Northern Prairie Wildlife Research Center
The number of occupied farms on the study area decreased from 21 in 1951 to 8 in 1992-93, and the number of landowners decreased from 38 to 22 during the same time period. Average farm size increased from 76 to 132 ha. Fourteen of 22 landowners in 1992-93 were either present as tenants or as offspring of tenants present during the original study.
Access was granted to 2,850 ha (97.8%) of the study area. Two landowners of small tracts did not allow access, but we were able to view these areas from adjacent properties.
The amount of land used for annual cropping decreased from 63.0% in 1950-53 to 29.7% in 1992-93 (Table 2), a 52.9% drop.
Much (27.5%) of the land that was annually cropped in 1950-53 was classified as highly erodible and enrolled in the CRP (Johnson and Schwartz 1993) for 10-year contracts beginning in 1986. Once enrolled, the land was seeded to mixtures of cool-season grasses and legumes. This is the main reason for the large decrease in annually cropped land on the study area since 1953.
Agricultural crops in the study area in the early 1950s were limited to spring wheat, flax (Linum), barley, and oats. Crops in 1992-93 also included winter wheat, rye, buckwheat, corn, soybeans, and alfalfa (Table 2). Flax is still grown in the area but was not planted in the study area during 1992-93. Sunflowers (Helianthus) are a commonly grown row crop in the region but were not planted in the study area during 1992-93. Overall, fewer hectares of row crops were planted in 1993 than in 1992 because of the wet growing season.
Wetland Classification. Eight different classes of wetlands (Stewart and Kantrud 1971) were found in the study area in 1992-93 (Fig 6). Of 504 wetlands, 43.5% were categorized as class III seasonals and 24.3% were in tillage classes III t, T-2, T-3, and T-4 during 1992-93 (Fig 6).
|Figure 6. Percent of wetlands in different wetland classes (Stewart and Kantrud 1971) in the Waubay study area, Day County, South Dakota, 1992.|
Wetland Drainage. Drainage on the study area is extensive, with evidence of some type of drainage on 180 of 504 (35.7%) wetlands (Table 3).
On the basis of wetland class, proportionately more drainage occurred in class I ephemeral (45.6%) and class II temporary (56.8%) basins and in tillage classes T-2, T-3, T-4, and III t (Table 3). All drainage in the study area was by open ditches.
Drainage records from previous studies in the Waubay study area were only available for 173 of 180 wetlands that have at least been partially drained since the 1950s. The number of drained basins in each wetland class was greatest after 1968 and smallest between 1954-68 (Table 4). Although fewer basins were drained before 1954 than after 1968, more wetland area was drained before 1954 than after 1968 (Table 4).
The USFWS restored three of five (60.0%) previously drained class IV wetlands in the study area and smaller percentages of temporary (32.0%) and seasonal (22.1%) wetlands (Tom Wickstrom, Waubay National Wildlife Refuge, pers comm) (Fig 7).
|Figure 7. Percent of 173 wetland basins by wetland class (Stewart and Kantrud 1971) that have been drained or restored in the Waubay study area, Day County, South Dakota, as of 1992-1993.|
More total ducks and breeding pairs were present in the study area in 1993 than in 1992 (Table 5). Eight species of dabbling ducks, or surface-feeding ducks, [blue-winged teal, gadwall, mallard (Anas platyrynchos), northern pintail (A. acuta), northern shoveler, American wigeon, green-winged teal (A. crecca), wood duck (Aix sponsa)]; five species of diving ducks [redhead, canvasback (Aythya valisineria), ruddy duck, lesser scaup (A. affinis), ring-necked duck (A. collaris)]; and giant Canada geese were observed during 1992-93.
The same species occurred in the study area in 1950-53 (Table 6). However, only two pairs of Canada geese were seen in this period (Evans and Black 1956).
Species Composition. Dabbling ducks made up 83.8% and diving ducks 16.2% of total pairs during 1992-93 (Table 6). Pairs of blue-winged teal were the most common (39.3%) dabbling ducks during 1992-93 and 1950-53 (47.3%) (Table 6). Pairs of mallards were the second most common (22.0%) dabbling ducks during 1992-93 and third most common (13.4%) during 1950-53 (Table 6).
Redhead pairs were the most common (11.7%) diving ducks during 1992-93, but ruddy duck pairs were the most common (3.6%) in 1950-53 (Table 6). Redhead pairs were the second most common (3.2%) diving duck during 1950-53 (Table 6).
