Agassiz National Wildlife Refuge Research Articles

Increased Bufflehead (<i>Bucephala albeola</i>) Breeding Activity in Minnesota

Buffleheads (Bucephala albeola) predominantly nest in the boreal forests and aspen parklands of Canada and Alaska. Historically, Buffleheads were common migrants but not summer residents in Minnesota. However, recent observations in Minnesota and surrounding states suggest increased breeding activity in the region. In 1978, the first Bufflehead brood in Minnesota was recorded at East Park Wildlife Management Area. Annually, Agassiz National Wildlife Refuge (NWR) has conducted pair and brood surveys, with species-specific records available from 1990 to 2014. We report an increase in numbers of Bufflehead breeding pairs and broods at Agassiz NWR and new pair observations in surrounding areas.

Understanding and Finding Solutions to the Problem of Sedimentation in the National Wildlife Refuge System

Abstract The National Wildlife Refuge System (Refuge System) is a collection of public lands maintained by the U.S. Fish and Wildlife Service for migratory birds and other wildlife. Wetlands on individual National Wildlife Refuges (Refuges) may be at risk of increased sedimentation because of land use and water management practices. Increased sedimentation can reduce wetland habitat quality by altering hydrologic function, degrading water quality, and inhibiting growth of vegetation and invertebrates. On Refuges negatively affected by increased sedimentation, managers have to address complex questions about how to best remediate and mitigate the negative effects. The best way to account for these complexities is often not clear. On other Refuges, managers may not know whether sedimentation is a problem. Decision makers in the Refuge System may need to allocate resources to studying which Refuges could be at risk. Such analyses would help them understand where to direct support for managing increased sedimentation. In this paper, we summarize a case study demonstrating the use of decision-analytic tools in the development of a sedimentation management plan for Agassiz National Wildlife Refuge, Minnesota. Using what we learned from that process, we surveyed other Refuges in U.S. Fish and Wildlife Service Region 3 (an area encompassing the states of Illinois, Indiana, Iowa, Ohio, Michigan, Minnesota, Missouri, and Wisconsin) and Region 6 (an area encompassing the states of Colorado, Kansas, Montana, Nebraska, North Dakota, South Dakota, Utah, and Wyoming) about whether they experience sediment-related impacts to management. Our results show that cases of management being negatively affected by increased sedimentation are not isolated. We suggest that the Refuge System conduct a comprehensive and systematic assessment of increased sedimentation among Refuges to understand the importance of sedimentation in context with other management problems that Refuges face. The results of such an assessment could guide how the Refuge System allocates resources to studying and managing widespread stressors.

Isolation of Type A Influenza Viruses from Red-necked Grebes (Podiceps grisegena)

Six type-A low pathogenic influenza viruses from 14 Red-necked Grebes (Podiceps grisegena) from Agassiz National Wildlife Refuge were sequenced. The grebe viruses were closely related to North American duck viruses. The genetic and temporal subtype consistency between the duck and grebe isolates suggest spillover events, potentially enhanced by feather eating.

Heavy Metals and Selenium in Grebe Feathers from Agassiz National Wildlife Refuge in Northern Minnesota

Metal levels in feathers can often be used as an indicator of exposure and of potential effects in birds. In previous work at Agassiz National Wildlife Refuge, northwestern Minnesota, pied-billed grebe (Podilymbus podiceps) eggs had significantly higher levels of manganese and mercury and significantly lower levels of selenium than eared (Podiceps nigricollis) or red-necked grebes (Podiceps grisegena), but in 1999, pied-billed grebes had significantly higher levels of mercury, but lower levels of selenium and tin than the other grebes. This led us to examine whether these patterns held up in feathers of grebes as a function of age. The feathers of young birds represent local exposure. We collected feathers of flightless young and adult grebes from 1997 to 1999 in the marshes at Agassiz National Wildlife Refuge. Regression models indicated that year, age, or species were significant factors accounting for variations in the levels of arsenic, cadmium, chromium, lead, manganese, mercury, and selenium, depending on the metal. Overall, there were significant intraspecific differences for all metals. Pied-billed grebes had the highest levels of arsenic, chromium, and selenium, and eared grebes had the highest levels of cadmium, manganese, and mercury. Pied-billed and western grebes (Aechmophorus occidentalis) had the highest levels of lead. There were significant age-related differences in cadmium, chromium, and mercury for both eared and red-necked grebes, for arsenic in eared grebes, and for lead and manganese in red-necked grebes. Adults had higher levels of all metals, except young had higher levels of chromium. Mercury in the feathers of eared grebes were higher than found from other studies with a wide range of aquatic and marine birds and were above those known to cause adverse effects in laboratory studies, suggesting some cause for concern.

