Water: Monitoring & Assessment
11.0 Bird Communities
Impacts on Quality of Inland Wetlands of the United States:
A Survey of Indicators, Techniques, and Applications of Community Level Biomonitoring Data
Excerpts from Report #EPA/600/3-90/073
(now out of print)
11.1 USE AS INDICATORS
This discussion addresses the monitoring of wetland birds, e.g., waterfowl, shorebirds, wading birds, and wetland-dependent songbirds. The use of birds as environmental indicators is discussed by Morrison (1986), Reichholf (1976), and particularly, by Temple and Wiens (1989). Statistical aspects of regional bird trend analysis are discussed in Sauer and Droege (1990). Advantages and disadvantages of using birds as indicators are summarized in Appendix A.
Because most vertebrates use wetlands at some time during the year, defining what truly constitutes a "wetland-dependent" bird species is difficult. One could argue dependency based on diet, energetics/metabolism, requirement for a particular structural component found only in wetlands, or duration of seasonal use. As with some wetland plant groups, many degrees of dependency occur, from species that spend their entire life in wetlands to species that use wetlands opportunistically and/or for only brief periods. Species that may be casual users of wetlands of a particular type in one region may be obligate users of a different type in the same region, or of the same type in a different region. Dependency in highly altered landscapes may be less related to the intrinsic characteristics of wetlands than to the fact that little other undeveloped habitat remains, forcing species that normally occur less frequently in wetlands to use what remains, regardless of its condition. In such cases, bird density may be a poorer indicator of habitat quality (the ability of the habitat to sustain successfully reproducing individuals over the long-term) than measurements of population demographics or measurements made at the organism level (Van Horne 1983). An empirical approach for testing wetland-dependence of birds is demonstrated by Finch (1990).
Monitoring of wetland birds, particularly waterfowl, has been extensive in many regions. Wetland birds can be categorized as (a) those most strongly dependent on larval insects, non-insect aquatic invertebrates, amphibians, fish, and submersed plants, and (b) those most strongly dependent on adult (terrestrial) invertebrates, emergent plants, and rodents. In general, the former group--which includes waterfowl and wading birds--tends to respond more immediately to contamination and water level changes than does the latter group--which usually includes marsh wrens, certain warblers, red-winged blackbirds, and swallows. Diets (and thus, guild assignments) of particular species can be confirmed through stomach content analysis or, less destructively, through close-range, automated photography of nest visits. In general, though, habitat requirements, life histories, and species assemblages of wetland birds are relatively well-known. Still, information on community-level response to particular stressors has been difficult to collect, in part because most bird species--as very mobile organisms--may be better at integrating overall landscape conditions than they are at indicating the conditions in a particular wetland.
Enrichment/Eutrophication. The effects of enrichment on overall community structure of birds are poorly documented in wetlands, and indicator assemblages of "most sensitive species" remain mostly speculative for this stressor. Weller and Spatcher (1965) defined a species assemblage that inhabits a "late marsh" successional stage, and species that inhabit the upland transitional zones of wetlands are well-known. However, the dominance of these species assemblages may be related as much to physical factors (geomorphology, fire, extreme climate events) as to nutrient enrichment. For Great Lakes wetlands, Crowder and Bristow (1988) hypothesized the following series of events that might lead to a waterfowl decline as a result of eutrophication:
"For the waterfowl, the effect of inshore eutrophication is thus an initial increase in food plants, a gradual replacement of favorite species by less desirable plants, and finally a total loss of submersed and floating-leaved plants coincident with an extension of cattail marsh. The extended marsh in turn declines, having been exposed to wave erosion through loss of the deeper zones of vegetation."
However, not all aquatic plants that increase with eutrophication are poor waterfowl foods. For example, in a study at Lake Okeechobee, Florida, Johnson and Montalbano (1984) found that waterfowl diversity in Hydrilla beds (a widespread, exotic species) was significantly greater than in several indigenous wetland plant communities (bulrush, cat-tail, pondweed, spikerush, and others).
