---

---

Table of Contents

• Introduction
• Study Area
• Methods
• Results
• Discussion
• Management Implications
• Literature Cited

Tables and Figures

• Table 1 -- White-fronted goose use of habitat
• Table 2 -- Diurnal activity of white-fronted geese
• Table 3 -- Results of regression analysis
• Table 4 -- Area of corn harvested for Clay County

• Figure 1 -- Changes in carcass fat
• Figure 2 -- Area of corn harvested in Nebraska

---

Introduction

The midcontinent population of greater white-fronted geese breed across the Arctic from the North Slope of Alaska to Hudson Bay and winter from coastal Louisiana westward through Texas and into Mexico (Bellrose 1980). Although dispersed during most of the year, a major part of the midcontinent population stages during March at a limited number of sites in the RBA and adjacent Platte River Valley (PRV) of south-central Nebraska. Wetlands in the RBA have been the site of disease (primarily avian cholera [Pasteurella multocida]) outbreaks among waterfowl since 1975 (Friend 1981; J. D. Kauffeld, U.S. Fish and Wildl. Serv., Kearney, Nebr., unpubl. data). Concern prompted by the observation that agricultural development has destroyed >90% of the original wetlands in the RBA (Tiner 1984) and a major disease outbreak in March 1975 that killed an estimated 25,000 waterfowl of which 35% were white-fronted geese (Zinkl et al. 1977), along with a general lack of information on the importance of RBA/PRV staging areas to migration and reproduction, led to this study. Our objectives in studying the spring-staging ecology of white-fronted geese in south-central Nebraska were to (1) determine agricultural habitats used and foods consumed, (2) assess diurnal activity budgets on agricultural lands and wetlands used as roosts, (3) evaluate importance of the RBA/PRV for nutrient storage, and (4) relate corn availability to energy requirements for maintenance and fat deposition.

We thank R. J. Atkins, C. R. Frith, B. A. Hanson, G. R. Lingle, and F. C. Rowher for assisting in data collection and C. Jorgenson, Canadian Wildlife Service, Last Mountain Lake, Saskatchewan, for providing facilities during fieldwork in Canada. We thank W. C. Dobbs and S. E. Keller, Nebraska Agricultural Statistics Service, Lincoln, for providing data on corn production in Nebraska during 1900-93; I. D. Godtel, Clay Center, Nebraska, and V. L. Hofman, Department of Agricultural Engineering, North Dakota State University, Fargo, for information on the history of corn harvesting methods in the Great Plains; and J. D. Kauffeld, Rainwater Basin Wetland Management District, Kearney, Nebraska, for information on goose mortality from disease outbreaks in the RBA during 1975-88. We thank D. H. Johnson for assistance in study design; J. A. Beiser, D. W. Howerter, R. C. Khan-Malek, W. E. Newton, and T. L. Shaffer for statistical advice and analyses, K. M. Lahlum, D. A. Brandt, and J. W. Solberg for obtaining technical reports and other assistance. We appreciate comments provided by J. E. Austin, R. J. Greenwood, D. W. Howerter, D. L. Larson, L. L. Strong, and 2 anonymous reviewers.

Study Area

The RBA encompassed approximately 7,000 km² adjacent to, and south of, the PRV in south-central Nebraska. An estimated 3,907 wetlands (38,338 ha) existed in the RBA in 1900, but by 1983 only 374 (8,479 ha) remained (Gersib et al. 1990). Each spring, remaining wetlands of the RBA and PRV attracted several million migrant waterfowl including a major segment of the midcontinent white-fronted goose population. Within the RBA/PRV, wetlands in Clay and Kearney counties received the most use as spring staging areas by white-fronted geese (D. S. Benning, U.S. Fish and Wildl. Serv., Denver, Colo., unpubl. data). As part of our studies, we determined use of habitat in the vicinity of Harvard Marsh and Prairie Dog Marsh, major white-fronted goose staging areas in Clay and Kearney counties. During March 1980, Harvard and Prairie Dog marshes had estimated peak white-fronted goose populations of 23,000 and 11,000 birds, respectively (Krapu, unpubl. data), or about 14% of the 250,000 white-fronted geese present in south-central Nebraska in 1980 (D. S. Benning, unpubl. data).