Selenium and Metal Concentrations in Waterbird Eggs and Chicks at Agassiz National Wildlife Refuge, Minnesota

Exceptionally high cadmium (Cd) and chromium (Cr) concentrations were reported in eggs, feathers, or livers of selected waterbird species nesting at Agassiz National Wildlife Refuge (Agassiz) in 1994. Ten- to 15-day-old Franklin's gull (Larus pipixcan), black-crowned night-heron (Nycticorax nycticorax), and eared grebe (Podiceps nigricollis) chicks were collected in 1998, 1999, and 2001 at Agassiz and analyzed for selenium (Se) and metals including Cd and Cr. Freshly laid eggs were collected in 2001 from Franklin's gull, black-crowned night-heron, eared grebe, and pied-billed grebe (Podilymbus podiceps) nests at Agassiz. Based on a multivariate analysis, the pattern of Se and metal concentrations differed among species for eggs, chick feathers, and chick livers. Low Cd and Cr concentrations were measured in eggs, chick livers, and chick feathers of all four species. Mercury concentrations in black-crowned night-heron and eared grebe eggs collected from Agassiz in 2001 were lower than concentrations reported in 1994. Se and metal concentrations, including Cd and Cr, in waterbird eggs and chicks collected at Agassiz in 1998, 1999, and 2001 were not at toxic levels.

Black Tern Nest Habitat Selection and Factors Affecting Nest Success in Northwestern Minnesota

We documented nest habitat selection, nests success, and factors affecting nest success of Black Terns (Chlidonias niger) at Agassiz National Wildlife Refuge in northwestern Minnesota. During 1992-1994, 289 Black Tern nest sites and 400 random sites were sampled on five search areas totaling 1,325 ha. Four habitat characteristics were measured at each nest and random site: (1) mean water depth, (2) distance to open water, (3) dominant vegetation within a 2-m radius, and (4) amount of open water within a 2-m radius. Habitat variables were highly correlated with each other, making it difficult to estimate independent effects of each habitat variable on nest-site selection. However, conditional logistic regression models indicated that locations closer to open water and in deeper water were more likely to be associated with nest sites. Locations in bulrush (Scirpus acutus) and sedge/grass were preferred, although 68% of nests were in cattail (Typha spp.) reflecting the greater availability of that habitat in the study area. Nest success ranged from 48-69% (Apparent) and 33-62% (Mayfield) among years. Except for five nests that were abandoned or had infertile eggs, nests that failed to hatch apparently were depredated. Nest success was higher for nests with larger clutch sizes and for nests located farther away from other nests. Nests with 3-egg clutches were 2.8 times as likely to hatch as 2-egg nests. The odds of a nest being successful increased by 25% for each 5 m increase in distance to the nearest nest. Earlier nests were also more successful (the odds of a nest being successful was estimated to decrease 7% for each additional day that passed before the nest was initiated). Nest success was not related to nest cluster size and was negatively related to habitat factors associated with nest site-selection.

Trace Element Concentrations and Bioindicator Responses in Tree Swallows from Northwestern Minnesota

Extremely high concentrations of cadmium (3.5 microg/g dry wgt.) and elevated concentrations of chromium (>10 microg/g dry wgt.) and mercury (1.6 microg/g dry wgt.) were reported in waterbird tissues at Agassiz National Wildlife Refuge in northwestern Minnesota in 1994. Tree swallows (Tachycineta bicolor) were studied during 1998-2001 at three drainages into the Refuge, two pools on the Refuge, and at a nearby reference location to document whether high levels of contaminants were still present, and if so to quantify the source and severity of the contamination. Trace elements were measured in tree swallow eggs, livers, and diet. Reproductive success and bioindicator responses were monitored. In 2000, water was drawn down on Agassiz Pool, one of the main pools on the Refuge. This presented an opportunity to evaluate the response of trace element concentrations in the diet and tissues of tree swallows after reflooding. High concentrations of trace elements were not detected in swallow tissues, nor were there differences among locations. Less than 20% of swallow samples had detectable concentrations of cadmium or chromium. Mercury concentrations were low and averaged <0.25 microg/g dry wgt. in swallow tissues. Trace elements, including mercury, did not increase in tree swallows following the 2000 drawdown at Agassiz Pool. Hatching success and survival of nestlings to 12 days-of-age for tree swallows on the Refuge were similar to the national average and consistent with background trace element concentrations. Bioindicator measurements were within the normal ranges as well.