Organic Loading/Reduced DO. The effects of severe organic loading, e.g., from wastewater outfalls, on overall community structure of wetland birds have been investigated in a few cases. Generally, abundance and/or on-site diversity of songbirds (Brightman 1976, Hanowski and Niemi 1989) and sometimes waterfowl (Belanger and Couture 1988, Piest and Sowls 1985) have tended to increase with increased abundance of aquatic invertebrates. The effect may depend on the type and configuration of the particular wastewater treatment system (Fuller and Glue 1980). Other bird groups have responded more to water levels (and associated effects on vegetation and invertebrates) than to contamination status (e.g., Ramsay 1978). In the Houghton Lake, Michigan, wetland that was exposed to treated wastewater, Kadlec (1979) reported no major shifts over a 3-year period in bird abundance or species composition.
Where introduction of organic wastes results in anoxic conditions lethal to fish and some amphibian larvae, community composition may shift from fish-eating species (e.g., herons, loons, grebes) to invertebrate-eating species and opportunists (e.g., shorebirds, songbirds, gulls, terns). Indeed, migrant shorebirds and gulls often appear to concentrate at sewage lagoons, turf farms, and wetlands mildly polluted with organic wastes (e.g., Campbell 1984, Fuller and Glue 1980).
Contaminant Toxicity. The effects of bioaccumulation of contaminants in wetland bird tissues have been widely measured, and the disasterous effects of naturally-occurring toxicants on community structure of wetlands have occasionally been documented (see discussion of Salinization below). Species assemblages for indicating the physical effects of oil spills can be easily identified based on characteristic behaviors of some wetland birds. However, the effects of pesticides, heavy metals, and other contaminants on overall structure of wetland bird communities are poorly documented in wetlands, and indicator assemblages of "most sensitive species" remain mostly speculative for these stressors (Grue et al. 1986).
Bird reproductive failure in wetlands from effects of heavy metal contamination (e.g., Scheuhammer 1987, Kraus 1989) and pesticides have been documented, but only for a few species. Lethal thresholds for metals and synthetic organics are reported in Hudson et al. (1984), EPA's "AQUIRE" database, and the US Fish and Wildlife Service's "Contaminant Hazard Reviews" series that summarizes data on arsenic, cadmium, chromium, lead, mercury, selenium, mirex, carbofuran, toxaphene, PCBs, and chlorpyrifos. However, relatively few field data are available for judging which wetland species are most sensitive. Additional testing of chemical toxicity to wildlife is currently being sponsored by EPA.
Numerous anecdotal reports exist describing relatively stable bird assemblages in traditionally-used wetlands even after years of progressive contamination. This might be attributed to the loss of nearby wetlands that otherwise would have been preferred, to behavioral avoidance of contaminated microenvironments and foods, and/or to replacement of contaminated individuals by immigrants.
Acidification. Naturally acidic wetlands sometimes have lower densities and species richness of birds, particularly in winter, than do non-acidic wetlands (Brewer 1967, Ewert 1982). Bird use of acid mine drainage wetlands in Pennsylvania was found to be less than use of natural wetlands, probably because of physical degradation of habitat rather than inferior water quality alone (Hill 1986). Acidification has also been demonstrated to reduce reproductive success and juvenile survival of some species in wetlands (e.g., tree swallows--Blancher and McNichol 1988, black ducks--DesGranges and Hunter 1987, ring-necked ducks--McAuley and Longcore 1988). Bird responses to anthropogenic acidification, summarized by McNichol et al. (1987), are felt indirectly as a result of alteration in the dominance of various food sources and possibly, changes in physical habitat (e.g., composition and distribution of submersed macrophytes). Shifts from fish to aquatic insects in lakes and streams can cause a corresponding shift from fish-eating species to those that critically depend on aquatic invertebrates, to those that feed on aquatic invertebrates opportunistically (assuming other habitat features remain relatively constant). Wetland bird groups in each category are listed in Table 14. Strong presence of a particular feeding group relative to others might be used to suggest acidification effects, if the role of other stressors (such as others listed in this section) can be ruled out.
