey were sampled in both flooded rice fields and flooded fallow fields each month (May-June in 2008, April-June in 2009). I estimated prey density using a 1-m 2 Kushlan throw trap (Kushlan 1981b) to quantify fish, amphibians, and crayfish at multiple locations within randomly selected fields. These taxa have been found to make up a large portion of wading bird diets (Baynard 1912; Jenni 1969; Hoffman 1978; Telfair 1981; Kushlan 1986; Frederick and Collopy 1988; Rifé 1997; Campos and Lekuona 2001). I also included insects in 2009 based on field observations from the previous year. By dividing the number of individuals captured by the area of the throw trap, I obtained a relative prey density for the field. Prey sampling occurred once a month for both field types. For each sampling event, I threw a trap 15 times. Samples were taken more than 5 m apart and could thus be considered as independent (Heyer et al. 1994). Although I did not determine the level of detectability, this method has been shown to be useful for studies of fish in shallow marshes (Kushlan 1981b). Each captured individual was weighed to the nearest 1.0 g using a Pesola spring scale and 23

24 identified. Only fish were identified to species. Body length (total length) was measured to the nearest 1.0 mm. Data Analyses To determine the use of the EAA regionally by wading birds, I analyzed the effects of variables on abundance. I used a chi-square and Fisher s exact test for independence (PROC FREQ, SAS Institute Inc. 2008) to identify the effects of year, month, the interaction of year and month, and species on abundance. To analyze the effect of habitat on abundance, I used a generalized linear model (PROC GLIMMIX, SAS Institute Inc. 2008) and negative binomial distribution. Precipitation data for the Everglades was collected from representative locations through the South Florida Water Management database, DBHYDRO ( accessed October 2009). To determine the effects of precipitation on abundance, I used general linear models (PROC GLM, SAS Institute Inc. 2008), which include ANOVA and linear regression in its analyses. Water depth data was collected from the Everglades Depth Estimation Network ( accessed November 2009). I analyzed ground survey data for effects on abundance and species richness using a general linear model with a square root transformation. I determined the effect of month on overall wading bird abundance using a chi-square test for equal proportions. Because flooded fallow fields were only available in June, I excluded this habitat from analyses of the effect of month. In order to make any comparisons between flooded rice and fallow fields, a combined variable was created to reflect month and habitat (month*habitat). Therefore, a rice field in April was represented as April 24

25 Rice. I used a general linear model to determine the effect of month, month*habitat, and environmental variables on the abundance and species richness of wading birds. I calculated species richness and used a Shannon diversity index to evaluate wading bird species diversity for data collected from ground surveys. To determine whether there was a difference in habitat use by species, I used a chi-square test. I evaluated foraging success of focal species between flooded rice and fallow fields using general linear models and incorporating a square root transformation. I analyzed captures/strike and captures/min for the effect of species. Within species, I used a general linear model to determine the effects of year, month, month*habitat, and environmental variables. I measured habitat characteristics representative of the whole field and linked them with all focal observations of foraging wading birds during that day for that field. Additionally, I analyzed environmental variables by month and month*habitat to identify potential differences. I used Student s t-test to evaluate the differences between habitat characteristics at specific foraging sites within a flooded agricultural field and the mean values for these measurements for the whole field. I analyzed the effect of year, month, and month*habitat. I evaluated the effects of species, year, month, and month*habitat using a generalized linear model and negative binomial distribution on the number of different types of prey and prey sizes captured by wading birds. I calculated the amount of foraging time required to obtain daily energy needs for Great Egrets and Little Blue Herons based on the proportion of prey types consumed by foraging individuals. I used weighted means of median captured prey sizes from foraging observations and length-mass equations from Kushlan et al. (1986) to 25

