Final Report. W. David Shuford, Matthew E. Reiter, Khara M. Strum, Cory J. Gregory, Michelle M. Gilbert, and Catherine M. Hickey

Transcription

1 THE EFFECTS OF CROP TREATMENTS ON MIGRATING AND WINTERING WATERBIRDS AT STATEN ISLAND, Final Report W. David Shuford, Matthew E. Reiter, Khara M. Strum, Cory J. Gregory, Michelle M. Gilbert, and Catherine M. Hickey PRBO Conservation Science 3820 Cypress Dr. # 11 Petaluma, CA April 2013 Submitted to: The Nature Conservancy 190 Cohasset Road, Suite 177 Chico, CA 95926

2 EXECUTIVE SUMMARY Agricultural landscapes provide benefits for wildlife, particularly birds. There is little information, however, on the influence of specific post-harvest management practices on waterbird use in many crops. During fall and winter in and , we conducted a study at Staten Island, in the Sacramento San Joaquin River Delta of California s Central Valley, to assess the effects on waterbird use of 10 distinct treatments (crop and management combinations) of corn, potatoes, flood-irrigated pasture, and dry-farmed winter wheat. We surveyed waterbirds in each treatment and collected data on the characteristics of the fields (e.g., water depth and proportion flooded, moist, or stubble). In winter , we also conducted comprehensive (all-island) surveys of 11 species of large waterbirds to assess spatial and temporal patterns of distribution and abundance and proportional use of crop treatments for foraging and roosting/loafing. Use of crops and treatments varied considerably among waterbirds. Although most waterbirds were predictably common in flooded treatments, geese, cranes, and long-legged waders also were numerous in some dry ones. Regardless of whether corn was flooded or dry, overall the chop-and-roll treatment was more beneficial to waterbirds than the harvest-only practice. More specifically, in winter, shorebirds were most strongly associated with flooded chop-and-roll corn, wheat, and potatoes; highest densities were in flooded potato fields. In fall, shorebirds had significantly higher densities in both flooded potato and wheat fields than in irrigated pastures. In winter, the densities of Sandhill Cranes were higher in dry than in flooded treatments for all crops but in fall densities showed no significant differences in treatment comparisons. In winter, densities of long-legged waders varied little across the range of crop treatments; the only significant relationship was a higher density in flooded chop-and-roll than in dry harvest-only corn. In fall, wader density was significantly higher in irrigated pastures than in dry potato or wheat fields. In winter, dabbling and diving ducks both showed large differences in densities in comparisons of some flooded treatments. The only significant relationship, however, was a higher density of dabbling ducks in flooded potato fields than in flooded winter wheat. Waterbird species richness varied seasonally and by crop, and was further influenced by variation in water depth and amount of flooding. In fall, species richness was significantly higher in irrigated pasture than in flooded wheat. In winter, species richness was positively associated with flooded chop-and-roll corn but negatively associated with flooded potato fields. Further analyses suggested this difference reflects different degrees of variation in the amount of flooding and water depths within these two treatments and the low sample size of flooded potato fields. In winter, species richness was significantly higher in fields with higher seasonal variation in water depths and the amount of flooding. We identified several significant mechanisms driving waterbird distribution. Shorebirds, dabbling ducks, and diving ducks all were significantly associated with water depth, with densities generally increasing with water depth but then declining after a depth threshold was reached. Shorebirds were positively associated with the amount of moist soil, whereas dabbling

3 ducks and diving ducks were positively associated with the amount of flooded habitat. Dabbling ducks had a significant negative association with the amount of residual stubble. Sandhill Cranes, however, were negatively associated with the amount of flooded habitat and water depth. Spatial and temporal patterns of waterbird use and bird behavior varied among species groups and across crop treatments. Dark geese generally were foraging in higher proportions in dry than in flooded treatments, where roosting was more common. White geese were predominantly foraging in two treatments; those roosting were in flooded fields. Overall, geese and cranes increased in abundance in the southern portion of the island from early to late winter. Sandhill Crane abundance increased through the winter, whereas geese generally declined, mainly from a large decline in the number of Cackling Geese. The spatial distribution patterns of long-legged waders and the Tundra Swan remained relatively constant across the winter, though overall waders increased, while swans decreased, in abundance. Our study confirmed the value of most wildlife-friendly management treatments specific to fields (i.e., flooding and chopping and rolling) on Staten Island. We suggest that by maintaining a diversity of dry and flooded crop types and continuing the chop-and-roll practice for corn managers can achieve a high level of waterbird diversity. The optimal amount of specific crops and flooding treatments on the island, however, will require the development of clear conservation objectives for waterbirds. This first-ever rigorous test of the associations of waterbird groups with crop management practices in corn, winter wheat, and potatoes provides the basis to inform management decisions at Staten Island. Likewise, it may aid in the development of incentive-based programs to foster similar management on other properties. At a broader scale, these detailed data from Staten can help refine regional models that use birdhabitat relationships to estimate habitat objectives required to meet population goals, to gauge progress toward those goals, and to inform an adaptive management framework for major decisions on land use changes in the Delta. 2

4 BACKGROUND There are many benefits of agriculture for wildlife, particularly birds (Taft and Elphick 2007). Although varying among crop types and management practices, crops may provide birds with foraging and loafing habitat in winter or migration, and breeding and foraging habitat in the nesting season (Sterling 2011). Because agricultural production is the dominant land use in some regions of North America, favorable agricultural crops have, to varying degrees, offset the loss of historic native habitats to some wetland-dependent birds. Yet the benefits to wildlife have infrequently been quantified. The agriculturally dominated Central Valley of California is a very important region for wintering and migratory waterfowl (Heitmeyer et al. 1989, Fleskes et al. 2005), shorebirds (Shuford et al. 1998), and other waterbirds (Shuford in press). Within the Central Valley, the Sacramento San Joaquin River Delta (hereafter Delta) is particularly important to wintering Sandhill Cranes, including the state threatened Greater Sandhill Crane (Grus canadensis tabida) (Pogson and Lindstedt 1991, Ivey and Herziger 2003). Although rice and corn currently are the most important crops to waterfowl and other waterbirds in the Central Valley, corn is by far the most important crop to these bird groups in the Delta (CVJV 2006). In California, winter wheat (Triticum spp.) and corn (Zea mays) are ranked third (~754, 000 acres) and fifth (~546,000 acres) in total crop acreage planted statewide (NASS 2007). Although the benefits of wheat and corn for selected avian species during the winter and migration period have been documented (see Taft and Elphick 2007), there is little information on the influence of post-harvest management practices (e.g., harvest only, chopping and rolling, flooding) on waterbird use in these crops. In the Central Valley, there are an estimated 30,000 acres of flooded and 70,000 acres of dry corn fields available to birds (CVJV 2006, p. 46). The Delta holds about 60% of the corn acreage available to waterfowl in the Central Valley and 100% of the valley s winter-flooded corn. The Central Valley Joint Venture also considers winter-flooded corn in the Delta to be comparable to flooded rice as an invertebrate food resource for shorebirds (CVJV 2006, p. 242). The validity of this assumption, however, is not known. The availability of flooded winter wheat and use of winter wheat fields in the Central Valley by waterbirds is also largely unknown. In 2010, we initiated a two-year study at Staten Island a large farm in the heart of the Delta to evaluate how migrating and wintering waterbirds currently use its agricultural landscape. Due to its large size and managers receptivity to changing crop management practices to benefit wildlife, Staten Island is well-suited for assessing the interface of wildlife management and agriculture practices. The island is owned by The Nature Conservancy and operated by Conservation Farms and Ranches, Inc. with a dual mandate of economic viability and wildlife conservation. Here we report how the composition, abundance, and distribution of waterbirds at Staten Island were affected by post-harvest management practices in corn, winter wheat, and potatoes and by flood-irrigated pasture and dry-farmed winter wheat while growing. We also document the relationships between habitat metrics and the observed associations of 3