|Predominantly feeding on fish or amphibians (at some season or life stage, in some regions):||loons,
some herons and egrets,
|Aquatic invertebrate obligates (at some season or life stage, in some regions):||some herons and egrets,
some dabbling ducks,
|Aquatic invertebrate facultatives:||most dabbling ducks,
many other songbirds (see Adamus 1987 for list for the Northeast)
Salinization. Breeding waterfowl in hypersaline wetlands reportedly prefer fresher portions of these wetlands, and inland wetlands that are naturally saline generally have fewer nesting waterfowl (Kantrud and Stewart 1977). However, high densities of a few species, e.g., Northern Phalarope, can occur during migration in some naturally saline wetlands. The effects of salinization on structure of wetland bird communities have not been widely studied, despite publicity given to events such as the catastrophic mortality at Kesterson National Wildlife Refuge. Assemblages of species that might be used as indicators of salinization remain speculative.
Sedimentation/Burial; Turbidity/Shade. Wetland bird species that prefer soft-bottomed wetlands can be defined, but probably with insufficient precision to warrant their use as indicators of excessive sedimentation. Sedimentation affects community structure of wetland bird communities primarily by altering the type and distribution of submersed plants, and perhaps also by affecting invertebrate food sources and interfering with feeding of birds that rely on visual cues.
Vegetation Removal. Effects of vegetation removal associated with grazing and/or fire are described by Fritzell 1975, Landin 1985, Schultz 1987, and others summarized by Kantrud (1986). "Moderate" levels of grazing and/or mowing, if occurring at a time in the season when nests are not disturbed, can increase wetland bird species richness in floodplain ponds (Landin 1985) and emergent wetlands (Nelson and Kadlec 1984). However, severe grazing, mowing, or fire at inappropriate times is detrimental (Duebbert and Frank 1984, Kantrud and Stewart 1984), and total removal of woody riparian vegetation dramatically alters species composition, density, and richness of the mammalian community (Cross 1985, Malecki and Sullivan 1987, Possardt and Dodge 1978).
Many species benefit from increased openings in dense stands of vegetation and from reduced floodplain ground cover, while others, including ground-nesting species such as Northern Harrier and Short-eared Owl (USDA Soil Conservation Service 1985), do not. As patches of open water are created in formerly continuous stands of emergent vegetation, the diversity of species using a wetland typically increases (Harris et al. 1983, Kaminski and Prince 1981). These may be species that are generally widespread in the region, so the contribution of vegetation removal to overall regional diversity of birds may be slight. Species assemblages associated with vegetation structural changes can be defined by region and wetland type. Brown et al. (1989), and Durham et al. (1985) have done so for vertebrates in bottomland hardwood wetlands, and Short (1983, 1989) for midwestern and Arizona wetlands.
Effects of silvicultural activities in forested wetlands have received only limited study. Birds in forested wetlands respond very strongly to changes in vertical and horizontal vegetation structure (Finch 1990, Rice et al. 1980). Because the habitat structural needs of most forested wetland birds are relatively well-known, at least qualitatively (e.g., see Durham et al. 1985, Swift et al. 1984), indicator associations could probably be easily developed that reflect bird response to different levels and types of silvicultural practices in forested wetlands. An old-growth forested floodplain wetland in South Carolina was compared by Hamel (1989) to clearcut and selectively cut portions of the same area. More species (and particularly cavity-nesters) achieved their highest densities in the old-growth habitat than in the disturbed forested wetland, and those species that did achieve higher density in the disturbed forested wetlands were widespread throughout the region. In a southwestern riparian wetland, Carothers et al. (1974) reported 46 percent fewer breeding birds where vegetation had been thinned to 25 trees per acre, as compared to a similar reference wetland with 116 trees per acre.
Thermal alteration. The effects of thermal alteration on overall community structure of birds are poorly documented in wetlands, and indicator assemblages of "most sensitive species" remain mostly speculative for this stressor. Effects of heated wastewater are mostly indirect, affecting habitat and bird distribution by prolonging ice-free conditions in northern wetlands, altering vegetation type and structure, and affecting the type and seasonal availability of food sources (e.g., Haymes and Sheehan 1982). On a regional level, species most sensitive to changes in temperature are often those occurring at the periphery of their geographic ranges. These are easily defined by local ornithologists.