26 determine the masses of prey items in flooded agricultural fields because all captured prey from prey sampling were less than 1.0 g. Using caloric conversions for fish (Cummins and Wuycheck 1971), crayfish (Pope et al. 2001), amphibians (Evenson and Kruse 1982), and insects (Cummins and Wuycheck 1971), I calculated the amount of energy obtained for a capture of each prey type and the amount of energy that is usable (80%; Kahl 1964; Kushlan 1977). I multiplied captures/min for each wading bird species in both years by the proportional caloric values obtained based on prey consumption to estimate overall intake of calories/min foraging. I then compared these values to the non-breeding daily energetic requirements for both Great Egrets and Little Blue Herons provided by Frederick and Powell (1994). From these comparisons, I calculated the amount of foraging time necessary for both species in both years to meet daily energy requirements. 26

27 CHAPTER 3 RESULTS Wading Bird Surveys: Regional Regional wading bird use of the EAA in 2008 and 2009 differed both in overall pattern and monthly totals (Table 3-1, Fig. 3-1), and these differences corresponded to density (Table 3-2, Fig. 3-2). Two peaks in abundance were observed during the 2008 season (March and June), whereas a single peak occurred in June during Total wading bird abundance was influenced by month (χ 2 = 2473, df = 3, p<0.01) and year*month (χ 2 = 1110, df = 3, p<0.01) but not year. A total of 60.7% of all wading bird observations occurred in the month of June. Table 3-1. Abundance of wading birds observed in aerial surveys (2008, 2009). January February March April May June 2008 NA NA Wading bird abundance within the EAA was not equally distributed across species (χ 2 = 2807, df = 9, p<0.01). Great Egrets (61.1%) and White Ibises (18.3%) were the most commonly identified species during the aerial surveys. However, aerial surveys had a strong bias towards light-colored species, and I was unable to identify a large portion of birds to species. During 2008 and 2009, 82% and 70% of wading birds observed during aerial surveys were not identified to species, respectively. Habitat type (field, ditch, or canal) significantly affected wading bird abundance (F = 31.88, df = 2, 24, p<0.01). Wading birds were found standing in or flying over fields more often (67%) than either ditches (32%) or canals (1%). During 2008, usage of both 27

28 Figure 3-1. Total wading bird abundance for aerial surveys during the 2008 and 2009 breeding seasons. Table 3-2. Density of wading birds (individuals/ha) observed in the Everglades Agricultural Area during the 2008 and 2009 breeding seasons. January February March April May June 2008 NA NA Figure 3-2. Density of wading birds in the Everglades Agricultural Area during the spring breeding seasons of 2008 and

29 fields and ditches was similar in April, but use of ditches declined and use of fields increased in May and June (Fig. 3-3). During 2009, wading bird abundance increased in fields beginning in April and continuing into May and June (Fig. 3-3). These results suggest that fields became more important in May-June of 2008 and in April-June of Although fields were used in greatest abundance, wading bird density was greatest in ditches. Incorporating the total coverage of each habitat type, wading bird density was highest in ditches in all months. Wading birds were observed in ditches in greater proportion than the habitat s availability, suggesting that ditches may be a preferred resource within the EAA for wading birds (Fig. 3-4). Figure 3-3. The number of wading birds observed in fields, ditches, and canals in aerial surveys within the Everglades Agricultural Area during the spring breeding seasons of 2008 and Note the break in time between 2008 and

30 Figure 3-4. Wading bird densities in fields, ditches, and canals as observed in aerial surveys within the Everglades Agricultural Area during the spring breeding seasons of 2008 and Densities are in number of birds per hectare. Hydrology in the Greater Everglades Ecosystem influenced wading bird abundance in the EAA. A significant relationship existed between rainfall in the Everglades and wading bird abundance in the EAA during 2009 but not 2008 (Fig. 3-5, 3-6). In 2009 precipitation within the Everglades from the previous month was strongly and positively correlated (R 2 =0.88) with wading bird abundance in the EAA (F = 28.42, df = 1, 4, p<0.01). Although previous months precipitation in 2008 was not significantly correlated to wading bird abundance, precipitation in the EAA from the previous month still explained 63% of the variation (R 2 =0.63) and mirrored wading bird abundance in overall pattern (Fig. 3-5). A lag was incorporated to provide time for a response between a rainfall event and wading bird abundance. When no lag was incorporated between precipitation and wading bird abundance, no effect was observed. 30