5 waterbirds with the selected crops and crop treatments. Finally, we discuss how our results might improve management for waterbirds on Staten Island, provide the basis for extending this management evaluation to other farming operations in the Delta and the entire Central Valley, and bolster the scientific underpinnings for broad-scale conservation planning in agriculturally dominated regions (e.g., CVJV 2006). STUDY AREA Staten Island is a 9,200-acre farm located in the central part of the Delta, one of the major subdivisions of the larger Central Valley. Historically the Delta was a maze of sloughs and swampy islands, but an extensive levee network now protects multiple large islands or tracts from floods and tidal surges. Because of aeration of the Delta s peat soils, most of the islands, including Staten, are below sea level. Major environmental concerns in the Delta include potential inundation of islands from catastrophic breaks in levees from earthquakes, rising sea levels, further island subsidence, conversion of wildlife-friendly crops to vineyards and orchards, and increasing salinity and its impacts on threatened and endangered fish. Some of these factors threaten the state s water supply and the viability of agriculture in the Delta and other parts of the Central Valley. The over 8,000 acres in agriculture on Staten Island are dominated by corn (76% 81% of total acreage) followed by winter wheat (5% 12% of total acreage), irrigated pasture, and sometimes other crops (e.g., potatoes) (Figs. 1 and 2; Golet 2011). In the season winter wheat was grown, but in this was replaced by triticale, a cross between wheat (Triticum) and rye (Secale). For convenience, we refer to these two crops throughout as winter wheat, as the timing and methods of cultivation and of post-harvest management are the same. Currently, wildlife-friendly practices are in place to benefit avian and terrestrial species, particularly waterfowl and cranes (Ivey et al. 2003). Conservation tillage light disking that leaves some crop residue on the surface is employed during corn harvest, and various management practices for this and other crops are thought to benefit wildlife, by increasing food availability, and agriculture, by hastening the decomposition of corn residue, controlling weeds, reducing soil erosion and compaction, and lowering fuel and labor costs (Golet 2011). METHODS CROP TREATMENTS Based on life history knowledge of various waterbirds, we expected that many species would preferentially use flooded fields and fewer would use dry fields. Likewise, we judged that the amount and height of standing crop stubble and water depth in flooded fields were factors likely to influence species use of particular crops or treatments. To test these hypotheses, we evaluated bird use of four crops (corn, winter wheat, potatoes, irrigated pasture) and a variety of management actions (Table 1, Figs. 1 and 2), most of which are employed annually. Of the 10 treatments defined by crop and management combinations, 5 were employed in fall (August- October) and all 10 during the winter (November-February). Within crops and treatments, 4

6 there was substantial variation within and across fields in the amount of flooding, moist soil, residual stubble, and water depth (Table 1, Fig. 3). There is considerable variation both within and among crops in the timing and duration of events in the growth and post-harvest phases. After corn is harvested in September and October, in recent years the primary practice at Staten Island is to chop and roll the remaining corn stubble, which reduces weed growth and retains soil moisture. To accommodate our study, the farm manager left some fields as is after harvest so we could compare bird use between chop-and-roll and harvest-only corn fields. About 25% of the corn is flooded from October or November into February, the remainder is left dry except for surface moisture and scattered puddles from winter rains. Winter wheat is typically harvested in July, and some fields are flooded post-harvest to provide habitat for early arriving cranes, waterfowl, and migrating shorebirds. The period of post-harvest flooding of winter wheat varied among years from about 1-4 months. Winter wheat is seeded in the early winter with a 20% increase above the normal seeding rate to account for loss to foraging cranes and other waterbirds. Also, some species, particularly geese, forage extensively on the blades of growing wheat. Moisture for growth of wheat comes from seasonal rains and occasional flood irrigation during extended dry periods. After harvest in 2011, some potato fields were tilled finely and then flooded from October thru early February. Pastures are flood-irrigated intermittently from spring through fall, when cattle are present, but thereafter moisture is supplied by winter rains; standing water is very limited in pastures at any season. SAMPLING SURVEYS We selected a random sample of fields using Generalized Random Tessellation Stratified (GRTS) sampling methodology, which enabled the selection of a set of spatially balanced random locations with respect to crop type and treatment (Stevens and Olsen 2003). We conducted surveys from pre-determined points at the edge of each selected field. Because distance can bias counts of birds, we restricted counts to within a 200-m arc from the survey point, which was typically truncated on the sides by the lateral boundaries of the field. From 11 August 2010 to 8 March 2011 and from 1 September 2011 to 28 February 2012, we surveyed waterbirds at Staten Island on 30 occasions across our 10 crop treatments (Table 1). Observers conducted 11 surveys (6 in 2010, 5 in 2011) of 18 survey areas during the fall (August-October) and 19 surveys (10 in , 9 in ) of up to 68 surveys areas during the winter (November-February). We surveyed birds in each sampling area approximately once every 14 days, and we varied the order in which we surveyed areas to avoid bias in counts from the effects of time of day. We attempted to start surveys in early morning, but sometimes delayed the initiation of surveys in winter until after fog burned off. We did not survey when winds exceeded 20 mph or during steady rain. Observers used binoculars and spotting scopes to scan all survey areas from the survey point for at least two minutes; there was no maximum time limit for completing a count. We did not count birds that were flying over the field during the survey (except foraging raptors). 5