Inundation/Dehydration. The response of bird community structure to water level alteration has been the subject of dozens of studies, many conducted to improve the management of waterfowl habitat. Water level alterations (either increases or decreases) can increase or decrease overall bird abundance and richness, depending on their duration and many other factors.
Both sustained increases and sustained decreases in water levels directly affect habitat availability and dramatically shift community composition. For example, construction of dams on the lower Colorado River produced a relatively stable environment that favored high invertebrate densities and consequently increases in diving ducks, but diminished numbers of riparian species (Anderson and Ohmart 1988). Alteration of the flooding regime of a southern forested wetland from seasonal flooding to permanent flooding (for a greentree reservoir) had little overall effect on bird diversity; waterfowl and common grackles increased while white-throated sparrow and a few other species decreased (Newling 1981).
Water level changes of short durations (weeks or months), while having less affect on habitat availability, have the potential for long-term impacts to habitat quality by altering water chemistry, invertebrate populations, and seed germination. For example, dam-induced alterations in hydrologic regime have decreased bird richness partly by encouraging the spread of non-native salt cedar (Tamarix spp.)(Ohmart et al. 1977, Hunter et al. 1985).
Addition of permanent open water to a non-permanently flooded wetland usually increases the opportunity for use by waterfowl and fish-eating birds. Moreover, the typical increase in submersed and floating-leaved plants that accompanies creation of a permanent pool within a wetland provides for a more diversified plant and invertebrate food source. This consequently can result in an increase in on-site species richness of birds. Many studies have found that productivity and diversity of waterbirds are greatest within basins having a permanently flooded pool or channel that is surrounded by shallowly flooded (<10 inches depth) wetlands that are gradually dehydrated at regular seasonal or frequent (3-5 year) annual intervals (Fredrickson and Taylor 1982, Reid 1985). Among a series of Massachusetts forested wetlands, Swift et al. (1984) found that the driest wetlands supported the lowest abundance and richness of birds, even though in some regions such wetlands have the greatest diversity of vertical habitat structure and plant species richness.
Wetland bird species vary in their water depth requirements and sensitivity to water level change. Much information on depth requirements is summarized in Fredrickson and Taylor (1982) and Fredrickson and Reid (1986). This information could be used to define hydroperiod "response guilds" of birds. The most sensitive species may be those which (a) nest along water edges (e.g., Western Grebe, Redhead--Wolf (1955), or (b) feed on mudflats (e.g., shorebirds), or (c) require a particular combination of wetland hydroperiod types in a region (e.g., Kantrud and Stewart 1984). In contrast, species with nests typically well-above water level (e.g., marsh wren, prothonotary warbler) may be less vulnerable. For arid, deep-water marshes in eastern Oregon, Littlefield and Thompson (1989) suggested that presence of yellow-headed blackbird might be a good indicator of ecologically "healthy" conditions.
Fragmentation of habitat. Only a single study (Brown and Dinsmore 1986, 1988) has looked directly at the effects of fragmenting regional wetland resources. Others had previously noted the effects of fragmentation, using knowledge of species-specific life histories or data from non-wetland forest systems. Essentially, as the distance between wetlands containing certain species becomes greater, and/or hydrologic connections and vegetation corridors become severed by dehydrated channels, bank-clearing, or (particularly) roads, species most dependent on wetlands and/or which do not disperse easily could be most affected. Moreover, some species require not just a particular density of wetlands, but a particular combination of wetland types (or wetland types and other land cover types) at a particular density on the landscape or in close proximity to each other (Cowardin 1969, Kantrud and Stewart 1984, Ohmart et al. 1985, Weller 1979, Flake 1979, Patterson 1976). Although individual birds, being highly mobile, can disperse to new areas having the proper combination of types at a sufficient density, this can cause diminished reproductive success and thus, non-sustainable populations.
Territorial size requirements of wetland birds are highly variable, but can be used (with empirical observations of presence/absence in wetlands of various sizes and degrees of isolation) to define assemblages of species that are likely to be most sensitive to habitat fragmentation (Brown et al. 1989). Such studies must employ a standard level of effort (e.g., censusing time) per unit area if results are to be comparable. Radiotelemetric methods can be used to track individuals and determine home range sizes under various combinations of landscape cover patterns (Hegdal and Colvin 1986 describe techniques).