HEAVY METALS AND SELENIUM IN GREBE EGGS FROM AGASSIZ NATIONAL WILDLIFE REFUGE IN NORTHERN MINNESOTA

Metal levels in eggs can often be used as an indicator of exposure and of potential effects. In previous work at Agassiz National Wildlife Refuge, northwestern Minnesota, the levels of several heavy metals were shown to be significantly higher in the eggs of eared grebes (Podiceps nigricollis) compared to those in the eggs of Franklin's gulls (Larus pipixcan), black-crowned night-herons (Nycticorax nycticorax) and double-crested cormorants (Phalacrocorax auritus, except for mercury). In the present study we test the hypothesis that there are no differences in the levels of heavy metals in eggs of three species of grebes nesting at Agassiz National Wildlife Refuge (1997, 1999). There were significant differences in levels of selenium, manganese and mercury in the eggs of the grebes collected in 1997, with pied-billed grebe (Podilymbus podiceps) having significantly higher levels of manganese and mercury, and significantly lower levels of selenium, than eared or red-necked grebes (Podiceps grisegena). In 1999, pied-billed grebes had significantly higher levels of mercury, but lower levels of selenium and tin than the other species. The only pattern that was significant and consistent among years was selenium; in both years pied-billed grebes had lower levels than the other species. For eared grebes, there was a decline from 1997 to 1998, and again to 1999 for arsenic, cadmium, and selenium. Levels of mercury in the eggs of grebes were not as high, however, as those found in cormorants or night-herons sampled in 1994 at Agassiz National Wildlife Refuge. There were few consistent patterns in the relationships among metals in eared grebe eggs (with the largest sample sizes). The possible reasons for the high levels of some metals in eggs of grebes are unknown, but presumably egg levels represent exposure on the wintering grounds or migratory routes. In comparison to eggs of other birds: 1) the mean levels for manganese were at the high end of the range, and the mean was an order of magnitude higher than the median for the studies examined, 2) mean levels were above the median in the eggs of other birds for lead (red-necked grebe), mercury (pied-billed grebe) and selenium (eared and red-necked grebe).

First Described Renesting Attempt by an American Bittern

Most life history traits of the American Bittern (Botaurus lentiginosus) have not been studied and are poorly understood. The ability of the American Bittern to renest has not been confirmed previously. A second nesting attempt by an American Bittern was observed on Agassiz National Wildlife Refuge on 8 July 1996. This information provides insight into American Bittern fecundity by showing that additional reproductive capability exists when nests are destroyed by predation or weather related events. Future studies of nesting bitterns will need to consider renesting when estimating density of nesting females.

Direct Recovery Rates of Lesser Scaup Banded in Northwest Minnesota: Sources of Heterogeneity

Heterogeneity in direct recovery rates is an important area of study because deductions about survival rates and hunting mortality are made from recovery data. We tested the hypothesis of condition bias in hunter-killed lesser scaup (Aythija affinis) and tested differences among direct recovery rates (f= proportion of birds banded that are killed by a hunter ≤3 months after release and reported) of scaup staging on Thief Lake Wildlife Management Area (TLWMA) and Agassiz National Wildlife Refuge (ANWR) in northwestern Minnesota Scaup were trapped, held for 24 hr, weighed, banded with standard aluminum leg bands, and released during 4 October-8 November 1983-86 at TLWMA, and 1984-86 at ANWR. During 1984-86 e also sprayed areas of white plumage with fluorescent paint, colorcoded by age and location. Recovery data obtained from cooperative hunters and the U.S. Geological Survey Bird Banding Laboratory were subjectec to 3 sets of analyses. In an analysis of variance (ANOVA), after adjusting for sex, age, year, location of capture, and date of capture body mass at capture of recovered birds was lower, on average, than those not recovered Logistic regression mdicated (1) probability of recovering a bird decreased with increasing body mass; (2) within years, probability of recovery increased with date of first capture; (3) scaup at TLWMA were recovered at greater rates than scaup at ANWR: (4) hatching-year (HY) scaup were recovered at greate r rates than were after-hatcbing-year (AHY) birds; and (5) recovery was more likely during 1984-86 than in 1983. Finally, in a crude test of recovery distribution, recoveries of birds banded at ANWR were about half as likely as those banded on TLWMA to occur in Minnesota. We conclude that cohort membership (age, capture site) and condition, as measured by body mass, influence susceptibility to hunting mortalitv. These and other sources of heterogeneity in recovery rates should be considered when modeling survival of scaup.