Our study areas encompassed fields used for feeding by geese that roosted on the marshes during spring 1980. The existing road system enabled us to survey all fields within the study areas. Harvard and Prairie Dog were owned by the U.S. Fish and Wildlife Service and managed as Waterfowl Production Areas; surrounding agricultural lands were privately owned and most were farmed annually with the principal crops being corn, winter wheat, and milo. Agricultural habitats within the study areas by type (in ha) were pasture/hayland (381 [Harvard], 377 [Prairie Dog]), growing winter wheat (668, 884), harvested winter wheat (142, 399), harvested corn (4,436, 5,134), harvested milo (918, 569), harvested soybeans (132, 95), and canal/dugouts (15, 461). We further classified fields within the harvested corn category (in ha) as idle (855 [Harvard], 1,427 [Prairie Dog]), cultivated (3,072, 1,823), grazed (485, 1,883), and moldboard plowed (17, 0). Idle cornfields were not cultivated or grazed between harvest and spring staging. Grazed cornfields were where livestock fed on corn residues.

Methods

Data Collection

Habitat Availability and Use.—To document habitat availability, we drew fields within Harvard and Prairie Dog study areas on 1:24,000-scale topographic maps and classified them to habitat type in early March 1980. We measured field areas with an electronic digitizer and summed them by habitat.

We conducted 10 and 9 surveys of use of habitat on the Harvard and Prairie Dog study areas, respectively, between 12 March and 5 April. We randomly assigned the time interval (0700-1000, 1200-1500, or 1600-1900) during which we made counts. At Harvard, we conducted 4 surveys during 0700-1000 and 3 each during 1200-1500 and 1600-1900. At Prairie Dog, we conducted 3 surveys during 0700-1000, 2 surveys during 1200-1500, and 4 surveys during 1600-1900. We checked all agricultural lands within each study area for white-fronted geese during each survey and recorded the number of geese and field type. We made goose counts with a spotting scope mounted on a vehicle window.

Foods.—We determined foods consumed by geese on agricultural lands from the esophageal contents of 42 geese collected during 22 February-3 April 1979 and 29 February-25 March 1980 while feeding in fields (n = 10) or returning to roosting marshes (n = 32) in Clay, Kearney, and Hall counties. We removed food items in a bird's esophagus in the field and froze them. Later, we sorted, identified, and dried food samples for 48 hours at 55 C, and weighed them to the nearest 0.01 g. We expressed food importance as frequency of occurrence and as aggregate percentage of total dry mass (Swanson et al. 1974).

Time Budgets.—We estimated diurnal activity budgets of white-fronted geese during 9-21 March 1979 and 14-25 March 1980 for agricultural lands and wetlands, respectively, in the Harvard study area. We sampled individuals by focusing a spotting scope on a flock of geese, moving the scope horizontally or vertically, and selecting for observation the individual nearest to the center of the field of view. We recorded activities of individuals at 12-second intervals (Wiens et al. 1970) for ≤ 30 minutes. Temporal distribution (in min) of time budget observations was 0700-0859 (n = 98), 0900-1059 (113), 1100-1259 (81), 1300-1459 (154), 1500-1659 (119), 1700-1859 (176), and 1900-2059 (112). Behaviors recorded were resting, feeding, alert, agonistic interactions, and walking. Walking usually was associated with searching for food. We monitored 76 individuals (55 in 1979 and 21 in 1980) during 14.2 hours of observation. Mean length of an observation was 11.2 minutes.

Carcass Composition.—We collected 52 geese (1/flock) for carcass nutrient analyses in Clay, Kearney, and Hall counties from 22 February to 3 April 1979 and 25 February to 25 March 1980. We collected 6 geese in the Last Mountain Lake area of central Saskatchewan (51°26'N, 105°26'W) from 28 April to 5 May 1979 and 7 on the Colville River Delta on the Alaskan North Slope (70°24'N, 150°30'W) from 29 May to 19 June 1983. Last Mountain Lake was at the northern edge of the Great Plains and presumably was a departure point for geese leaving for the breeding grounds. The Colville River Delta was a breeding area for the midcontinent population (King 1970).