31 Figure 3-5. Wading birds observed in the Everglades Agricultural Area and previous months precipitation values for the Everglades, 2008 Figure 3-6. Wading birds observed in the Everglades Agricultural Area and previous months precipitation values for the Everglades, 2009 Two different patterns of water depth were observed in the Everglades in 2008 and 2009 based on over 2,000 stations located in Everglades National Park. In 2008, water depth was extremely low due to drought conditions, and there was a reversal in the drying pattern in April (Fig. 3-7). In 2009 there was a steady, continuous drydown of 31

32 water from January to May and then a spike in water depth in June during the onset of the wet season (Fig. 3-8). Figure 3-7. Mean Everglades water depth from March-June in comparison to wading bird abundance in the Everglades Agricultural Area in 2008 based on over 2,000 stations located in Everglades National Park. Figure 3-8. Mean Everglades water depth from January-June in comparison to wading bird abundance in the Everglades Agricultural Area in 2009 based on over 2,000 stations located in Everglades National Park. 32

33 Wading Bird Surveys: Local Scan sampling during spring breeding seasons of April-June 2008 and 2009 occurred over a total of 62 days and included 43 different fields. Scans in 2008 occurred in rice fields 23 times and in flooded fallow fields 7 times. Scans in 2009 occurred in rice 23 times and in flooded fallow fields 9 times. The smaller number of flooded fallow fields sampled in both years was because these fields were not flooded until June, whereas rice fields were flooded April-June. A total of 1,168 wading birds were counted in 2008, and 1,284 were counted in For both years, the number of wading birds counted per day of sampling increased each month (Fig. 3-9). I observed the greatest mean number of birds per day in June. Sampling in the month of June was restricted almost exclusively to flooded fallow fields, suggesting that flooded fallow fields were a more important resource for wading birds than rice fields in this month. Figure 3-9. The number of wading birds observed in scans of flooded agricultural fields during the spring breeding seasons in 2008 and Error represented is the standard deviation. Wading birds did not occur in flooded fallow fields during April or May because this habitat type did not exist (Figure 3-10). For this reason, I excluded flooded fallow 33

34 fields and analyzed only flooded rice fields to identify an effect of month. The mean number of wading birds per survey in flooded rice fields was different by month (F = 19.73, df = 2, 227, p<0.01), but the relationship was fairly weak (R 2 =0.15). The mean number of wading birds per day for both years was highest in May, followed by June, and April (Table 3-3). Table 3-4 shows the results of paired difference tests for the mean abundance of wading birds in flooded rice fields. Wading bird abundance in 2008 (4.19 ± 5.89) and 2009 (3.10 ± 4.99) were similar (F = 3.28, df = 1, p = 0.07). Figure A comparison of abundance by month for wading birds in flooded rice and fallow fields of the Everglades Agricultural Area. Table 3-3. A comparison of wading bird abundance per day in flooded agricultural fields within the Everglades Agricultural Area. Year Habitat Month N (fields) Abundance ( x ± SD) 2008 Rice April ± 2.08 May ± June ± Flooded Fallow June ± Rice April ± 4.42 May ± June ± 0.00 Flooded Fallow June ±