Heavy metal and selenium levels in birds at Agassiz National Wildlife Refuge, Minnesota: Food chain differences

The levels of heavy metals and selenium in the eggs and in breast feathers of adult doublecrested cormorant (Phalacrocorax auritus), black-crowned night heron (Nycticorax nycticorax), and franklin's gull (Larus pipixcan) nesting at Agassiz National Wildlife Refuge in Marshall County, northwestern Minnesota were examined. Also examined were metal levels in the feathers of fledgling night herons and gulls, in the feathers of adult and fledgling American bittern (Botaurus lentiginosus), in eggs of American coot (Fulica americana) and eared grebe (Podiceps caspicus), and in feathers of adult Canada geese (Branta canadensis). These species represent different levels on the food chain from primarily vegetation-eating species (geese, coot) to species that eat primarily fish (cormorant). A clear, positive relationship between level on the food chain and levels of heavy metals occurred only for mercury in feathers and eggs. Otherwise, eared grebes had the highest levels of all other metals in their eggs compared to the other species. No clear food chain pattern existed for feathers for the other metals. For eggs at Agassiz: 1) lead, selenium, and manganese levels were similar to those reported in the literature, 2) mercury levels were slightly higher for cormorants and night herons, 3) all species had higher chromium and cadmium levels than generally reported, and 4) eared grebes had significantly higher levels of cadmium than reported for any species from elsewhere. For adult feathers: 1) gulls had higher levels of lead than the other species, 2) cadmium levels were elevated in gulls and adult herons and cormorants, 3) mercury levels showed an increase with position on the food chain, 4) selenium and chromium levels of all birds at Agassiz were generally low and 5) manganese levels in adults were generally higher than in the literature for other species. Adults had significantly higher mercury levels than fledgling gulls, night herons, and bitterns.

Survival of Radiomarked Canvasback Ducklings in Northwestern Minnesota

Duckling survival, an important factor affecting annual recruitment, has not been determined adequately for canvasbacks (Aythya valisineria). We investigated the magnitude, timing, and causes of mortality of canvasback ducklings from hatch to fledging at the Agassiz National Wildlife Refuge (NWR) in northwestern Minnesota during 1987-90. During the 4 years, 217 day-old ducklings were radiomarked and released in 52 broods. Another 141 ducklings were radiomarked at ≥ 4 weeks of age. Survival was estimated with the Kaplan-Meier nonparametric estimator and the Weibull parametric model. Most mortalities occurred within 10 days after hatch. Total brood loss occurred in 18 (35%) of 52 broods released. The primary sources of mortality were predation, principally by mink (Mustela vison), and exposure to precipitation and cold temperature. For combined years, females had lower survival than males (P = 0.03). If the disparate survival between sexes of canvasbacks observed in this study is representative of canvasbacks in their breeding range, this phenomenon contributes to reduced reproductive potential and the male-biased sex ratio of the species.

Determining Sex Ratios from Collected Specimens

Ornithologists have recently shown considerable interest in determining the sex ratios of birds, particularly on the breeding grounds. Field counts are possible only for sexually dimorphic species. Investigators working on monomorphic species have therefore turned to the data accompanying museum specimens. The method of collecting these specimens is usually unknown, and may bias the sex ratio, as I will show. In 1971 I collected a number of Franklin's Gulls (Larus pipixcan) in the vicinity of Agassiz National Wildlife Refuge, Marshall County, Minnesota. The gulls were collected during May and June at or within 8 km of a breeding colony of 10,000-15,000 pairs of gulls (see Burger, Anim. Behav. 22:521-567, 1974). Birds were collected by three methods; nest-trapping (Burger, Bird-Banding 42:123124, 1971), shooting, and netting birds that were feeding behind a plow. I used all three methods only to increase my sample of specimens. I nest-trapped between 09:00 and 11:00 throughout the incubation period. Shooting and netting were done on the same days and in the same two fields where 200-500 gulls were feeding. A local farmer