We measured body mass of collected specimens to the nearest gram, and total culmen length, flattened wing chord, and tarsus length to the nearest mm with calipers. We determined age from plumage characteristics and sex by gonadal examination. We classified individuals as adults if they had dark brown or black blotches and irregular bars on their bellies and white feathers at the base of the bill, and as immatures if they had pale bellies and few or no white feathers at the base of the bill (Bellrose 1980).

To prepare for analyses of carcass fat, protein, and calcium, we plucked each specimen, removed gizzard contents, weighed the featherless carcass, and homogenized the remainder in a commercial meat grinder. All chemical analyses were on subsamples of carcass homogenates using Official Methods of Analysis (Horwitz 1975). We determined fat content in Soxhlet extractors using petroleum ether, with duplicate analyses for each specimen. We determined protein using the Kjeldahl method, and calcium by atomic absorption spectrophotometry from residue remaining after we burned samples in a muffle furnace at 550 C to constant mass.

Corn Requirements.—We used a simple model of individual energy requirements to determine the amount of corn needed by staging geese to support observed rates of fat synthesis and tolerate ambient temperatures recorded at Hastings, Nebraska, during March 1979 ( = 3.7 C) and 1980 ( = 1.6 C) (Natl. Oceanic and Atmos. Adm. 1979, 1980). In the model, daily energy requirements were the sum of maintenance energy plus productive energy for fat deposition. We estimated maintenance energy as the sum of (1) existence energy calculated from equation 5.31 of Kendeigh et al. (1977:143), (2) a 10% increment to adjust existence energy for free-living conditions (Kendeigh et al. 1977:178), and (3) energy for thermal regulation calculated from equation 5.19 in Kendeigh et al. (1977:141). We estimated productive energy as a product of (1) caloric density of fat (9.5 kcal/g [Kendeigh et al. 1977:151]), (2) estimated daily rate of fat deposition (this study), and (3) energetic efficiency of fat deposition (assumed to be 75% [Walsberg 1983]).

To determine corn intake necessary to satisfy estimated energy requirements, we assumed the diet was corn (this study) and that corn provided 3.97 kcal metabolizable energy/g dry matter intake (Joyner et al. 1987). We determined availability of corn by multiplying area of corn fields mapped on the Harvard and Prairie Dog study areas times the density of waste corn from concurrent studies in the PRV (Reinke and Krapu 1986) and from publications describing effects of post-harvest treatments on corn densities (Baldassarre et al. 1983, Warner et al. 1985).

Data Analyses

Time Budgets.—We used 1-way multivariate analysis of variance (MANOVA) (Johnson and Wichern 1988) to test for variation in time budgets among habitats because estimated percentages of walking, resting, feeding, alert, and agonistic behaviors were potentially correlated. We employed multivariate pair-wise contrasts to compare time budgets between habitats, and univariate pair-wise contrasts to compare habitats within each of the 5 behavior classes. We used an arcsine-square root transformation (Martin and Bateson 1986) to stabilize variances of percentages. We weighted estimates by the square root of the number of observations, because lengths of observations varied (Neter et al. 1990).

Carcass Composition.—Because body mass, fat, protein, and calcium content potentially are related to overall body size, we attempted to remove the effect of structural size before testing for temporal trends in these variables. To do this, we used a combination of principal component and regression analyses similar to Ankney and Alisauskas (1991). First, we used PROC PRINCOMP (SAS Inst. Inc. 1989) to extract a principal component (PC1) describing variation in structural size from the correlation matrix of culmen, tarsus, and flattened wing chord measurements. Then, using PC1 as a measure of overall size, we performed regression analyses to test for linear relationships between size and body mass, protein, calcium, and fat.

When there was a regression relationship (P ≤ 0.05) between structural size and a dependent variable for specimens from any age and sex class, we transformed measurements of this dependent variable for all sex and age classes by expressing them as residuals from the regressions (Ankney and Alisauskas 1991). The form of the transformation was y_i = y_obs - (a + b[PC1]) + _obs. Then, we used linear regression of adjusted (body mass, protein, and calcium) and unadjusted (fat) carcass variables on Julian date to test for temporal trends (i.e., daily rates of nutrient deposition). We tested differences in nutrient deposition rates between sexes with analysis of covariance (ANCOVA). We performed regression analyses with PROC GLM (SAS Inst. Inc. 1989), and we pooled data from 1979 and 1980.