7 Our surveys focused on waterbirds (waterfowl, shorebirds, herons and egrets, ibis, coots, cranes, grebes, gulls) and raptors. We counted all individuals of waterbird and raptor species within the defined survey area and, with a few exceptions, we identified all birds to species. We also recorded the presence, but not the abundance, of upland birds (pheasants and, mainly, passerine birds). Under survey conditions, it was not possible to adequately distinguish between Longbilled (Limnodromus scolopaceus) and Short-billed (L. griseus) dowitchers, so we pooled all observed as dowitcher spp. These likely represent mainly Long-billed Dowitchers, as that is the only species known to occur in the region in winter and is by far the most dominate in migration. Likewise, in the few cases in which we were unable to distinguish between Western Sandpipers (Calidris mauri), Least Sandpipers (C. minutilla), or Dunlin (C. alpina) we pooled such data as Western/Least sandpipers or Western/Least/Dunlin. We recorded water depth and other characteristics of the field condition at the time of the survey. At each sampling location (except in irrigated pasture), we placed two survey stakes marked with multiple, 2-inch bands of colored paint. One stake was about 25 m out at the far edge of the headlands a strip at the field end tilled during harvest to allow tractors to easily turn around and the other 200 m out at the far boundary of the survey area. On each survey, we recorded water depth at the two stakes and we also estimated the proportion of the survey area that was flooded, moist, or dry and the proportion that consisted of residual stubble or green vegetative growth. Because the corn fields that were flooded were invariably long and narrow, sloped along the long axis, and accessible only at their short ends, our surveys were from the deep and shallow (sometimes dry) ends of fields. COMPREHENSIVE SURVEYS In winter , we supplemented our sampling surveys with bi-weekly surveys of all individuals of selected waterbird species (see below) in all fields across the entire island to better understand their temporal patterns of distribution and abundance. We conducted a total of eight comprehensive (all-island) surveys at Staten from 17 November 2011 to 22 February Two to three people counted simultaneously, each starting in different areas of the island and continuing until we collectively had counted all fields. We attempted to complete surveys in the morning, but sometimes surveys extended into early afternoon, particularly when early morning fog delayed our start time. On these all-island surveys, we recorded the number of birds by field, the proportions of each species that were foraging versus roosting/loafing (to assess variation in behavior by crop or treatment), and the percent of the field that was flooded. We already had data on the crop type and treatment in each field. Because most of the fields on Staten are very long and accessible only at their short ends, we judged that in counting over the entire field we could obtain accurate counts of only the large waterbird species. Hence the primary species counted on the all-island surveys were the Sandhill Crane and various species of geese, though we also counted smaller numbers of pelicans, herons, egrets, and swans. We sometimes were unable to distinguish between some species of geese when distances were great, birds were roosting tightly together with heads 6

8 tucked in areas of stubble, or when mobile large flocks constrained available count time. In such cases, we pooled counts of the Cackling Goose (Branta hutchinsii) and Greater White-fronted Goose (Anser albifrons) as unidentified dark geese and the Snow Goose (Chen caerulescens) and Ross s Goose (Chen rossii) as unidentified white geese. The dark geese might also have included small numbers of the Canada Goose (Branta canadensis), although we infrequently recorded this species on the island. ANALYSIS Abundance and Richness We modeled the effect of each crop treatment on species abundance and richness of waterbirds pooled by guild (long-legged waders [herons/egrets/ibis], cranes, shorebirds, dabbling ducks, diving ducks [diving ducks/grebes], and geese), survey location, and visit within each year. Pooling the data facilitated model convergence given the large number of observations with zero birds counted. Despite efforts to reduce zero-data observations, we still required a zero-inflated, over-dispersed Poisson (ZIP) distribution to model the effect of each of 10 treatments for some guilds (Kery 2010). We compared the fit of both ZIP models and over-dispersed Poisson models without the zero-inflation parameter using the Deviance Information Criterion (DIC). Whichever model had a lower DIC, indicating it was a relatively better fit to the data, was used for inference. We fit all models using the R2WinBUGS package in R (Sturtz et al. 2005), and Markov Chain Monte Carlo (MCMC) simulations were completed in WinBUGS software (Spiegelhalter et al. 2003). We used non-informative priors for all model parameters, and for each model run we used three Markov chains of five million iterations. We discarded the first four million iterations of each chain and sampled every 100 th value of the remaining one million estimates within each chain to reduce autocorrelation in the parameter estimates (Kery 2010). We used the combined 30,000 samples to make inference. To ensure that models had reached convergence, we waited until the r-hat statistic for each parameter was less than 1.5 before using the sampled parameter estimates (Gelman and Hill 2007). We used the fitted model to estimate the mean bird density (birds per ha) for treatments in each iteration of the MCMC model-fitting process and calculated the 95% credible intervals of the mean using the percentile method (Spiegelhalter et al. 2003). We also conducted simultaneous pairwise comparisons of the estimated densities between each treatment during each iteration of the MCMC simulation. We considered differences in densities of different treatments within avian groups to be significantly different if the 95% credible interval of the difference did not overlap zero. We included the natural logarithm of the area (ha) surveyed as an offset term in the models since the area surveyed at each sampling location was not equal (Kery 2010). We calculated the area (ha) surveyed from each sample point using ArcMap Version ( ESRI Inc.). We modeled the effect of crop-treatment combinations by season and defined the fall as August through October and winter as November through February. For shorebirds, waders, cranes, and geese, we evaluated all combinations of 10 crop treatments for the winter and all combinations of 5 crop treatments for the fall. Because we recorded no dabbling or diving 7

