Kenneth F. Higgins

A LANDSCAPE APPROACH TO CONSERVING WETLAND BIRD HABITAT IN THE PRAIRIE POTHOLE REGION OF EASTERN SOUTH DAKOTA

Resource managers confronted with preserving ecosystems for prairie wetland birds in fragmented landscapes require landscape studies that direct conservation efforts over broad geographic regions. We investigated the role of local and landscape factors affecting habitat suitability by integrating remotely sensed wetland and land-cover data with wetland bird habitat models. We linked habitat models with locations of easement and fee-title wetlands to evaluate spatial location and extent of protected, suitable habitat. We also simulated impacts of the loss of small wetlands on suitability of larger wetlands for mobile species that use multiple wetlands. Lastly, we evaluated the efficacy of waterfowl habitat programs in preserving suitable habitat for nongame wetland bird species to recommend strategies for maximizing regional landscape connectivity. Regional databases constructed for this study indicate that easement and fee-title tracts encompass 13.9% (1.2 million ha) of land area and protect 19.8% of the wetlands in eastern South Dakota, USA Proportion of protected wetlands is highest for semi-permanent (32.3%), intermediate for seasonal (25.6%), and lowest for temporary wetlands (15.8%). A stratified, two-stage cluster sample was used to randomly select 834 semi-permanent and seasonal wetlands that were surveyed for birds in 1995 and 1996. Logistic analyses indicate that habitat suitability for some species (e.g., Virginia rail, pied-billed grebe) is related to local vegetation conditions within wetlands, while suitability for others (e.g., northern pintail, black tern) is related to landscape structure at larger scales. As a result, unfragmented prairie wetland landscapes (i.e., areas with wetland complexes embedded within upland grasslands) provide habitat for more species than isolated wetlands in tillage fields. Models developed from survey wetlands were used to classify habitat suitability for all semi-permanent and seasonal wetlands in eastern South Dakota. Small wetlands are critical components of the surrounding landscape that influence habitat suitability of larger wetlands. Models used to reclassify suitability of larger remaining wetlands after small wetlands (<0.5 ha) were removed indicate that species most vulnerable to loss of small wetlands are vagile species that exploit resources over broad spatial scales. Number of wetlands suitable for northern pintails, a mobile species that uses multiple wetlands within a season, decreased 20.7% when wetlands <0.5 ha were removed. Historic paradigms dictating waterfowl habitat protection efforts also have conserved habitat for nongame bird species. Modern paradigms that acknowledge the importance of small shallow wetlands to breeding waterfowl have shifted the focus of protection towards preserving habitat for species that occupy more abundant seasonal wetlands. Cessation of protection efforts would result in further fragmentation of regional wetland landscapes. We recommend that wetlands be acquired not only to consolidate suitable habitat within protected core areas but also to ensure that core areas coalesce to preserve connectivity among regional wetland landscapes.

A Cable-Chain Device for Locating Duck Nests

Abgract: A cable-chain device towed between two vehicles was developed for locating occupied duck nests in brushy, herbaceous, and grassy cover types. Twenty-three of 29 previously located gadwall (Anas strepera) and blue-winged teal (A. discors) hens were flushed from their nests with the drag for an efficiency of 79 percent. Eighty acres of nesting cover can be searched in 44; hours by the method described.

Bird mortality associated with wind turbines at the Buffalo Ridge wind resource area, Minnesota.

Abstract Recent technological advances have made wind power a viable source of alternative energy production and the number of windplant facilities has increased in the United States. Construction was completed on a 73 turbine, 25 megawatt windplant on Buffalo Ridge near Lake Benton, Minnesota in Spring 1994. The number of birds killed at existing windplants in California caused concern about the potential impacts of the Buffalo Ridge facility on the avian community. From April 1994 through Dec. 1995 we searched the Buffalo Ridge windplant site for dead birds. Additionally, we evaluated search efficiency, predator scavenging rates and rate of carcass decomposition. During 20 mo of monitoring we found 12 dead birds. Collisions with wind turbines were suspected for 8 of the 12 birds. During observer efficiency trials searchers found 78.8% of carcasses. Scavengers removed 39.5% of carcasses during scavenging trials. All carcasses remained recognizable during 7 d decomposition trials. After correction for biases we estimated that approximately 36 ± 12 birds (<1 dead bird per turbine) were killed at the Buffalo Ridge windplant in 1 y. Although windplants do not appear to be more detrimental to birds than other man-made structures, proper facility siting is an important first consideration in order to avoid unnecessary fatalities.