Results

Use of Habitat and Foods

White-fronted geese roosted at night in Harvard and Prairie Dog marshes and flew at dawn to surrounding agricultural lands to feed. During midmorning, geese rested in fields or flew back to the wetlands. Geese that returned to the marshes during midmorning generally flew back to the fields in mid- to late afternoon and stayed until dusk. Geese often remained in fields throughout the day when standing water was present after rains.

White-fronted geese were located in harvested cornfields and growing winter wheat fields 76 and 23%, respectively; harvested winter wheat, milo, and soybeans accounted for the remaining 1% of use (Table 1). Within the Harvard and Prairie Dog study areas, geese were located in a mean of 9.4 ± 2.4 and 2.0 ± 0.5 cornfields and 2.9 ± 1.4 and 1.8 ± 0.5 growing wheat fields, respectively, per survey. When data were pooled over study areas, 65, 31, and 3% of goose use of cornfields occurred in cultivated, idle, and grazed sites, respectively.

Corn grain and winter wheat shoots accounted for nearly all foods ingested by geese while foraging on agricultural lands. Corn formed the following percentages of the diet by age and sex (by frequency of occurrence and aggregate % dry mass, respectively): adult males (n = 13), 77 and 77%; immature males (n = 11), 100 and 99%; adult females (n = 11), 91 and 91%; and immature females (n = 7), 100 and 99%. Shoots of winter wheat taken from fields planted during the previous autumn occurred in the following percentages: adult males, 23 and 23%; immature males, 9 and 1%; adult females, 9 and 5%; and immature females, 0 and 0%. Overall, corn grain and winter wheat shoots formed 90 and 9%, respectively, of the diet by aggregate percent dry mass.

Activity Budgets

Activity budgets were not the same among habitats (MANOVA, Wilks' lambda statistic, λ = 0.1882; F = 4.53; 30, 262 df; P < 0.001). Geese rested most in moldboard-plowed cornfields and growing winter wheat fields and least in grazed and chisel-plowed cornfields (Table 2). Feeding activity did not vary among cornfield treatments except that little feeding occurred in moldboard-plowed fields. Geese were less alert in habitats used least for foraging (i.e., wetlands [<1.0%], moldboard-plowed fields [1.9%], and winter wheat fields [1.1%]). Agonistic interactions occurred infrequently (Table 2). The amount of walking was similar among cornfield treatments except in moldboard-plowed fields where resting predominated.

Body Composition

We collected 13 adult females, 19 adult males, 8 immature females, and 12 immature males from 22 February through 3 April 1979-80 for carcass composition analyses. We combined ages and sexes for the principal components analysis of body size. The PC1, which had loadings ranging from 0.552 to 0.590 and an eigenvalue of 2.09, accounted for 69.6% of the variation in the morphometric data and was interpreted as a general measure of body size.

Body mass was related to body size (PC1) only among immature males (t = 4.49, 10 df, P < 0.001). For adult females, protein (t = 4.61, 11 df, P < 0.001) and calcium (t = 2.31, 11 df, P = 0.042) were related to body size. Protein also was related to body size for immature males (t = 7.95, 10 df, P < 0.001) and immature females (t = 2.11, 6 df, P = 0.079). None of the regressions of fat on body size was significant (P > 0. 10). Therefore, we used measures of body mass, protein, and calcium adjusted for body size and unadjusted fat when testing for temporal trends.

Body mass was positively related to Julian date for all sex and age categories (Table 3). There was a temporal trend in carcass protein only for adult females and no trend in body calcium for any sex or age group (Table 3). In contrast, body fat was related to Julian date for all sex and age categories. Fat deposition rates (regression slopes) ranged from 8.8 g/day for immature males to 17.7 g/day for adult females, whereas rates of mass gain ranged from 11.4 g/day for immature females to 22.2 g/day for adult females (Table 3). Slopes and elevations of regressions of carcass fat on Julian date were similar between sexes for adults (ANCOVA, slope, F = 1.97; 1, 28 df; P = 0.17; elevation, F = 1.37; 1, 29 df; P = 0.25) and immatures (slope, F = 1.09; 1, 16 df; P = 0.31; elevation, F = 0.87;1, 17 df; P = 0.36). Resulting age-specific estimates for fat gain were 14.2 ± 1.9 g/day for adults (Fig. 1a) and 10.6 ± 1.9 g/day for immatures (Fig. 1b).