9 ducks in dry fields, we compared crop-treatment combinations for these guilds only if they were intentionally flooded. For geese, we were unable to achieve model convergence due to severe overdispersion in the data caused by the clustering behavior of these birds. Consequently, we applied a non-parametric bootstrap procedure (Manly 2007) to estimate the mean and 95% confidence intervals for each crop and treatment combination. For these comparisons, we considered the treatments significantly different when the 95% confidence interval of their estimated means did not overlap. We also evaluated four variables in analysis that we believed were the driving mechanisms influencing variation in waterbird use by crop treatment: water depth (cm) and the percent cover of the survey area that was flooded, moist soil, and stubble. We fit two models to winter data for each guild, except for geese due to problems with model specification and convergence. We also modeled species richness as a function of the variance in the same four variables just mentioned across the season. We hypothesized that species richness would be related to variation in the type of habitat rather than the amount. Due to low sample sizes, we did not evaluate mechanism models for fall data. We did not include covariates that had Spearman rank correlation coefficients of greater than 0.60 in the same model. We considered model covariates to be significant if their 95% credible interval did not overlap zero. Comprehensive Surveys We summarized the total birds on the island by species or species group for each of the eight survey events. We summarized bird behavior by comparing the percent of foraging birds to roosting/loafing birds by crop-treatment. Lastly, we mapped the distribution of birds by field and month to visualize spatio-temporal variation in bird distribution on Staten Island. For the behavior analysis and spatial-temporal mapping, we pooled bird observations into the following groups: (1) dark geese, (2) white geese, (3) long-legged waders, and (4) Sandhill Cranes, as described above. We used ArcMap v.9.3 ( ESRI Inc.) for spatial analyses and mapping. RESULTS SAMPLING SURVEYS Relative Abundance and Species Richness On sampling surveys in and , we counted 70,822 individuals of 57 species of waterbirds and raptors representing 8 avian groups; we also recorded the occurrence of 22 species of upland birds (Table 2). Totals for six species Greater Whitefronted Goose, Cackling Goose, Northern Shoveler, Canvasback, Sandhill Crane, and American Coot exceeded 5000 individuals. Totals for an additional seven species Tundra Swan, Northern Pintail, Ruddy Duck, Killdeer, Least Sandpiper, Dunlin, and dowitchers (presumably mostly or all Long-billed Dowitcher) exceeded 1000 individuals. The number of avian groups and species recorded in particular crop treatments ranged from 4 to 6 and 6 to 37, respectively (Table 3). 8

10 Winter Period The use of crops and treatments varied considerably among groups of waterbirds as detailed below. Shorebirds. Shorebirds were most strongly associated with flooded post-harvest corn, wheat, and potatoes (Table 4). Flooded chop-and-roll corn had a significantly higher shorebird density than all other corn treatments (including flooded harvest-only), irrigated pasture, dry harvested wheat, and newly planted wheat. Although shorebird density in flooded chop-and-roll corn was considerably lower than in flooded post-harvest fields of both winter wheat and potatoes, these differences were not significant (Table 4; see Appendix 1 for density estimates and 95% credible intervals). Still, flooded post-harvest potato fields and flooded winter wheat had significantly higher densities compared to most other treatments. Although there were some significant relationships when comparing treatments in dry fields (Table 4), shorebird densities in them were so low (Appendix 1) that overall dry treatments provide limited benefits to shorebirds. Mechanism models suggested that shorebirds had a significant positive association with the amount of moist soil (Table 5). Shorebirds were also significantly associated non-linearly with water depth, with density increasing as water gets deeper but subsequently declining as water gets too deep. Cranes. Overall, the densities of Sandhill Cranes were higher in dry than in flooded treatments for all crops (Table 6). Crane density was significantly higher in dry than in flooded chop-and-roll corn, but there was no significant difference between densities in the dry and flooded variants of harvest-only corn. The highest density for cranes was in dry potato fields (see Appendix 1), and the density was significantly higher in that treatment in 5 of 9 pairwise comparisons (Table 6). Flooded potato fields, however, had significantly lower densities in 9 of 9 pairwise comparisons. Crane density in planted winter wheat was significantly lower than in other treatments in 7 of 9 pairwise comparisons and was significantly higher only with respect to flooded potato fields. Predictably, based on the comparisons above, mechanism models indicate that crane abundance had a significant negative association with the amount of field flooding and water depth (Table 5). Waders. Overall, densities of long-legged waders varied little across the range of croptreatment combinations (Appendix 1). The only significant pairwise difference was a higher density in flooded chop-and-roll corn than in dry harvest-only corn. None of the mechanisms evaluated showed a significant association with the abundance of long-legged waders (Table 5). Evaluation of a non-linear depth effect was not successful, as the quadratic depth term would not converge to an r-hat less than 1.5. Geese. Geese had high densities in some flooded and some dry crop treatments (Appendix 1). The density of geese was significantly higher in flooded chop-and-roll corn than in any of the other corn treatments, flooded or non-flooded (Table 8). All corn treatments, except for flooded chop-and-roll corn, had significantly lower densities of geese than in irrigated pasture and planted winter wheat. Two of the four corn treatments flooded harvest-only and dry chop-and-roll had significantly lower densities than in flooded potatoes. Overall planted 9

11 winter wheat had the highest density of geese (significantly higher in 7 of 9 comparisons). Likewise, irrigated pasture and flooded potatoes also had significantly higher densities in 6 of 9 and 3 of 9 comparisons, respectively. Mechanism models were not evaluated for geese because of severe overdispersion in the count data. Ducks. Dabbling and diving ducks both had large differences in densities between some flooded treatments, but, given low sample sizes of treatments, the only significant relationship was the higher density of dabbling ducks in flooded potato fields than in flooded winter wheat (Table 9). Generally, flooded potato fields had the highest density of both dabbling and diving ducks. Dabbling ducks had a significant negative association with the amount of stubble and a significant non-linear association with water depth (Table 5). Though not significantly associated with stubble, diving ducks also exhibited a non-linear association with water depth, with abundance initially increasing with water depth then declining at the deepest depths (Table 5). Species richness. The number of waterbird species was positively associated with flooded chop-and-roll corn, which had a significantly higher mean richness in 5 of 9 treatment comparisons (Table 10). Conversely, richness was negatively associated with flooded potato fields, which had a significantly lower richness in 9 of 9 comparisons (however, see Discussion). Species richness was not significantly different in dry chop-and-roll corn than in dry harvest-only corn. Dry harvest-only corn, however, had significantly lower species richness in 4 of 9 treatment comparisons, making it the winter treatment with the second lowest value for species richness of waterbirds. Mechanism models indicated a significant positive effect of variation in water depth and percent flooded on species richness (Table 5). Fall Period Patterns of waterbird use were not as clear in fall as they were in winter. Shorebirds in fall had a significantly higher density in flooded potato and wheat fields than in irrigated pastures. Shorebird densities in the former two crops, however, were not significantly different than in dry potato or wheat fields (Table 11). Overall, Sandhill Cranes densities showed no significant differences in crop-treatment comparisons for the fall. Still, cranes had their highest density in dry potato fields, but low samples sizes prevented strong inference for this treatment relative to the four others evaluated. Long-legged waders had a significantly higher density in irrigated pastures than in dry potato or wheat fields, but their density in pastures was not significantly different than in flooded potato and wheat fields. Wader densities were higher in both flooded potato and wheat fields than in dry wheat fields. The density of geese was highest in irrigated pasture, and was significantly greater (>10 birds per ha more) in pastures than in dry potato fields; densities also were significantly higher in flooded potato and flooded wheat fields than in dry potato fields (Table 11). Dabbling duck density was higher, but not significantly, in flooded potato than in flooded wheat fields. The density of diving ducks was highest in flooded potato fields; diving ducks were not observed in flooded wheat in the fall. Waterbird species richness was significantly higher in irrigated pasture than in flooded wheat; we found no other significant differences in waterbird richness (Table 11). 10