Other Human Presence. Several studies (e.g., Brooks et al. 1990, Robertson and Flood 1980, Todt 1989) have reported changes in species composition of wetland bird communities in response to general watershed "development," reduction in natural land cover types surrounding the wetlands, and increased visitation of wetlands by humans. Developed areas are characterized by a typical suite of species that include European Starling, Rock Dove, American Crow, House Sparrow, American Robin, and perhaps a few others (Graber and Graber 1976).
Human disturbance can discourage use by wildlife (Pomerantz et al. 1988), especially (a) hunting (Conroy et al. 1987, Gordan et al. 1987) and people traveling on foot (Burger 1981), and (b) during the breeding season or under harsh weather conditions. Effects of noise disturbances on wildlife are summarized by Gladwin et al. (1988). The most sensitive species appear to be ducks, geese, and other long-distance migrants which feed in large flocks at the ground or water level (Burger 1981), as well as colonially-nesting species (e.g., Markham and Brechtel 1979, Tremblay and Ellison 1979) and large species (e.g., Stalmaster and Newman 1978). Sensitivity to human disturbance may also be species-specific. Reduced use of human-visited wetlands by waterfowl or nongame waterbirds has been demonstrated by Hoy (1987), Josselyn et al. (1989), and Kaiser and Fritzell (1984). To some extent, presence of screening vegetation can permit closer approach to waterbirds by humans (e.g., Milligan 1985).
Many waterbirds take flight when humans approach within 75 to 175 feet (e.g., Josselyn et al. 1989). Wintering bald eagles may take flight when approached from a distance of 800-1,000 feet (Knight and Knight 1984; Stalmaster and Newman 1978). Motorboat activities can disturb waterfowl up to 1,000 m away (Hoy 1987). This results in more time being spent in energetically costly behaviors. Disturbance can also increase the food consumption needs of waterbirds. Korschgen et al. (1985) found that only 5 boating disturbances per day increased the energy requirements of canvasbacks by 20 percent, requiring consumption of an additional 23 g of food daily.
Other direct human influences on wetland birds include mortality from collisions with vehicles and powerlines, and predation by hunters and housecats. Hunting comprises an obvious impact to certain wetland bird species, in some cases resulting in changes at the population level.
11.2 SAMPLING METHODS AND EQUIPMENT
Some factors that could be important to measure and (if possible) standardize among wetlands when monitoring anthropogenic effects on community structure of birds include:
distribution of water depth classes, vegetation (type, and vertical and horizontal diversity and arrangement), conductivity and baseline chemistry of waters and sediments (especially conductivity), current velocity, distance and connectedness to other wetlands of similar or different type, surrounding land cover (particularly within 500 feet of wetland perimeter), shoreline slope, wetland size, ratio of open water to vegetated wetland and its spatial interspersion, and the duration, frequency, and seasonal timing of regular inundation, as well as time elapsed since the last severe inundation or drought.
Methods for surveying bird communities are described by Burnham et al. 1980, Halvorson 1984, Ralph and Scott 1981, Verner 1985, Verner and Ritter 1985, and others. Censusing marsh and shorebirds specifically is discussed in detail by Connors (1986) and Weller (1986); censusing of waterfowl by Eng (1986) and Kirby (1980); censusing of colonial waterbirds by Speich (1986); and censusing of birds in bottomland hardwood wetlands, by Durham et al. (1985). An effort to refine techniques for monitoring wetland birds is presently being sponsored by the Maine Department of Inland Fisheries and Wildlife.
Even when apparently similar wetlands are censused, it is sometimes impossible to attribute changes in wetland bird communities to human activities within the wetland being sampled, because birds move widely across regions and continents. However, by calculating density-weighted ratios of declining resident to declining non-resident species (with similar habitat requirements), the possible role of this factor might be estimated.