Lightning fires in North Dakota grasslands and in pine-savanna lands of South Dakota and Montana

Lightning strike fires which occurred between 1940 and 1981 were studied in mixed-grass prairie grasslands and in pine-savanna lands in the Northern Great Plains region. A majority (73%) of ignitions occurred during July and August, while a lesser number was recorded in April, May, June, and September. The AprilSeptember period is also the average time of the freeze-free period and approximates the average distribution period for thunderstorm activity in this region. The area burned by each of 293 lightning fires (most of which were suppressed) ranged from 0.0041158.3 ha (X = 10.8 ha). The frequency of lightning fires in mixedgrass prairie grasslands averaged 6.0/yr per 10,000 km2 in eastern North Dakota, 22.4/yr per 10,000 km2 in southcentral North Dakota, 24.7/yr per 10,000 km2 in western North Dakota, and 91.7/yr per 10,000 km2 in pine-savanna lands in northwestern South Dakota and southeastern Montana. The ecological role of lightning-set fires is discussed relative to the development of resource research and management plans and to the interpretation of historical records of natural fire occurrence in the Northern Great Plains region. Climate and fires are strongly supported in recent reviews as primary agents affecting the origin and development of native grasslands in North America (Vogl 1974, Wright and Bailey 1980). Lightning has for many years been recognized as an important cause of forest fires (Komarek 1966, 1968; Stokes and Dieterich 1980) but there is less agreement about its importance as a natural cause of grassland fires. Rowe (1969) and Vogl (1974) summarized historical perspectives of the arguments relative to the purported unimportance or absence of lightning fires in North American grasslands. However, authentic instances of lightning-caused fires have been reported in grasslands or steppes in Alberta (Nelson and England 1971), Saskatchewan (Raby 1966, Rowe 1969, Coupland 1973), Kansas (Brock 1925, Malin 1956), Nebraska (Kirsch and Kruse 1973, Wolfe 1973, Westover 1977, Bragg 1978), and Washington (Uresk et al. 1976). Komarek (1966) reported lightning-caused fires for South Dakota and Nebraska, all of which were associated with forest or forest-savanna vegetative types. Thus, lightning-caused fires are a natural phenomenon in the Central and Northern Great Plains grasslands but there is a paucity of information about their seasonality, frequency, and extent. The present paper summarizes and characterizes the events of 294 known occurrences of lightning-caused fires in grasslands in North Dakota and in pine-savanna lands of nearby South Dakota

LOCAL AND LANDSCAPE-LEVEL FACTORS INFLUENCING BLACK TERN HABITAT SUITABILITY

Wetlands throughout eastern South Dakota were surveyed (1995-97) for foraging and nesting black terns (Chlidonias niger) to evaluate local and landscape factors influencing habitat suitability We surveyed 834 randomly selected, semipermanent, and seasonal wetlands that were stratified by physiographic domain, wet- land density, and wetland surface area. A discriminant function model was used in a geographic information system (GIS) to classify habitat suitability of all semipermanent wetlands in eastern South Dakota. We calcu- lated number of suitable, protected wetlands by combining wetlands with easement and fee-title tracts in the GIS. Black terns nested in 7.8% and foraged in an additional 17.9% of semipermanent wetlands. Significant variables in the discriminant function were wetland area, total semipermanent wetland area within the wetland complex, and grassland area in the upland matrix. Black terns were an area-dependent species that occupied large (f = 18.9 ha) wetlands located within high-density wetland complexes. Black terns typically occurred in wetlands within landscapes where <50% of upland grasslands were tilled. Classification rates were high (76- 100%), indicating that the model identified unsuitable wetlands using wetland area and landscape-level attri- butes. Characteristics of entire landscapes must be considered in habitat assessments because wetlands that do not correspond to landscape-scale habitat requirements may not be suitable despite favorable local condi- tions. Lower correct classification rates (22-78%) for occupied ponds indicated that suitability also is dependent on local conditions for wetlands which correspond to habitat requirements from a landscape perspective. Suitable black tern nest sites occurred within regenerating or degenerating wetlands where vegetation structure rather than species of vegetation dictated suitability of nest substrates. Wetland acquisition programs have protected 44% of wetlands suitable for black terns. Future wetland acquisitions should maintain the integrity of entire prairie landscapes in additional to attributes of individual wetlands.

Spatiotemporal Patterns of Wetland Occurrence in the Prairie Pothole Region of Eastern South Dakota

We evaluated changes in wetland abundance, size, and classification between average (1979–1986) and above-average (1995–1999) precipitation periods for two physiographic regions in eastern South Dakota. Temporal shifts in wetland numbers, area, and class varied by topographic location. In high wetland density areas (> 8 wetlands/100 ha), our data suggests that larger, semipermanent wetlands expanded and absorbed nearby wetland basins into their margins, resulting in a net “loss” or disappearance of temporary and seasonal wetlands in above-average water condition years. “Losses” described here are not deemed permanent as in cases of draining or filling, and wetlands may re-form when water conditions return to normal. Nevertheless, temporary disappearance of smaller more isolated wetlands may have implications for breeding waterfowl and other fauna. Percent change of semipermanent basin numbers was positively correlated with wetland density, whereas the opposite was true for seasonal wetlands. Loss of temporary wetlands was correlated with wetland aggregation within the sample area. However, in low wetland density areas, the number and size of seasonal and temporary wetlands generally increased following above-average precipitation. We suggest that wetlands’ spatial arrangement be considered along with traditional wetland quantification techniques to better account for shifts in wetland habitat in dry versus wet years.