12 COMPREHENSIVE SURVEYS Abundance On the eight all-island surveys of large species of waterbirds from November to February , we counted a total of 109,515 individuals of 11 species. Collectively, geese, cranes, and swans accounted for about 94% of birds counted. Of the geese, 55,989 were dark forms (Cackling Goose 37,273, Greater White-fronted Goose 16,745, unidentified dark goose 1536, Canada Goose 435) and 12,163 were white forms (unidentified white goose 6392, Snow Goose 5756, Ross s Goose 15). The Sandhill Crane (34,591) was the second and the Tundra Swan (6346) the fifth most numerous species counted (assuming most unidentified white geese were Snows). Other large waterbirds counted included three species of long-legged waders (Great Egret 321, Snowy Egret 42, and Great Blue Heron 41) and the American White Pelican (22). Behavior Patterns by Crop On all-island surveys, the ratio of birds foraging to those roosting/loafing varied among species groups and across crop treatments (Figs. 4 and 5). Dark geese generally were foraging in higher proportions in dry treatments than in flooded ones, where roosting was more common. We recorded white geese in only dry and flooded chop-and-roll corn, where they were predominantly foraging; those that were roosting/loafing occurred just in flooded fields. Tundra Swans roosted in slightly higher proportion in flooded chop-and-roll corn than in flooded harvest-only corn or flooded potatoes. Conversely, the proportion of swan foraging was higher in flooded harvest-only corn than in the other flooded treatments (Fig. 5). Overall, dry harvest-only corn and dry potato fields had the highest proportions of loafing Sandhill Cranes, but high proportions of foraging cranes occurred in dry post-harvest winter wheat. Foraging cranes were also more common than loafing cranes in flooded potato fields, though this may reflect birds foraging in the dry portions of partially flooded fields. Overall, long-legged waders were foraging the majority of the time in all crop treatments with few birds observed loafing. Distribution and Temporal Patterns Patterns of distribution of waterbirds at Staten Island varied among species and over the course of the winter (Figs. 6 15). Overall, for geese there was a trend of decreasing abundance in the northern and west-central portions of the island and increasing abundance in the southern portion from early to late winter (with limited use of dry corn fields on the eastern side of the island at any time). For all dark geese combined, this appears to be driven largely by the concentration of Cackling Geese in the north in November to December and Greater White-fronted Geese dominating in the south in January to February. Further contributing to this pattern was the temporal trend of island-wide abundance of Cackling Geese dropping sharply by mid-december and remaining low thereafter (Fig. 13); their spatial use of the island remained roughly similar over the winter, but the total number of fields they used declined each month. 11

13 By contrast, numbers of Greater White-fronted Geese remained fairly constant (Fig. 13), but their distribution shifted, across the season (Fig. 9). In early winter, they occurred almost exclusively in flooded fields to the north and west but later most extensively in the island s southern portion, particularly in dry corn fields. Though overall less abundant than Greater White-fronteds, white geese (mainly Snow Geese) occurred primarily in the northern portion of the island in early winter and the southern portion in late winter. Our anecdotal observations indicate that early in the season most white geese were foraging off the island but coming to roost in flooded fields in the northern portion of the island at midday. Conversely, later in the season white geese were foraging at the southern end of the island in the morning, but observations indicate they were leaving the island to roost elsewhere at midday. Sandhill Cranes were the most widespread of the large waterbird species on Staten Island, but they also became more numerous (Fig. 14) and used more fields in the southern portion of the island in late winter than earlier in the season (Fig. 6). Tundra Swans were highly aggregated in flooded fields in the northern and west-central portion of the island throughout the winter (Fig. 11), though by early February their numbers had declined substantially (Fig. 13). Herons and egrets occurred widely across the island in small numbers throughout the winter (Fig. 12), with egret numbers increasing through the winter (Fig. 14). These wader species foraged in both flooded and dry fields and along the main drainage canals between fields. DISCUSSION Agricultural landscapes provide important habitat for many bird species (Taft and Elphick 2007). This is particularly the case in the Central Valley of California where >90% of the natural wildlife habitat has been lost (CVJV 2006) and studies have quantified the benefits of selected agriculture crops to waterbirds (e.g., Elphick and Oring 1998, Shuford et al. 1998). Increasingly, there are opportunities for wildlife managers and biologists to work with farmers and ranchers to promote wildlife-friendly practices. Understanding the relative value of specific crops and crop management practices provides essential data to guide management and conservation. Our study quantified the relationship of different crops treatments to the abundance of waterbirds at Staten Island in the Delta region of the Central Valley. We documented significant variation in bird use among crop types and treatments. Although most waterbirds were predictably common in flooded treatments, geese, cranes, and long-legged waders also were numerous in some dry treatments. Our data suggest that maintaining a mosaic of habitats on the landscape at Staten Island will promote a diverse community of waterbirds during fall migration and winter. The precise recommended composition of these habitats, however, requires defining population and management objectives for species using Staten Island. RELATIVE VALUE OF CROPS A primary interest in this study was the effect on waterbird use of post-harvest management in corn, the dominant crop type used by waterbirds at Staten Island and in the broader Delta region. Flooded chop-and-roll corn was important for dabbling ducks, 12