Birds are present in wetlands throughout the year, but densities of birds vary greatly by season, depending on region. As with many other taxa, if only a single annual visit can be made, it should be timed to account for major life history events, such as nesting, molting, dispersal, migration, or wintering. The most severe reductions in bird density and richness occur in winter in northern emergent wetlands and bogs that completely freeze over. In southern wetlands, density and diversity are generally greatest in winter, while in northern wetlands, density and diversity are usually greatest in summer (Harris and Vickers 1984). During spring and fall, large numbers and high diversity may be present in either northern or southern wetlands. The species richness of wetlands in arid regions often increases the greatest during spring and fall, as many species seek temporary refuge during migration (e.g., songbirds in riparian oases, shorebirds in flooded fields).
A survey covering several wetlands should occur simultaneously or within consecutive days, unless severe weather conditions intervene. If the objective is to compare between-year trends in a species, total species, or species richness, then simple count methods (e.g., transects) are probably appropriate. However, if the objective is to rank wetland types or relative abundance of species, more time-consuming censusing to develop estimates of density are required (Steele et al. 1984). Determination of indices of relative annual abundance, rather than exhaustive population censusing, is suitable for most purposes (Emlen 1981).
For reasonably accurate estimates of breeding bird richness in a wetland, three visits spread over the breeding season may be desirable (Brooks et al. 1989, Weller 1986). Sampling non-wetland environments, Steele et al. (1984) reported that three repetitions of a 2 km transect were adequate to estimate bird abundance and richness of a habitat. In an inventory of birds in 87 Maine wetlands, Longcore et al. (pers.comm.) counted birds from an overview for two hours at dawn and two hours at dusk on at least two dates; as many observation points as necessary to view the entire wetland were used. The actual number and duration of visits required in a particular instance will depend on the size of the wetland, its habitat heterogeneity, visibility, and other factors.
If not only richness, but density, must be determined, then at least eight visits may be needed (Ralph and Scott 1981). Although most common songbirds will not be disturbed by frequent visits by monitoring personnel, raptors, waterfowl, other large or colonial species, and ground-nesting species may be susceptible. Wetland songbird surveys are commonly conducted in during May - July, when breeding birds are most detectable by song.
Species detection (especially of most songbirds) is greatest during early morning hours. However, thrushes and a few other species are more detectable in the evening, and in winter, some species may be most active at mid-day. Night-time coverage may be warranted, not only for typically nocturnal species such as owls, but also for waterfowl and wading birds which use different wetland types for roosting and feeding. Secretive species (e.g., rails, some passerines) have sometimes been surveyed more effectively by playing back of tape recorded calls, use of predator decoys, use of dogs, and by dragging ropes or chains through wetlands (e.g., Glahn 1974, Ralph and Scott 1981).
Surveys may be conducted from ground level, from elevated observation posts, or aerially. In the case of species that nest or roost colonially and in exposed locations, photography may be used to assist counting of individuals. Ground-level, visual techniques cannot be used effectively in wetlands with tall vegetation (mid-season emergent marshes, forested wetlands). Boats are typically used for surveys of wetlands wider than about 100 meters, as visibility from shore, even using a spotting scope, becomes restricted.
Many methods have been developed for monitoring wetland bird communities using visual, auditory, and capture techniques. These include point counts, line transects, nest counts, mist netting, and regional surveys (Brooks et al. 1989). Methods differ mainly in the degree of quantification they provide, the level-of-effort required, and the taxa they are most effective in censusing. These methods can be used in virtually all types of wetlands.
11.3 SPATIAL AND TEMPORAL VARIABILITY, DATA GAPS
Quantitative community-level data on birds have not been uniformly collected from a series of statistically representative wetlands in any region of the country. Thus, it is currently not possible to state what are "normal" levels in wetlands of various types for parameters such as bird density, species richness, biomass, or productivity. Data on temporal and spatial variability of wetland birds among wetlands and years has been systematically collected in only a few instances. These few data sets are available largely because of the existence of two important national data collection networks, which are described as follows.