Planted Grasslands and Native Sod Prairie: Equivalent Habitat for Grassland Birds?

ABSTRACT. Little is known about how avian relationships to tracts of native sod prairie compare with avian relationships to single and multiple species of cool- and warm-season grassland plantings. We compared grassland bird species richness and density in 5 grassland cover types (n = 97) in the tallgrass prairie region of eastern South Dakota and western Minnesota, 2001–2004. Grassland bird species richness was significantly higher in native sod prairies than it was in all planted cover types except warm-season mixes. Grasslands dominated by exotic species did not support as many grassland bird species or have species densities as high as grasslands containing native species. Intermediate wheatgrass monotypes and cool-season mixes comprised of exotic species contained 40%–60% fewer grassland bird species than native sod prairie. Bobolink (Dolichonyx oryzivorus) density was 68% and 51% lower in intermediate wheatgrass monotypes and cool-season mixes, respectively, than it was in switchgrass monotypes. Clay-colored Sparrow (Spizella pallida) density was 75%–91% higher in native sod prairies than it was in any other cover type. Savannah Sparrow (Passerculus sandwichensis) density was 72% higher in native prairie than it was in grasslands dominated by exotic species. We recommend incorporating a diversity of native plant species into grassland plantings for biomass fuels or wildlife, rather than using monotypes or exotic species, to provide habitat for grassland birds. Although replacing croplands with planted grasslands would benefit grassland bird populations, we caution that replacing existing native sod tracts with planted grasslands would be detrimental to populations of several grassland bird species.

Temporal distribution of waterfowl in eastern South Dakota: Implications for aerial surveys

The U.S. Fish and Wildlife Service and Canadian Wildlife Service incorporate information from annual aerial and ground counts of waterfowl into harvest management strategies. Timing of population surveys is thought to be optimal for early nesting species but may not reliably reflect the status of species with later migration chronology. The purpose of this study was to evaluate effects of survey timing on diving duck and duck brood abundance indices in eastern South Dakota. Findings indicate that timing of aerial breeding surveys occurred too early in eastern South Dakota to accurately reflect diving duck population status because birds had not yet settled into breeding habitats and aerial production surveys did not coincide with peak duck brood abundance. Diving duck abundance from aerial surveys conducted during 10–17 May were higher than indices from ground surveys conducted 2 weeks later because lesser scaup (Aythya affinis), ring-necked duck (Aythya collaris), and bufflehead (Bucephala albeola) were still migrating through eastern South Dakota. Ground-survey estimates of redhead (Aythya americana) and canvasback (Aythya valisineria) abundance exceeded the upper limit of 95% confidence intervals for aerial survey estimates. Results from our second year of study showed an 11-fold decrease in lesser scaup abundance and an 8-fold decrease in ring-necked duck abundance in the two weeks following aerial breeding surveys. Brood abundance of dabbling and diving ducks peaked 1–1.5 months after aerial production surveys conducted in early July. Late-nesting indices calculated from aerial surveys were unrelated (P=0.21) to number of late-hatching broods, whereas number of broods hatching after aerial surveys was correlated with late-nesting indices from ground surveys (r2=0.74, P<0.01) and with number of July ponds (r2=0.77, P<0.01). We encourage scientists to initiate a more thorough evaluation of survey biases to ensure that trends accurately reflect status of duck populations and to explore the possibility of using a separate late May/early June aerial survey after diving ducks have settled into habitat as a way of assessing breeding population trends for these species.

Planting Trees in Prairie Landscapes: Do the Ecological Costs Outweigh the Benefits?

Abstract Tree plantings are striking landscape features that symbolize settlement of the West. Although grassland birds require large tracts of treeless grasslands, planting trees in prairie landscapes is still regarded by many as a positive management practice. We compared bird use of tree plantings (n=182) and natural riparian woodlands (n=37) to evaluate whether ecological costs of tree plantings to grassland birds is compensated for by the benefit they provide to forest birds of management concern. Findings indicate that ecological costs of tree plantings outweigh their benefits because they fail to provide habitat for forest birds of management concern. Thus, loss of native grassland bird communities is the currency in which we pay the ecological costs of plantings trees in prairie landscapes. We recommend that managers refrain from planting trees in or adjacent to grassland habitats. We further recommend that managers refocus their tree plantings efforts to restoration of riparian woodlands that maximize avian diversity in prairie landscapes.

Factors influencing wetland use by Canada geese

Seasonal and semi-permanent wetlands in eastern South Dakota were surveyed in 1995 and 1996 to identify habitat characteristics influencing wetland use by Canada geese (Branta canadensis maxima). Position of a wetland within the landscape and its area were important landscape-scale features influencing wetland use by geese. Our delineation of potential Canada goose habitat using a wetland geographic information system indicated that distribution and area of semi-permanent wetlands likely limit Canada goose occurrence in regions outside the Prairie Coteau. Periodicity in hydrologic cycles within landscapes also may influence goose use of wetlands in eastern South Dakota.