Predicted mean fat levels of adults staging in the RBA increased from 221 g on 22 February to 803 g on 3 April. Further increases in fat deposition were evident in the small sample of geese collected in Saskatchewan during 28 April-5 May 1979. Mean fat levels were 1,005 ± 61g for 4 adult males and 1,169 ± 105 g for 2 adult females, or 1,060 ± 59 g for the 6 adults.

Corn Requirements

The model of energy requirements predicted an adult white-fronted goose weighing 2,750 g and depositing 14.2 g/day of fat would need 104.2 g/day of corn to satisfy 90% of its 459.8 kcal/day energy requirement at the mean March 1980 temperature of 1.6 C. Predicted energy requirements were not sensitive to temperature variation, because the difference of 2.1 C in mean monthly temperature between March 1979 and 1980 only increased energy requirements and corn intake by 1.5%. Assuming geese were already maximizing food intake, this difference in temperatures would have decreased fat deposition by only 0.5 g/day. We calculated cumulative corn requirements of the white-fronted goose population in the Harvard study area as 69,000 kg by multiplying the daily corn intake (about 100 g) by the number of birds present (23,000) by the estimated length of the staging period (30 days). Estimated corn requirements for the 11,000 white-fronted geese in the Prairie Dog study area were 33,000 kg.

Corn Availability

Spring 1980.—White-fronted geese obtained corn primarily from idle and cultivated fields on the Harvard and Prairie Dog study areas. Corn densities on recently harvested fields in the adjacent PRV (Reinecke and Krapu 1986: 75) and elsewhere (Baldassarre et al. 1983:26, Warner et al. 1985:187) generally exceed 300 kg/ha, and estimated corn densities for cultivated fields were about 50 kg/ha (Baldassarre et al. 1983:28, Warner et al. 1985:187). Thus, corn availability in the Harvard study area likely exceeded requirements of geese (69,000 kg) on idle fields (855 ha at 300 kg/ha = 256,500 kg) as well as cultivated fields (3,072 ha at 50 kg/ha = 153,600 kg). In the Prairie Dog study area, the estimated corn requirement of 33,000 kg also was exceeded on idle fields (1,427 ha at 300 kg/ha = 428,100 kg) and cultivated fields (1,823 ha at 50 kg/ha = 91,150 kg).

Historical Trends.—Corn yields and production in Nebraska were relatively stable from 1900 through the mid-1950s (Fig. 2), averaging 1.6 ± 0.1 metric tons·ha^-1·year^-1 (linear regression; t = -0.54, 54 df, P = 0.59). In the late 1950s, corn yield began an increase that continued into the 1990s. From 1956 to 1979, corn yield increased 5-fold, due to growing use of irrigation (Fig. 2), greater input of fertilizers, and the development of hybrids capable of producing higher yields (Quick and Buchele 1978).

Prior to the mid-1940s, corn was picked primarily by hand and most harvest losses were unhusked ears that probably were difficult for waterfowl to exploit (Table 4). After mechanical cornpickers came into widespread use in the late 1940s, husked ears and shelled kernels dominated harvest losses (Johnson and Lamp 1966) and provided a more accessible food source. Corn residues increased during the 1960s as yields increased and the percentage of harvest losses remained relatively constant. The years after 1970 represent a period of abundant corn residues for migrating waterfowl.

Discussion

Use and Availability of Food

The proportion of time geese fed in harvested cornfields and results of esophageal analyses indicated that corn grain supplied most of the energy for maintenance and lipid synthesis. Because of its high digestibility and gross energy content (Joyner et al. 1987), corn has become a major food of geese on many wintering grounds and staging areas in North America (Alisauskas and Ankney 1992a, Krapu and Reinecke 1992). Patterns of use of habitat and data from food samples also indicated that winter wheat shoots were the primary source of dietary protein.