The Breeding Bird Survey (BBS) database has existed since 1966, and includes all 50 states and some Canadian provinces. Data on bird relative abundance have been collected, usually recurrently, from about 2500 transects ("routes"), each 25 miles in length and containing 50 evenly-spaced data collection points. Density of coverage varies from 1 to 16 routes per degree (latitude-longitude) block. The survey routes are not located to intentionally intersect wetlands, so wetlands are included only randomly. Routes are run only once annually, so many species may be missed. Also, some routes are conducted later in the season than is optimal for detecting some wetland species. Because routes follow roads and rely largely on auditory detection more suitable for forest birds, they may further underestimate wetland species. Nevertherless, the BBS database, by its sheer quantity of spatial and temporal coverage, represents a valuable resource for helping define "average" bird densities (in relative terms) and for aiding detection of regional trends in wetland birds. Locations of routes are included on the state maps in Appendix B.
The Breeding Bird Census (BBC) database also provides useful information. This database is a compilation of individual censuses conducted by volunteers throughout the United States. Compared to methods used by the BBS, the BBC protocols are more intensive, but coverage is not nearly as extensive. Whereas the BBS measures only relative abundance using a single annual visit to an area, the BBC attempts to measure population density using repeated visits. The BBC also differs from the BBS in that some habitat data are collected, but habitat heterogeneity within census plots is not quantified, the acreage of censused plots is not consistent among censuses, and only a small portion of the plots are revisited annually. In most cases, census plots are too small and heterogeneous to adequately census species with large home ranges (Terborgh 1989), as is typical in wetlands. A few of the BBC's have focused exclusively on wetlands, but these wetlands have not been chosen randomly or systematically. These are included on the state maps in Appendix B. Selected data are presented in Tables 15 and 16, located at the end of this chapter. These are based on data compilation conducted by the Cornell Laboratory of Ornithology and sponsored by the EPA Wetlands Research Program. These tables are summarized in the following paragraphs.
Median number of breeding species ranged from 3.5 for all censused Florida wetlands to 51 for all censused Montana wetlands, where the national maximum of 68 species was found in one censused wetland (a bulrush-cattail marsh). As expected, salt marshes at all locations had the lowest number of breeding species. The greatest variability in species richness occurred among a set of 21 Wisconsin wetlands, a set of five Kansas wetlands, and a set of seven Florida wetlands. Most repetitively-censused wetland types (NUM>1) had less than 15 percent variation in species richness among years, and less than 10 percent variation in pair density among years.
Median density of breeding birds (i.e., pairs per square kilometer) ranged from 138 in Alaskan wetlands to 1857 in North Carolina wetlands. The two highest densities of all counts were from riparian willow woodlands in California. One, with 4547 pairs and 35 species, was dominated by Chesnut-backed Chickadee, Bewick's Wren, Song Sparrow, and Yellow Warbler. The investigator attributed the high density to extreme density of vegetation and abundant food, despite low plant diversity. The other remarkable California riparian count, 3208 pairs per km2 and 13 species, was dominated by Mourning Dove, Lazuli Bunting, Bewick's Wren, and Wilson's Warbler. Other high densities were in a California lacustrine marsh (3684 pairs, mainly Tricolored Blackbird), and in a cattail bulrush wetland in North Dakota (3418 pairs, mainly Yellow-headed Blackbird). The greatest variability of pair density among censused wetland types occurred among a set of four Nebraska wetlands (114 percent).
Table 15 summarizes the same parameters for each state/province, but does so by individual years of census. With regard to number of species, during a given year most states had less than 38 percent variability among their wetlands. Within any single year, the greatest variability in species richness among censused wetland types occurred between 2 Florida wetlands in 1983, which differed by 110 percent. With regard to pair density, during a given year most states had less than 54 percent variability among their wetlands. The greatest variability of pair density among censused wetland types occurred among 3 Colorado wetlands in 1973, which differed by 144 percent. In general, analysis of these 478 census plots showed the following statistically significant (p<0.05), linear relationships, based on log-transformed data:
- the median number of species was correlated with pair density and number of repeat censuses (years) on a plot;
- variability in number of species was inversely correlated with number of species;
- the median pair density was not correlated with number of repeat censuses (years) on a plot;
- variability in pair density was correlated with pair density and number of repeat censuses (years) conducted on a plot;
- variability in pair density was correlated with variability in number of species.