14 shorebirds, geese, and cranes, though this treatment was not necessarily more important for these groups than all other corn treatments. Of the two dry treatments of corn, chop-and-roll (primary method at Staten) had significantly higher densities of geese and shorebirds than harvest-only (common throughout the corn-growing regions of California). Among avian groups using dry corn treatments, geese and cranes had some of the highest densities, shorebirds some of the lowest (Appendix 1). For geese, the chop-and-roll treatment may be preferable because it provides greater access to residual grain, but food availability studies would be required to confirm this. Overall, sampled corn fields that were only harvested, whether subsequently flooded or not, did not have significantly greater densities of waterbirds than fields that were chopped and rolled postharvest. Collectively, this evidence provides a scientific basis for promoting the chop-and-roll post-harvest practice in corn, whether the fields are subsequently flooded or not. The Central Valley Joint Venture Implementation Plan (CVJV 2006) considered winterflooded rice and corn as the primary agriculturally based habitat resources for shorebirds and waterfowl in the Central Valley. Our data confirm that both shorebirds and waterfowl will use flooded corn, although those bird groups, along with long-legged waders, occur in lower densities in flooded corn at Staten than in winter-flooded rice in the Sacramento Valley (Strum et al. 2010). Other crops we evaluated at Staten Island, including potatoes, wheat, and irrigated pasture, also provided substantial benefits for waterbirds yet varied by crop treatment. In winter, shorebird densities in flooded treatments were 16x higher in potato fields and 6x higher in wheat fields than in corn. Densities of shorebirds in flooded potatoes and wheat were more comparable to those found in flooded rice fields in the Sacramento Valley (Elphick and Oring 1998, Strum et al. 2010) than those in flooded corn. Dry potatoes also had a relatively high density of shorebirds compared to flooded corn, but this may mainly reflect shorebird use on just one survey when 20% of the field was flooded by recent rains. This observation highlights the importance of winter precipitation as a factor contributing to the availability of waterbird, particularly shorebird, habitat in the agricultural landscape of the Central Valley. In fall, shorebird densities again were highest in flooded wheat and potatoes (Appendix 2). Density estimates for dabbling and diving ducks also were higher in flooded potatoes and wheat than in flooded corn. Also, planted winter wheat had some of the highest use by geese relative to other crop treatments. Although we documented significant use of both potato and winter wheat fields flooded post-harvest, we recommend caution in extrapolating these results. Staten Island has a history of wildlife-friendly management practices (including flooding), which may influence the overall use of the island. Whether flooded potato or wheat fields would have similar benefits at a farm with no history of flooding is unknown. Neither wheat or potatoes are currently grown extensively in the Delta, and hence it is unclear how much of this habitat could be created. It may be, though, that post-harvest flooding of other crops, including corn, may provide comparable benefits if fields are tilled in a manner similar to that for wheat and potatoes before 13

15 flooding. The tilling practice for those crops on Staten Island greatly reduces the extent of exposed residual stubble, which should be advantageous to dabbling ducks and shorebirds given both groups show a negative association with stubble. An advantage to wheat and potatoes over corn, however, is their harvest schedule, which allows them to be flooded in the fall for early migrating cranes, ducks, and shorebirds before flooding of corn fields is possible. MULTISPECIES MANAGEMENT Our study highlights the challenges associated with managing for multiple waterbird species. In winter, overall, the densities of Sandhill Cranes were higher in dry than in flooded treatments for all crops. The highest density for cranes was in dry potato fields, and density was significantly higher in dry than in flooded chop-and-roll corn. Shorebirds, however, were found at their highest densities in flooded treatments during winter. By contrast, planted winter wheat had little benefit for shorebirds and cranes relative to other crop treatments but was one of the most beneficial crops for geese. Similarly, irrigated pasture had little value for shorebirds but substantial value for geese. Species richness was highest in fields that had the highest diversity of water depths and amount of flooding over the course of the winter. The differential value of crop treatments across bird groups at Staten Island emphasizes the need to adequately define waterbird population objectives and then establish the needed composition of cover types on the landscape to meet those objectives. MECHANISMS EXPLAINING USE Some field characteristics helped explain waterbird use of certain crops and treatments and by extension identified best management practices. Our analyses support the need to manage for a variety of water depths and to minimize residual stubble to support a diversity of waterbird guilds. Shorebirds, dabbling ducks, and diving ducks all had a significant non-linear (quadratic) association with water depth, similar to the relationships observed in winter-flooded rice in the Sacramento Valley (Elphick and Oring 1998, Strum et al. 2010). Waterbird abundance increased with water depth up to a depth threshold beyond which the value of the habitat declined. The depth at which bird use began to decline in flooded fields at Staten Island was about 15cm for shorebirds, 34 cm for dabbling ducks, and 39 cm for diving ducks. Unlike in the Sacramento Valley (Strum et al. 2010), we did not find a significant association of water depth with long-legged waders at Staten Island. Cranes at Staten had a negative association with water depth; similarly, cranes avoided roosting in wetlands in the northern Sacramento Valley at night when water depths exceeded a certain threshold (Shaskey 2012). We also found that dabbling ducks were negatively associated with areas with higher amounts of residual stubble. Field characteristics also appeared to influence species richness of waterbirds. Although our pairwise comparisons found that species richness of waterbirds in winter was highest in the flooded chop-and-roll corn and lowest in flooded potato fields, the low richness in potato fields appeared to be related to low variation in water depths in the sampled portions of those fields. Compared with the much larger set of flooded corn fields sampled, the three flooded potato fields sampled were nearly entirely flooded throughout the study period, generally had water depths that were deeper and varied less than in flooded corn, and had limited amounts of moist 14

16 soil (Fig. 3). When conducting a post hoc evaluation of data from surveys of the entire potato fields with that from the sampled areas, the former had much greater variation in water depth and hence much higher species richness. Our mechanism models for species richness further supported the cause of the perceived difference, in that species richness increased significantly with increasing variation in both water depth and the percent of the survey area that was flooded. These results similar to those in rice in the Sacramento Valley (Elphick and Oring 1998, Strum et al. 2010) and in managed wetlands of the San Joaquin Valley (Isola et al. 2000) suggest that managing for a diversity of depths, no matter the crop or wetland type, will likely maximize waterbird diversity. SEASONAL SHIFTS IN USE The comprehensive (all-island) surveys conducted during provided evidence of variable patterns of spatial use among species and species groups over the winter. The spatial distribution patterns of Tundra Swans and long-legged waders, though differing, remained relatively constant across the winter. By contrast, the overall pattern for geese and cranes was much more dynamic. Generally their abundance increased in the southern portion of the island from early to late winter, and throughout the season there was evidence of movements of these species between Staten and adjacent islands. For cranes, the increased use of the south part of Staten Island later in winter may simply reflect increasing numbers of cranes on the island overall rather than a north-to-south shift in field use. These variable patterns demonstrate the dynamic nature of spatial and temporal use by large waterbirds, and that the scale of use for some extends beyond the confines of Staten Island. Temporal changes in the resource base on Staten Island may influence these patterns. The ones most likely to affect on-island distribution patterns of waterbirds are the amount and location of flooded fields and the depletion of waste grain over the season. The temporal trends in flooding over the course of the winter were modest in Flooding was largely confined to the northern and west-central portions of the island, except for the addition of a limited number of flooded fields at the south end in December. There was substantial waterbird use of the later-flooded fields to the south, but the increasing numbers of geese to the south in late winter were foraging mainly in dry corn fields in the southwestern portion of the island. Corn harvest in fall progresses from north to south, so it is possible that waste grain available earlier in the season to the north may be depleted sooner than in the south. Food availability studies would be needed to confirm this, but the highly variable patterns of on-island distribution of geese and cranes, suggest this is not a major driving factor. Cranes and geese are well known to move among the Delta islands, so patterns of bird occurrence on Staten are quite likely influenced, to an unknown degree, by the availability of suitable habitat for these birds on the larger landscape. FUTURE DIRECTIONS Although our study evaluated the relative value to waterbirds of most current management on Staten Island, there is more to learn to refine best management practices and to develop exportable practices. It would be valuable to conduct additional studies on Staten 15