Use of the RBA/PRV by midcontinent white-fronted geese undoubtedly is related to availability of corn in close proximity to suitable roosting habitat. The limited time geese spent feeding and high daily rate and total magnitude of fat gain during March indicate favorable foraging conditions. Although corn residues available to geese in spring were reduced by tillage during the previous autumn, corn exceeded requirements of goose populations in both study areas. Moreover, these staging areas are capable of satisfying the energy requirements of greater numbers of geese than were present during the study, because the 5-km radius within which geese fed represents only a fraction of the area potentially available.

Corn probably became the principal food of spring-staging white-fronted geese in Nebraska when mechanized harvesters came into widespread use in the 1940s. Shelled corn became abundant in fields after yields began to increase rapidly in the 1960s. By the 1970s, corn residues were sufficient to feed large flocks of geese using Harvard and Prairie Dog marshes, and other RBA wetlands. Efficiency of harvesting corn may have increased from 1980 to 1990 but any decrease in the percentage of harvest losses was offset by increased corn production. Similarly, increases in the area of corn grown in the north-central United States (U.S. Dep. Agric. 1958-93) and in Manitoba (Manit. Agric. 1994) that began in the late 1970s probably has resulted in increased corn consumption in the northern Great Plains. Corn was the principal food of lesser snow geese (Chen caerulescens) during spring migration in southern Manitoba in 1983-84 (Alisauskas and Ankney 1992a).

Nutrient Dynamics

Calcium content did not change over time among adult or immature white-fronted geese, suggesting the RBA/PRV does not serve as a calcium deposition site. Carcass protein increased with Julian date only for adult females, consistent with the reliance of arctic-nesting female geese on endogenous reserves to satisfy most protein requirements for egg production (Krapu and Reinecke 1992). Fat deposition was the most dynamic component of carcass composition during spring in Nebraska and apparently was responsible for most increases in body mass. The pattern of changes in mass and fat among sex and age classes was similar to that of giant Canada geese (Branta canadensis) during late winter and spring in Minnesota (McLandress and Raveling 1981). However, daily rates of mass gain in 2.5-3.5-kg white-fronted geese were 11-22 g/day, whereas mass gains in 3.5-5.5-kg giant Canada geese were 18-36 g/day.

Midcontinent white-fronted geese breeding in the central Arctic apparently mobilize more fat reserves acquired on temperate staging areas to meet energy needs for migration and breeding than do Pacific Flyway white-fronted geese. Adult male and female white-fronted geese of the midcontinent population departed Nebraska staging areas during 28 March-5 April with fat levels that were 200 and 164% of those reported for adults males and females, respectively, of the Pacific Flyway population in Klamath Basin (Ely and Raveling 1989). Moreover, fat levels of adult male and female white-fronted geese departing from Saskatchewan were 265 and 208%, respectively, of those of Pacific Flyway geese.

The greater accumulation of fat by midcontinent white-fronted geese probably reflects limited availability of foraging habitat on breeding areas as well as plentiful grain supplies on spring staging areas. The midcontinent population breeds near the Arctic Ocean from the North Slope of Alaska eastward to Hudson Bay (Bellrose 1980), where short growing seasons (54-76 days) limit availability of plant foods to prenesting and laying females. Growing seasons on the Yukon-Kuskokwim Delta, where the Pacific Flyway population breeds, are about 117 days (Ely and Raveling 1984), and females are able to deposit 100-150 g of fat during the 2 weeks preceding nesting (Budeau et al. 1991). Whether differences in magnitude of fat stored during migration by individuals of the 2 populations represent an evolved strategy or a facultative response is not known.

Data from specimens collected during nesting and estimates of energy requirements for egg production and incubation suggest that white-fronted geese of the midcontinent population use nearly 50% of their lipid reserves between departure from temperate staging areas and nest initiation. With a mean clutch size of 4.2 eggs (n = 176 clutches) and mean egg mass of 130 ± 3 g (n = 9) (Simpson, unpubl. data), a 90% efficiency in converting fat reserves to egg energy, and an estimated energy density of 2.28 kcal/g of egg (Alisauskas and Ankney 1992b: 33), approximately 146 g of fat is needed for clutch production. Because energy requirements for incubation in white-fronted geese are equivalent to 467 g of fat (Budeau et al. 1991), total fat allocated to egg production and incubation accounts for only about 50% of the reserves acquired on midcontinent staging areas. Three females, 1 collected just prior to egg laying and 2 in early incubation, had fat deposits ( = 377 ± 68 g) similar to those of females during pre-incubation at the Yukon-Kuskokwim Delta (Budeau et al. 1991). A female collected during June in mid-incubation had 282 g of fat. Thus, female white-fronted geese from the midcontinent population mobilized as much as 50% of the fat deposited on temperate staging areas during migration and prenesting. The mean fat level of 3 males paired to the females was 191 ± 32 g, or approximately half those of their mates.