Despite their statistical significance, there was considerable scatter in all of these relationships, and the correlation coefficients (r) never exceeded 0.5.
Published studies (other than from the national databases described above) that have compared year-to-year or long-term variation in bird community structure in wetlands include Bellrose 1979, Blake et al. 1987, Harris et al. 1983, Hanowski and Niemi 1987, and Rice et al. 1980. Conceivably some unpublished data on annual variation in wetland bird communities may be available from sites of the U.S. Fish and Wildlife Service's Northern Prairie Research Station, the U.S. Department of Energy's National Environmental Research Park system, and the Illinois Pool 19 and Illinois-Mississippi Rivers sites of the National Science Foundation's Long Term Ecological Research (LTER) program.
Other national bird databases exist, and new ones are being developed, for example:
- International Shorebird Survey
- Christmas Bird Count database
- Colonial Wading Bird database
- Monitoring Avian Productivity (MAP) database
- Winter Bird-population Censuses
- Migratory Waterfowl Surveys
- Mid-winter Waterfowl Survey
- breeding bird atlases in dozens of states
None of these pertain exclusively to wetlands, and it is not always possible to separate the portion of the data that includes wetlands. Still, on a collective basis, these databases could be analyzed to yield more information on community structure in different regions and occasionally, in different wetland types. Overviews of some are provided by Muir and Davis (1989) and Terborgh (1989).
Lists of breeding wetland birds have been compiled by "block" (a unit generally smaller than about 50 sq. mi.) by statewide atlas projects in many states, and along with data from Christmas Bird Counts, other databases listed above, and records kept by thousands of volunteers, these can be used to define "expected" species in wetlands. Species that show highest affinity for wetlands of various types might be identified in discussions with local birders and by accessing the "Vertebrate Characterization Abstracts" database managed by The Nature Conservancy and various state Natural Heritage Programs. Limited qualitative information may be available by wetland type from the "community profile" publication series of the USFWS (Appendix C).
Quantitative data are most available for harvested groups, like waterfowl, and least available for the majority of wetland species, which are not harvested. In a survey of waterfowl migration/ wintering habitat in the United States, Bellrose and Trudeau (1988) reported the following to represent at least "moderate" densities of waterfowl (number of birds per acre per day):
|Dabbling Bay Ducks||Bay Divers||Geese|
Of studies that have compared bird community structure among many wetlands in a region (spatial variation), perhaps the most notable for their large sample sizes are those of bottomland hardwoods by Durham et al. 1985, and prairie potholes (Kantrud and Stewart 1984, Stewart and Kantrud 1973). The latter study--of 1321 wetlands--reported the following mean densities:
Other quantitative studies of multiple wetlands include:
Anderson and Ohmart 1985, Blake et al. 1987, Brewster et al. 1976, Briggs 1982, Brooks et al. 1987, 1989, Brown and Dinsmore 1986, DesGranges and Darveau 1985, Evans and Kerbs 1977, Flake et al. 1977, Hardin 1975, Harris and Vickers 1984, Heitmeyer and Vohs 1984, Hepp 1987, Hill 1986, Hudson 1983, Hunter et al. 1985, Klett et al. 1988, Knopf 1985, Landin 1985, Lawrence et al. 1985, Mack and Flake 1980, Maki et al. 1980, Menzel et al., Milligan 1985, Ohmart et al. 1985, Rector et al. 1979, Rice et al. 1980, Smith 1953, Stauffer and Best 1980, Swift et al. 1984, Wheeler and Marsh 1979, and others.
In summary, quantitative data on community composition of wetland birds is most available for breeding populations and least for wintering and migrating populations. Perhaps least-studied are montane wetlands; Northwestern wetlands; southeastern and southwestern herbaceous wetlands; and southern Great Plains wetlands. Information on impacts is most available for hydrologic alteration, vegetation removal, and acidification. Apparently the least information is available on impacts to community structure from eutrophication, sedimentation, contamination, and habitat fragmentation.
Table 15. Within-Year Variability of Breeding Bird Richness and Density, Among Wetlands, by State.
[not included in this Web page]
Table 16. Breeding Bird Richness and Density, by Wetland Type and State.
[not included in this Web page]