17 Island to better document the mechanisms that drive bird use of particular crops and treatments. Food availability studies could quantify how resources may vary on the island temporally. In particular, it would be important to assess if birds begin by heavily exploiting waste grain in corn fields to the north of the island harvested first then move to fields in the south harvested later in the fall. Also, it would be valuable to make frequent behavioral observations when new sets of fields are flooded away from others to see if this leads to any large shifts in distribution on the island. To broaden the scope of potential management options, we recommend also assessing the value of different practices used on other farms in the region to see if any of these might have benefits on Staten. Finally, it is clear that some species, particularly cranes and geese, depend on a much larger landscape than Staten Island and understanding their patterns of use will require work across a larger area of the Delta (see below). Concern that numbers of cranes have been declining on Staten Island in the last decade (G. Ivey and M. Eaton pers. comm.) should be evaluated and addressed. Given the importance of Staten Island for cranes, and their broader conservation concern, it seems important to design a robust monitoring program for cranes on the island that can detect biologically significant changes in abundance. The all-island surveys conducted in could serve, with modifications, as a basis for such a program. We recommend that a monitoring program also include other species of large waterbirds, particularly those that use the same food resources as cranes. A framework for adaptive management (Walters 1986) could be developed and adopted to guide annual planning on Staten. This would require setting specific population or conservation goals, developing and implementing strategies to achieve them, monitoring to see if goals are met, refining strategies as needed, and conducting research to answer questions that would help achieve conservation goals. Given the wide-ranging nature of many species of migratory waterbirds, it is also important to expand work beyond Staten Island and conduct surveys of the abundance and distribution of waterbirds in crop and wetland habitats on surrounding Delta islands. This research will evaluate to what degree the patterns of bird use on Staten are explained solely by crop treatment patterns at Staten versus the features of crops and other habitats on the surrounding landscape. Studies in the Sacramento Valley suggest that waterbird use of winterflooded rice fields increases when those fields are located near managed wetlands (Elphick 2008). Whether a similar relationship exists in the Delta is not known, but it is essential to understand the importance of such landscape factors to enable optimal waterbird management and conservation actions. Collecting data to better understand patterns of waterbird use across the Delta landscape, which post-harvest practices best support birds, and which bird species prefer which habitats would inform analyses to determine the optimal allocation of habitat across Staten and the broader Delta region. These broader-scale monitoring data and models could be used to identify conservation prioritization strategies using several weighting scenarios that incorporate variation in the 16

18 relative importance of crops to different guilds and seasonal variation in bird abundance and crop use. A spatially explicit prioritization exercise (e.g., Moilanen 2007) could also extend to evaluate impacts of climate change and of proposed changes to habitats as part of the Bay Delta Conservation Plan process. This work is needed now given the potential for significant habitat change in the Delta in the near future. This expanded research would inform conservation actions and establish a baseline needed to assess impacts of changes on waterbird populations and to inform mitigation strategies as changes occur. Lastly, as part of future work, it is important to quantify the overall prevalence of postharvest practices in the region and their agronomic costs and benefits. This would help to build a full cost-benefit model of these practices and to subsequently evaluate the potential to convince other farmers, perhaps under incentive-based programs, to implement beneficial practices and thereby expand waterbird conservation more broadly in this agricultural landscape. ACKNOWLEDGMENTS We are grateful for the participation and support of Brent Tadman and other staff of Conservation Farms and Ranches at Staten Island. Greg Golet at TNC and Gary Page at PRBO Conservation Science have been instrumental throughout the development and implementation of this project. Sara Sweet at TNC provided a research permit for work on Staten. We thank Tim Guida for conducting multiple surveys and performing most data entry in ; Ryan DiGaudio and Jennifer Roth also helped on occasional surveys. Ryan DiGaudio kindly shared the cover photo. Nat Seavy offered helpful discussions on study design, and Gary Ivey shared insights on bird use from prior studies on Staten Island. Mike Eaton, Greg Golet, Gary Ivey, Mike Savino, Sara Sweet, Paul Tebbel, and Sean Wirth provided valuable comments on an earlier version of the report. Funding for this project was provided by Contract No from the Northern Central Valley Office of The Nature Conservancy. This is Contribution No of PRBO Conservation Science. 17

19 LITERATURE CITED Aebischer, N. J., P. A. Robertson, and R. E. Kenward Compositional analysis of habitat use from animal radiotracking data. Ecology 74: Calenge, C The package adehabitat for the R software: A tool for the analysis of space and habitat use by animals. Ecological Modeling 197: Central Valley Joint Venture Central Valley Joint Venture Implementation Plan Conserving Bird Habitat. U.S. Fish and Wildlife Service, Sacramento. Elphick, C. S Landscape effects on waterbird densities in California rice fields: Taxonomic differences, scale-dependence, and conservation implications. Waterbirds. 31: Elphick, C. S., and L. W. Oring Winter management of Californian rice fields for waterbirds. Journal of Applied Ecology 35: Fleskes, J. P., J. L. Yee, M. L. Casazza, M. R. Miller, J. Y. Takekawa, and D. L. Orthmeyer Waterfowl distribution, movements and habitat use relative to recent habitat changes in the Central Valley of California: A cooperative project to investigate impacts of the Central Valley Habitat Joint Venture and changing agricultural practices on the ecology of wintering waterfowl. Final Report, U.S. Geological Survey-Western Ecological Research Center, Dixon Field Station, Dixon, Calif. Gelman, A., and J. Hill Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press, New York. Golet, G. H Conservation needs and opportunities at Staten Island Ranch, San Joaquin County, California. Unpublished report of The Nature Conservancy. Heitmeyer, M. E., D. P. Connelly, and R. L. Pederson The Central, Imperial, and Coachella valleys of California, in Habitat management for migrating and wintering waterfowl in North America (L. M. Smith, R. L. Pederson, and R. M. Kiminski, eds.), pp Texas Tech. Univ. Press, Lubbock, Texas. Isola, C. R., M. A. Colwell, O. W. Taft, and R. J. Safran Interspecific differences in habitat use of shorebirds and waterfowl foraging in managed wetlands of California s San Joaquin Valley. Waterbirds 23:

20 Ivey G. L., and C. P. Herziger Sandhill Crane monitoring at Staten Island, San Joaquin County, California, Available from The Nature Conservancy, Cosumnes River Preserve, Franklin Blvd., Galt, CA or at Ivey, G. L., C. P. Herziger, and M. Gause Farming for wildlife: An overview of agricultural operations at Staten Island, San Joaquin County, California. Report to The Nature Conservancy. Kery, M Introduction to WinBUGS for ecologists. Academic Press, Burlington, Massachusetts. Manly, B. F Randomization, bootstrap, and Monte Carlo methods in biology. Chapman and Hall, Boca Raton, Florida. Moilanen, A Landscape zonation, benefit functions and target based planning: Unifying reserve selection strategies. Biological Conservation 134: National Agriculture Statistics Office (NASS) California Cropland Data Layer. U.S. Department of Agriculture. Pogson, T. H., and S. M. Lindstedt Distribution and abundance of large Sandhill Cranes (Grus canadensis tabida) wintering in California s Central Valley. Condor 93: Shaskey, L. E Local and landscape influences on Sandhill Crane habitat suitability in the northern Sacramento Valley, CA. Unpublished M.S. thesis, Sonoma State University, Rohnert Park, California. Shuford, W. D. In press. Coastal California (BCR 32) Waterbird Conservation Plan. A plan associated with the Waterbird Conservation for the Americas Initiative. Published by U.S. Fish and Wildlife Service, Region 8, Sacramento. Shuford, W. D., G. W. Page, and J. E. Kjelmyr Patterns and dynamics of shorebird use of California's Central Valley. Condor 100: Spiegelhalter, D. J., A. Thomas, N. G. Best, and D. Lunn WinBUGS version 1.4 User Manual. MRC Biostatistics Unit, Cambridge, United Kingdom. Stevens, D. L., and A. R. Olsen Spatially balanced sampling of natural resources. Journal of the American Statistical Association 99:

21 Sterling, J Review of literature and information on the bird use of certain agricultural crops in California s Central Valley. Report to the Nature Conservancy. Strum, K. M., M. E. Reiter, C. A. Hartman, C. A. Hickey, and R. Kelsey Evaluating alternative approaches to managing winter water for waterbirds in Rice: Year 1 Report. Progress Report to the Migratory Bird Conservation Partnership. Sturtz, S., U. Ligges, and A. Gelman R2WinBUGS: A Package for Running WinBUGS from R. Journal of Statistical Software 12:1 16. Taft, O. W., and C. S. Elphick Waterbirds on working lands: Literature review and bibliography development. National Audubon Society, Inc., New York. Walters, C. J Adaptive Management of Renewable Resources. McGraw Hill, New York 20

22 Figure 1. Location of crop types at Staten Island in the season. 21

23 Figure 2. Location of crop types at Staten Island in the season. 22

24 Figure 3. Distribution of survey area characteristics in each treatment during fall and winter surveys of waterbirds at Staten Island, The box-and-whisker plots represent the median value of the distribution, the location of the first and third quartiles, and the minimum and maximum values observed. CCRD, dry chop-and-roll corn; CCRF, flooded chop-and-roll corn; CHD, dry harvest-only corn; CHF, flooded harvest-only corn; IP, flood-irrigated pasture; PHD, dry harvested potato fields; PHF, flooded harvested potato fields; WHD, dry harvested winter wheat; WHF, flooded harvested winter wheat; WW, planted winter wheat. 23

25 Forage Loaf CCRD CCRF CHF CHD IP PHF PHD WHD WW Figure 4. Average proportion of dark geese (top), white geese (center),and Tundra Swans (bottom)observed foraging and roosting/loafing by crop and treatment during comprehensive surveys of Staten Island, CCRD, dry chop-and-roll corn; CCRF, flooded chop-and-roll corn; CHF, flooded harvest-only corn; CHD, dry harvest-only corn; IP, flood-irrigated pasture; PHF, flooded harvested potato fields; PHD, dry harvested potato fields; WHD, dry harvested winter wheat; WW, planted winter wheat. 24

26 Forage Loaf CCRD CCRF CHF CHD IP PHF PHD WHD WW Figure 5. Average proportion of Sandhill Cranes (top) and long-legged waders (bottom) observed foraging and loafing by crop and treatment during comprehensive surveys of Staten Island, CCRD, dry chop-and-roll corn; CCRF, flooded chop-and-roll corn; CHF, flooded harvest-only corn; CHD, dry harvest-only corn; IP, flood-irrigated pasture; PHF, flooded harvested potato fields; PHD, dry harvested potato fields; WHD, dry harvested winter wheat; WW, planted winter wheat. 25

27 Figure 6. Distribution of Sandhill Cranes by survey month from comprehensive surveys of Staten island,

28 Figure 7. Distribution of dark geese (Cackling, Canada and Greater Whitefronted geese) by survey month from comprehensive surveys of Staten Island,

29 Figure 8. Distribution of Cackling Geese by survey month from comprehensive surveys of Staten Island,

30 Figure 9. Distribution of Greater White-fronted Geese by survey month from comprehensive surveys of Staten Island,

31 Figure 10. Distribution of white geese (Snow and Ross s geese) by survey month from comprehensive surveys of Staten Island,

32 Figure 11. Distribution of Tundra Swans by survey month from comprehensive surveys of Staten Island,

33 Figure 12. Distribution of long-legged waders (Great Blue Heron, Great Egret, and Snowy Egret) by survey month from comprehensive surveys of Staten island,

34 Figure 13. Seasonal patterns in abundance of dark geese (Cackling, Canada, and Greater White-fronted geese), white geese (Snow and Ross s geese), and Tundra Swans from comprehensive surveys of Staten Island, winter

35 Figure 14. Seasonal patterns in abundance of Sandhill Cranes and long-legged waders from comprehensive surveys of Staten Island, winter

36 Figure 15. Distribution of flooded fields on Staten Island by month as assessed on comprehensive surveys in