35 Table 3-4. Paired difference test (t-test) results for the mean abundance of wading birds observed in flooded rice fields. Values correspond to p-values. April May June April < May < June Incorporating flooded fallow fields, the mean density of wading birds observed per day was highest in June flooded fallow fields, followed by May rice fields, and June and April rice fields (Figure 3-10). Wading birds in flooded fallow fields tended to aggregate, thus explaining the large standard deviation observed in Figure Table 3-5 shows the results of paired difference tests for the mean abundance of wading birds by month*habitat. I incorporated the combined variable month*habitat because flooded fallow fields were only present in June. By combining the two variables, I was able to make comparisons between flooded rice and fallow fields. Month*habitat significantly affected the mean abundance of wading birds (F = 20.97, df = 3, 306, p<0.01). Table 3-5. Results of paired difference tests for the mean abundance of wading birds observed in ground surveys for flooded agricultural fields. Values given are significance levels (p-values). April Rice May Rice June Rice June Flooded Fallow April Rice May Rice June Rice June Flooded Fallow < <0.01 < < <0.01 <0.01 <0.01 <0.01 Certain habitat characteristics for flooded rice and fallow fields influenced wading bird abundance for different values of month*habitat (Table 3-6). Water depth significantly influenced abundance in April and May rice fields. Water temperature influenced abundance in May and June rice fields. Vegetation height and density 35

36 significantly influenced abundance in May rice fields and June flooded fallow fields. The influence of these habitat characteristics suggests that they were limiting factors for wading bird abundance. Table 3-6. Environmental variables and their effects on wading bird abundance for ground surveys during the spring of 2008 and Env. Variable Habitat Month N days F df p R 2 Water Depth (cm) Water Temp. ( C) Vegetation Height (cm) Vegetation Density (m 2 ) WB Abundance per site ( x ± SD) Rice April ± 1.61 May ± 6.12 June ± 5.67 Flooded June ± Fallow Rice April ± 1.61 May ± 6.12 June ± 5.67 Flooded June ± Fallow Rice April ± 1.61 May ± 6.12 June ± 5.67 Flooded June ± Fallow Rice April ± 1.61 May ± 6.12 June ± 5.67 Flooded June ± Fallow Env. Variable ( x ± SD) 6.56 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± , , , , , , , , , , , , , , , , < < < <

37 A total of 12 species of wading birds was observed over two years of sampling (Table 3-7). Wading bird species richness per survey in rice fields followed a similar pattern to abundance. Species richness was significantly related to month (F = 16.5, df = 2, p<0.01) and was highest in June (1.47 ± 0.13), followed closely by May (1.35 ± 1.05), and April (0.55 ± 0.72). Species richness was significantly different in April but not in May and June (Table 3-8). Water temperature in June (F = 5.23, df = 1,13, p=0.04; R 2 = 0.29), vegetation height in May (F = 16.90, df = 1, 148, p = 0.10; R 2 = 0.10), and vegetation density in May (F = 19.48, df = 1, 148, p<0.01; R 2 = 0.12) and June (F = 4.15, df=1, 13, p = 0.06; R 2 = -0.24) influenced richness. A Shannon diversity index revealed that species diversity in rice fields was highest in May (H=1.37), followed by June (H=1.14), and April (H=1.00). When flooded fallow fields were included, wading bird species richness per Table 3-7. Wading bird species and densities (individuals*ha -1 ) observed in flooded agricultural fields during April-June, Year Habitat Month Species Density 2008 Rice April Great Egret 0.17 Tricolored Heron 0.04 White Ibis 0.25 May Glossy Ibis 1.08 Great Blue Heron 0.03 Great Egret 1.44 Little Blue Heron 0.38 Snowy Egret 0.03 Tricolored Heron 0.01 White Ibis 0.03 Yellow-crowned Night-Heron <0.01 June Glossy Ibis 0.81 Great Blue Heron 0.19 Great Egret 0.81 Little Blue Heron