Breeding Pair Densities. Total pair densities (pairs per km²) were larger in 1992 (t=2.59, df=13, P<0.05) than in 1951 and were larger in 1993 (t=2.33, df=13, P<0.05) than in 1953. In fact, the total pair density in 1993 was the second largest ever recorded in the Waubay study area.
Density of all species increased from 1950-53 to 1992-93 except for northern pintails and lesser scaup. Density of all species except gadwalls, wood ducks, and lesser scaup also increased from 1992 to 1993 (Table 7).
The large annual variations in pair densities of waterfowl are strongly related to precipitation during the growing season (April-August; Fig 8). Annual patterns of local precipitation paralleled annual patterns of the local breeding population (Fig 8).
|Figure 8. Total pairs densities (pairs/km²) in relation to April-August precipitation (cm) in the Waubay study area, Day County, South Dakota, 1950-1965, 1992-1993.|
Wetland Conditions. More than twice as many ponds (54.7%) contained water in early May in 1993 than in 1992 (24.9%; Fig 9). Water conditions followed similar patterns during 1992 and 1993.
|Figure 9. Percent of basins in the Waubay study area, Day County, South Dakota, containing >10% standing water at four time periods from May-August, 1992-1993.|
The number of wet basins was smallest in late May 1992 and 1993, but abundant June and July precipitation (Table 1) filled a large percentage of the wetlands by late July (Fig 10). The number of wet basins was largest in late July of 1992 and 1993 (Fig 9).
|Figure 10. Road on east side of the Waubay study area, Day County, South Dakota, flooded from abundant summer rain fall in 1993.|
Use of Different Wetland Classes and Sizes by Pairs. In 1993, pairs of dabbling ducks were seen most frequently in class III and class IV wetlands (Table 8).
Northern pintails, northern shovelers, and blue-winged teal were seen most frequently in class III wetlands. Most (84%) pairs of diving ducks were seen in class IV wetlands (Table 8). However, redhead pairs (15.4%) were also quite common in class III wetlands. Ruddy duck and canvasback pairs were seen almost exclusively in class IV wetlands (Table 8).
Of the tilled wetland classes, pairs of dabbling ducks were seen most frequently in class III t (3.3%) and class T-2 (2.8%) basins (Table 8). Few diving duck pairs were found on tilled wetlands.
The greatest density of dabbling duck pairs per hectare of wetland area in each class were seen in class II (11.84) natural basin wetlands and class T-2 (19.56) tillage wetlands (Table 9). The greatest density of diving duck pairs per hectare of wetland area in each class were seen in class IV (1.13) natural basin wetlands and class T-4 (9.09) tillage wetlands (Table 9). However, only two class T-4 wetlands were in the study area in 1992-93.
The largest percentages of pairs of all species in 1993 were in wetlands larger than 4.81 ha (Figs 11, 12). Pairs of dabbling ducks were also well represented on all wetlands larger than 0.12 ha (Fig 11). Diving duck pairs were seen almost exclusively in wetlands larger than 4.81 ha (Fig 12). Redhead pairs were observed on more wetlands of different sizes than other diving duck species.
|Figure 11. Use (%) of different wetland sizes (ha) by dabbling duck pairs in the Waubay study area, Day County, South Dakota, 1993.|
|Figure 12. Use (%) of different wetland sizes (ha) by diving duck pairs in the Waubay study area, Day County, South Dakota, 1993.|
Pair observations of each species on wetlands of different sizes varied annually in 1992-93 and reflected the available wetland habitat (Table 10).
Observations of pairs of dabbling ducks in wetlands smaller than 0.36 ha were considerably fewer during the early 1950s than in 1993, whereas observations of pairs of dabbling ducks in wetlands larger than 1.98 ha were greater during the early 1950s than in 1993 (Table 10). Pairs of dabbling ducks used wetlands smaller than 0.76 ha during the extremely wet conditions of 1993.
Over-Water Nests. Proportions of 66 over-water nests found in 1992 were redheads (50.0%), mallards (31.8%), ruddy ducks (13.6%), pintails (3.0%), and canvasbacks (1.5%). Proportions of 64 over-water nests found in 1993 were redheads (46.9%), canvasbacks (25.0%), ruddy ducks (19.7%), and mallards (7.8%).
The average clutch sizes of over-water nests were 10.6 (redheads), 7.9 (mallards), 8.1 (ruddy ducks), 8.0 (pintails), and 8.8 (canvasbacks) eggs. Overall clutch success of over-water nests calculated using the Mayfield technique was 14.2% (95% CI=7.6%-26.3%) in 1992 and 23.6% (95% CI=14.6%-37.8%) in 1993 (Table 11).