Chemical Inputs by a Colony of Franklin's Gulls Nesting in Cattails

Franklin's gulls nest at the Agassiz National Wildlife Refuge, northern Minnesota, after migrating from South America. Nutrient inputs by the gulls during 1971 were examined by comparison of water and sediment samples from two shallow pools. On pool was colonized by about 30,000 gulls nesting among cattails; the second pool contained cattails without a gull colony. Large increases occurred in nitrogen and phosphorus concentrations of water in the immediate vicinity of the gull colony; peak concentrations were concomitant with the peak of gull nest-building and feeding activities (mid-May and mid-July). About 36% of the annual phosphorus input to the pool with the gull colony is attributable to the gulls. However, there were no net changes in concentrations of nitrogen and phosphorus in water of either pool as a whole, i.e., nutrient concentrations in inflow and outlet waters of both pools remained relatively constant throughout the year due to nutrient absorption by the pool sediments. The gull colony had little effect on the concentrations of sodium, potassium, calcium and magnesium in water. Outlet waters of both pools had higher potassium concentrations than inputs, due to release of potassium by the sediments to water, or by nutrient uptake by cattails and subsequent leaching of both living and dead cattails. Most nutrients accumulated in the sediments, especially in the pool with the gull colony. The Rufuge acts as a sink for nutrients in runoff from surrounding ag,ricultural land.

Breeding adaptations of Franklin's gull ( larus pipixcan) to a marsh habitat

The behaviour and ecology of Franklin's gull were studied at Agassiz National Wildlife Refuge in northwestern Minnesota to determine the adaptations of the species for nesting in marshes. Two factors seemed to be important in colony site selection: cattail dispersion pattern and cattail density. Franklin's gulls prefer to nest in cattail areas closest to open water. The number of nests per unit area decreased as cattail density increased. Nest site selection is dependent on aggression and visibility. Visibility from nest level is the result of cattail placement and height. The distance between nests was directly correlated with visibility. Aggression by gulls on nests was lowered experimentally by decreasing visibility and raised by increasing visibility. Nest platforms were constructed of cattail material, and were attached to cattail stems. Nest material was added to the nests throughout the incubation and brooding period. Material was usually added following nest relief. The egg laying period was from 6 to 28 May. There was more synchrony of egg laying in sub-areas of the colony than in the colony as a whole. Successive eggs in clutches were laid at 24- to 48-hr intervals. The distance between nests decreased during the season as pairs filled in areas that were not defended. Territorial pairs defended an area up to 10 m from their stations prior to egg laying, but defended only the area within 3 m of their nests during incubation. Both members of pairs incubated the eggs and cared for the young. The incubation period was 24 days. The primary predators on adults and young were marsh hawk, great horned owl and mink. Franklin's gulls do not eat eggs or young of gulls. Adults fed on earthworms, insects and grain. Most marked adults fed within 16 km of the colony. Chicks were fed primarily on earthworms. The hatching period was from 30 May to 21 June. Chicks of all ages tested on a visual cliff apparatus were able to perceive the drop. Chicks tested on a 30-degree incline apparatus walked up it when 6 days old and younger, and walked down at 12 days of age and older. Brood mobility was less than in ground nesting species of gulls. In an undisturbed colony the chicks remained on the nest platforms until they were 25 to 30 days old although they were capable of swimming shortly after hatching. Individual recognition between parents and chicks appeared later in this species than in ground-nesting gulls. Adults accepted alien chicks (experimentally exchanged) that were younger than about 14 days old until their own chicks were over that age. Adults accepted larger and older broods than their own, as well as broods of mixed ages. Chicks began to react differently to strange adults at about 16 days of age. The breeding chronology of Franklin's gull is compressed when compared to that of other gulls. Possible selection pressures affecting this synchrony are discussed. The behaviour of the marsh-nesting Franklin's gull is compared with that of typical ground-and cliffnesting gulls; the possibility that the ancestral gull may have been a marsh nester is discussed.

Moose Movement Patterns and Range Use in Northwestern Minnesota

During 1968-71, 3,881 observations were recorded on 36 marked moose (Alces alces) (21 bulls and 15 cows) on Agassiz National Wildlife Refuge in northwestern Minnesota. These data were used to calculate home ranges, habitat use patterns, and rates of movement. Home ranges during the summer and fall varied from 1.0 to 15.1 square miles, averaging 6.9 square miles for cows and 5.6 for bulls. Winter home ranges varied from 0.3 to 2.9 square miles, averaging 1.4 square miles for cows and 1.2 for bulls. About 20 percent of the Agassiz herd was found to be migratory; migration distances ranged from 9 to 21 miles. Rates of travel were greatest during the summer months (average, 0.6 mile/24-hour period). Bulls traveled widely during the rutting period while cows remained sedentary. Movements were generally greater at night than during the day. Moose preferred to use the more open willow (Salix spp.) type during the summer and fall months, but shifted to dense aspen (Populus tremuloides)--willow stands in late winter. Dispersal from wintering areas was associated with the disappearance of snow cover.