38 Table 3-7 Continued Year Habitat Month Species Density 2008 Flooded Fallow June Black-crowned Night-Heron 0.02 Glossy Ibis 0.07 Great Blue Heron 0.18 Great Egret 6.23 Green Heron 0.02 Little Blue Heron 0.18 Roseate Spoonbill 0.80 Snowy Egret 4.48 Tricolored Heron 0.07 White Ibis 0.13 Wood Stork Rice April Black-crowned Night-Heron 0.01 Great Egret 0.48 Little Blue Heron 0.09 Tricolored Heron 0.03 White Ibis 0.04 May Black-crowned Night-Heron 0.02 Glossy Ibis 0.02 Great Blue Heron 0.02 Great Egret 1.63 Little Blue Heron 0.24 Snowy Egret 0.83 Tricolored Heron 0.02 White Ibis 0.14 Yellow-crowned Night-Heron 0.01 June Glossy Ibis 0.38 Great Egret 2.38 White Ibis 0.25 Flooded Fallow June Black-crowned Night-Heron 0.01 Glossy Ibis 0.31 Great Blue Heron 0.11 Great Egret 7.35 Little Blue Heron 0.19 Roseate Spoonbill 2.00 Snowy Egret 1.96 Tricolored Heron 0.10 White Ibis 0.01 Wood Stork 0.85 survey was significantly affected by month*habitat (F = 14.94, df = 3, 306, p<0.01), although not all month*habitat combinations were significantly different from one 38

39 another (Table 3-9). Figure 3-11 illustrates how mean species richness was greatest in June flooded fallow fields (2.16 ± 1.94), followed by June rice fields (1.47 ± 1.30), May rice fields (1.35 ± 1.05), and April rice fields (0.55 ± 0.73). Vegetation height (F = 6.65, df= 1, 78, p=0.01; R 2 =0.08) and vegetation density (F = 6.47, df = 1, 78, p = 0.01; R 2 = 0.08) both significantly influenced species richness in flooded fallow fields. Wading bird diversity was highest in May rice fields (H=1.37), followed by June flooded fallow fields (H=1.31), June rice fields (H=1.14), and finally April rice fields (H=1.00). Table 3-8. P-values for paired difference tests comparing the mean wading bird species richness observed in flooded rice fields by month. April May June April < May < June Table 3-9. Results of paired difference tests comparing mean wading bird species richness observed in flooded agricultural fields by month*habitat. April Rice May Rice June Rice June Flooded Fallow April Rice < <0.01 May Rice < June Rice June Flooded Fallow < Foraging Success in Flooded Agricultural Fields I recorded a total of 256 focal samples over 260 days during the 2008 and 2009 seasons. Among all focal observations, 212 (82.8%) were Great Egrets and 44 (17.2%) were Little Blue Herons. While a significantly larger number of focal observations occurred in rice than in flooded fallow fields (χ 2 = 36.94, df = 1, p<0.01), an analysis of species by habitat indicated that the numbers of observed Great Egrets and Little Blue 39

40 Herons did not significantly differ from the expected value (χ 2 = 0.39, df = 1, p = 0.53), thus suggesting that habitat had little effect on which species was observed. Figure 3-11 Wading bird species richness in flooded agricultural fields within the Everglades Agricultural Area by month. I found a significant difference between the foraging success of Great Egrets and Little Blue Herons (Table 3-10). The mean capture rate per minute foraging was greater for Little Blue Herons than Great Egrets (F = 54.19, df = 1, 254, p<0.01). Similarly, the captures/strike of foraging attempts was greater for Little Blue Herons than Great Egrets (F = 5.89, df = 1, 231, p = 0.02). Although these statistics do not take into account the types of prey being captured or their caloric value, they do suggest that Little Blue Herons were more successful foragers than Great Egrets in terms of number of prey items taken in flooded rice and fallow fields. I analyzed rice fields alone to determine the effect of month on foraging success. Within rice fields, mean captures/min for Great Egrets differed significantly by month (F = 3.67, df = 2, 141, p = 0.03) and was highest in April (0.59 ± 0.67), followed by May (0.41 ± 0.37), and June (0.16 ± 0.21). Mean captures/strike was marginally different (F = 2.36, df =2, 131, p = 0.10). Mean captures/strike was highest in May (0.58 ± 0.33), 40