Species specific clutch success varied between years. However, our sample of northern pintail nests was small. The distribution of nest initiation dates of mallards (Appendix D), blue-winged teal (Appendix E), redheads (Appendix F), canvasbacks (Appendix G), and ruddy ducks (Appendix H) indicated an extended 1993 nesting season, because many basins were wet throughout the breeding season.
Nine giant Canada goose nests were found in 1993, and seven clutches (36.5% Mayfield) hatched. Average clutch size of giant Canada goose nests was 6.1 eggs.
Destruction of over-water nests in 1992-93 was caused by mammalian predation, flooding, and abandonment (Table 12).
Predation (36.2%), mainly by raccoons (Procyon lotor), was the leading cause of nest destruction in over-water nests. However, losses from mammalian predation decreased as nest initiation dates became later (Fig 13). Flooding was the second leading cause of nest destruction in over-water nests in 1992 (n=9, 13.6%) and 1993 (n=11, 17.2%). Evidence of a hen being killed was apparent in only one over-water nest.
|Figure 13. Percent of nests destroyed by mammalian predators during 6 different nest initiation periods in the Waubay study area, Day County, South Dakota, 1992-1993.|
Nest parasitism was highest in 1993. Twenty-five of 64 (39.1%; 10 canvasbacks, 12 redheads, three ruddy ducks) over-water nests were parasitized in that year, with 60.0% of the parasitized clutches hatching. Six of 66 (9.1%; four redheads, one mallard, one ruddy duck) were parasitized in 1992, with 16.7% hatching. Redheads were the parasitic layers in 30 of 31 (96.8%) parasitized nests.
In 1993, 62.5% of the canvasback nests, 40.0% of the redhead nests, and 23.1% of the ruddy duck nests that we found and monitored were parasitized.
Most (90.3%) over-water nests during 1992-93 were in class IV wetlands (Table 13), and most (72.2%) were in dense, monotypic stands of residual-growth cattail (Typha). Most redhead (77.8%), ruddy duck (57.1%), canvasback (76.5%), and mallard (65.4%) nests were in dense, monotypic stands of residual growth cattail. Ruddy ducks frequently constructed nests in new growth cattail.
Hardstem bulrush (Scirpus acutus), softstem bulrush (S. tabernaemontani), and river bulrush (S. fluviatilis) occupied greater than 5% of the basin in 65.6% of class IV wetlands, but only 11.9% of the nests were in these habitats.
Only 9.7% of over-water nests were in seasonal wetlands (class III; Table 13). Nests in seasonal wetlands were in whitetop (Scolochloa festucacea), slough sedge (Carex atherodes), river bulrush, or giant burreed (Sparganium eurycarpum).
Mean water depth and mean distance to the nearest shoreline from nest sites varied by species (Table 14). Mallards nested over considerably shallower water than did other over-water nesting ducks (Table 14).
Water depths at nest sites of all species varied, but no nests were found where water was deeper than 101.6 cm. Only 7 of 127 (5.5%) nests were over depths greater than 76.2 cm, even though wetlands with good emergent cover in water deeper than 77 cm were searched. Nests of ruddy ducks and mallards were closest to shoreline, and nests of redheads and canvasbacks were farthest from shoreline (Table 14).
Upland Nests. Twenty-three duck nests in 1992 and 17 nests in 1993 were found in upland habitats (Table 15). Largest percentages were on wetland edges (25.0%), grasslands enrolled in CRP (22.5%), and pastures (20.0%).
Smooth brome (Bromus inermis) (62.5%), Kentucky bluegrass (Poa pratensis) (35.0%), and alfalfa (27.5%) were the most common plant species at nest sites. The proportions of upland nests found in 1992 were blue-winged teal (56.5%), mallards (26.1%), northern shovelers (4.3%), gadwalls (4.3%), and northern pintails (8.7%). In 1993 the proportions were blue-winged teal (52.9%), mallards (23.5%), northern shovelers (11.8%), and gadwalls (11.8%). Average clutch sizes for upland nests by species were 10.1 eggs (blue-winged teal), 10.5 (northern shoveler), and 8.7 eggs (gadwall). Overall clutch success for upland nests calculated using the Mayfield technique was 8.2% (95% CI = 2.5%-26.2%) in 1992 and 5.7% (95% CI = 1.1%-26.9%) in 1993 (Table 15).
Destruction of most (67.5%) upland nests was caused by mammalian predation (Table 12), and predation continued at a high rate throughout the nesting season (Fig 13). A total of 238 ducklings hatched from monitored nests in 1992; 188 ducklings hatched in 1993 (Table 16).