Effects of Food Resources on Fat Deposition

Most North American goose populations have switched from natural to agricultural foods on temperate staging areas (Krapu and Reinecke 1992). We think it is reasonable to hypothesize that foraging conditions prior to agricultural development were less predictable and that the rate and magnitude of fat deposition on staging areas were influenced by variations in spring weather and population density of geese. There is circumstantial evidence supporting this hypothesis. Among brant (Branta bernicla) that feed on grasses and sedges during spring migration, body mass of migrants varies among years, with decreases occurring during springs when cold temperatures reduce plant growth or when abnormally warm periods cause plants to mature and decrease in digestibility (Prop and Deerenberg 1991). In eastern Canada, increases in greater snow goose populations were accompanied by changes in use of habitat on spring staging areas in the St. Lawrence River Valley (Bedard and Gauthier 1989). Initially, populations expanded into salt marsh habitats, where foods were less digestible (Bedard and Gauthier 1989) and fat deposition was less than in preferred freshwater tidal marshes (Gauthier et al. 1984). Later, greater snow geese began feeding in croplands and metabolizable energy intake increased (Bedard and Gauthier 1989). In association with the shift to cereal grains, the May population increased by an estimated 155% from about 170,100 geese in 1979 (Anonymous 1981) to 434,500 geese in 1992 (Hughes et al. 1994).

Whether favorable foraging conditions result in increased fat deposition on spring staging areas in the Great Plains and increase annual recruitment rates in white-fronted geese and other species is not known. However, Owen (1978: 207) compared recruitment during 1968-77 between a population of Greenland white-fronted geese (A. a. flavirostris) that ate primarily cereal grains during winter and spring and another population that ate natural foods. Mean brood size of the population that ate cereal grain was 3.8 ± 0.1, whereas mean brood size of the population that ate natural foods was 2.5 ± 0.2. If increased fat deposition in staging areas affects recruitment among geese, the likely mechanism is the positive correlation between fat reserves and clutch size (Ankney and MacInnes 1978) and possibly more constant incubation and decreased nest loss (Prop and de Vries 1993).

Management Implications

The long-term availability of corn and other grains in the Great Plains is uncertain but we expect it to remain adequate for the Foreseeable future. We recommend managers maintain and restore wetland roosting habitat so white-fronted geese can adjust their distribution to changes in food supply. Restoring wetlands to increase waterfowl dispersion is the best strategy to accommodate increasing goose populations while limiting risk associated with disease epizootics. An estimated 200,000 waterfowl died in the RBA during epizootics from 1975 to 1988 and 26% of 83,344 dead waterfowl recovered were white-fronted geese (J. D. Kauffeld, unpubl. data). Where suitable wetland roosting habitat exists, private landowners should be encouraged to limit fall tillage to a single discing or other practices that maintain availability of grains and facilitate dispersion of waterfowl.

Growth in goose populations has coincided with changes in farming practices and agricultural food sources. Research is needed to understand relationships between agricultural foods and survival rates and recruitment of geese. Current levels of most North American goose populations meet or exceed goals identified in the North American Waterfowl Management Plan (Can. Wildl. Serv. et al. 1994). The winter population of midcontinent lesser snow geese has increased from an estimated 0.8 million birds during 1969-70 to 2.7 million birds in 1994-95 (Sharp 1995). In some colonies, excessive grazing associated with population increases has contributed to degradation of brood-rearing habitats and decreased juvenile survival (Williams et al. 1993). A better understanding of the role of agricultural foods in goose population dynamics is needed to respond to challenges of increasing goose populations and the threats posed by disease in such areas as the RBA.

---

---

NPWRC Home | Site Map | About Us | Staff | Search | Contact | Web Help | Copyright