Diving ducks accounted for 80.6% of the total ducklings hatched from monitored nests in 1992 and for 79.3% in 1993. Redhead ducklings were the most numerous diving duck hatched from monitored over-water nests while blue-winged teal ducklings were the most numerous dabbling duck hatched from monitored upland nests (Table 16).
Recruitment of Ducks. One hundred broods were counted in 1992 and 183 in 1993 (Table 17). Dabbling-duck broods accounted for 66.0% of all broods in 1992, 74.3% in 1993, and 87.8% during 1950-53 (Table 17).
Counted on wetlands from distant observation points with binoculars or telescopes were 66% of all broods in 1992, 60.7% in 1993, and 75.0% in 1950-53.
Because many wetlands in 1992 were completely choked with emergents, brood counting was difficult. For example, only 20 of 505 (4.0%) wetlands had enough open water in 1992 for the survey of broods with binoculars or telescopes; whereas in 1953, broods could be surveyed by this technique in a minimum of 79 of 505 (16.0%) wetlands. In 1993, abundant rainfall allowed additional wetlands to be surveyed for broods with binoculars or telescopes.
In 1992, 25.0% of the counted broods were mallards, 19.0% were blue-winged teal, 16.0% were gadwalls, 5.0% were northern shovelers, 1.0% were wood ducks; 19.0% were redheads, 7.0% were ruddy ducks, and 4.0% were canvasbacks (Table 17).
In 1993, 37.7% of the counted broods were blue-winged teal, 14.2% were gadwalls, 13.1% were mallards, 6.6% were northern shovelers, 2.2% were northern pintails, 0.5% were green-winged teal; 9.3% were ruddy ducks, 4.9% were redheads, 4.9% were canvasbacks, and 0.5% were lesser scaup (Table 17).
During 1950-53, 52.2% of counted broods were blue-winged teal, 15.3% were gadwalls, 9.3% of the counted broods were mallards, 8.1% were northern pintails, 2.1% were northern shovelers, 0.4% were green-winged teal, 0.4% were American wigeon; 8.5% were ruddy ducks, 2.5% were redheads, 0.9% were lesser scaup, and 0.4% were canvasbacks (Table 17).
Brood Densities. Total brood densities did not differ statistically between 1992 and 1951. However, in 1993, total brood densities (t=2.11, df=12, P<0.10) were significantly less than in 1953. In 1992-93, total brood densities (t=1.94, df=12, P<0.10) were significantly less than during 1950-53 (Table 18).
The density of broods of all species decreased since the early 1950s except for northern shovelers, redheads, and canvasbacks (Table 19). Northern pintails showed the highest percentage decrease in brood densities since the early 1950s, and canvasbacks had the largest percentage increase (Table 19).
All species had percentage increases in brood densities from 1992 to 1993 except for redheads and mallards (Table 19).
Total brood densities during 1992 and 1993 were lower on average than during 1950-63 (Fig 14). For example, total brood density was 9.7/km² in 1963 and 3.4/km² in 1992 (Fig 14).
|Figure 14. Annual pair densities (pairs/km²) compared to annual brood densities (broods/km²) in the Waubay study area, Day County, South Dakota, 1950-1955, 1958-1963, 1992-1993.|
Indices of hen success in 1992-93 and 1950-53 follow the same pattern. Overall hen success was 14.2 in 1992-93 and 36.4 in 1950-53 (Table 20). Hen success in all species declined since 1950-53 (Table 20).
Use of Different Wetland Classes and Sizes by Broods. Broods of blue-winged teal (58.6%) and northern shovelers (88.2%) were seen most frequently in class III wetlands. Broods of mallards (55.1%) and gadwalls (71.4%) were seen most often in class IV wetlands (Fig 15). Canvasback (69.2%) and redhead (85.7%) broods were seen most often in class IV wetlands (Fig 15). Ruddy duck broods were seen exclusively in class IV wetlands (Fig 15).
|Figure 15. Use (%) of different wetland classes (Stewart and Kantrud 1971) by broods in the Waubay study area, Day County, South Dakota, 1992-1993.|
Broods of blue-winged teal, mallards, gadwalls, and northern shovelers were seen most frequently in wetlands larger than 0.81 ha in 1992 and 1993, whereas broods of redheads, canvasbacks, and ruddy ducks were seen most often on wetlands larger than 6.07 ha (Table 21). Broods of northern shovelers and canvasbacks were found in wetlands of various sizes.