Feeding Ecology of Northern Pintails and Green-winged Teal Wintering in California

Introduction

Recent evidence suggests that wintering habitat may play a major role in waterfowl breeding ecology (Krapu 1979, 1981; Heitmeyer and Fredrickson 1981). The wintering period is critical because it influences the physiological condition of ducks returning to nesting areas. Consequently, reproductive success on breeding areas may be closely related to wetland conditions on wintering areas (Heitmeyer and Fredrickson 1981). Studies on mallards (A. platyrhynchos) have demonstrated that most of the lipid requirements for reproduction come from endogenous reserves obtained on wintering or migration areas (Krapu 1981).

The importance of invertebrate foods to wintering waterfowl was not documented in most early studies (Cottam 1939, Martin and Uhler 1939, Glasgow and Bardwell 1962, McGilvrey 1966, and others). Recent studies based on esophageal data have shown that invertebrates, principally chironomid larvae, are major dietary items for wintering pintails and mallards in California (Beam and Gruenhagen 1980, Connelly and Chesemore 1980, Pederson and Pederson 1983).

Our study assesses the feeding ecology of pintails and green-winged teal in the southern Central Valley of California and relates this to habitat management patterns and habitat use by waterfowl. These species are major components of wintering waterfowl flocks in the Central Valley (Bellrose 1978) and have similar food habits (Glasgow and Bardwell 1962, McGilvrey 1966, Tamisier 1976) despite obvious size differences.

We thank T. J. Charmley, D. A. Hardt, G. W. Kramer, S. L. Lentz, F. K. Lindeman, and D. J. Severson for assisting with data collections and laboratory analyses. R. T. Alls, J. J. Chesi, and G. Grodhaus aided in the identification of seed and invertebrate samples and C. R. Luna prepared figures for publication. Financial assistance was provided by the U.S. Fish and Wildl. Serv. through Kern Natl. Wildl. Refuge (NWR).

Study Area

The study was conducted on the managed cropland units of Kern NWR, Kern County, California. The study sites represent about 65% (891 ha) of all the land flooded at Kern NWR each year. Size of wetlands average 14 ha, and the water levels are held at a relatively constant 20 cm depth during winter. The wetlands are dry during summer except when crops are irrigated. The management objective for these units is to produce nearly pure stands of barnyardgrass, swamp timothy, alkali bulrush (Scirpus paludosus), or ammannia (Ammannia coccinea). This is more or less achieved by differing irrigation schedules. Although most units develop mixtures of plants, the target plant usually emerges as the dominant species. Rank growths of tule bulrush (Scirpus acutus), cattail (Typha domingensis), baltic rush (Juncus balticus), or joint grass (Paspalum distichum) are removed periodically by mechanical means. Fall flooding begins by mid-September and about 4-6 weeks are required to inundate the study areas.

The amount of open water and cover varied among cropland types. Swamp timothy and alkali bulrush ponds both provided roughly equal proportions of open water and densely covered areas. In comparison, only about 10-15% (ocular estimation) of the surface area in barnyardgrass and ammannia ponds was open water. Open water areas in all cropland types were formed in areas where swamp timothy was dominant because of this plant's prostrate growth. However, the proportion of open water increased seasonally due to weather effects on senescent vegetation and due to the activities of waterfowl trampling dead, emergent vegetation. By the end of February, all cropland types provided roughly 70-80% open water areas.

Seasonal temperatures range from 9 to 43 C. Annual precipitation is approximately 15 cm and occurs mostly in winter and spring. Soils vary from neutral to strongly alkaline. Detailed descriptions of soils and climate of the lower San Joaquin Valley are provided by the U.S. Fish and Wildl. Serv. (1961) and Twisselmann (1967).

Methods

Habitat use patterns were determined by estimating the number of birds present in the specific marsh units. Study ponds were surveyed weekly during morning, mid-day, and afternoon hours. Ammannia habitats were not surveyed because they formed a minor area of the total croplands. In daytime, a spotting scope was used to count ducks and estimate proportions of birds feeding. At night, the number of birds of each species was estimated by visual count aided with a 12-volt spotlight.

Ducks were shot on the 4 major marsh types from October through February, 1979-80 and 1980-81 during diurnal and nocturnal periods. In daytime, ducks were shot only immediately after they had been observed actively foraging for ≥10 minutes. At nighttime, ducks were illuminated with a 12-volt floodlight and were shot as they flushed. Duck esophagi were immediately removed and their contents placed in 80% ethanol (Swanson and Bartonek 1970). Food items present in duck esophagi were sorted, identified, and measured volumetrically (Euliss 1984). Plant names used in the text are from Munz and Keck (1973) and invertebrate names follow Grodhaus (1967) for chironomids and Pennak (1978) for other invertebrates.

Food data were summarized as aggregate percentages (Swanson et al. 1974). Only birds that contained a minimum esophageal volume of 0.05 ml were included in the analyses. Aggregate percent data were compared using Mann-Whitney U-tests or Kruskal-Wallis tests. Chi-square analyses were used to compare waterfowl densities among habitat types (Siegel 1956). All statistical testing and data summaries were accomplished using Statistical Package for the Social Sciences software (Nie et al. 1975, Hull and Nie 1981).

Results

In daylight pintail densities were significantly higher in swamp timothy and alkali bulrush units during all months of the study (χ² = 6.31-1,077.10, 2 df, P <0.005 to P <0.001) than in other marsh types (Euliss 1984). Green-winged teal also followed this trend, occurring primarily in swamp timothy and alkali bulrush units (χ² = 6.08-112.52, 2 df, P <0.05 to P <0.001) during all months except February, October, and November 1980, when they were proportionately distributed among cropland types (P > 0.05). Bird densities ranged from 0.0 birds/ha for green-winged teal in barnyardgrass ponds in November 1980 to 11.9 birds/ha for pintails in alkali bulrush units in November 1979. About 50% of the pintails and 60% of the green-winged teal observed on the study ponds during the day were feeding (Euliss 1984).

At nighttime, pintails were concentrated in specific cropland types during all months (χ² = 10.90-1,654.70, 2 df, P <0.001) except during February 1981 when they were proportionately distributed among habitat types (P > 0.05) (Euliss 1984). Barnyardgrass ponds received over half of the nocturnal use on a monthly basis, and the dense cover areas of swamp timothy and alkali bulrush units received the remainder. Green-winged teal were also concentrated into specific cropland types at night during most months (χ² = 7.17-99.30, 2 df, P <0.025 to P <0.001) but were proportionately distributed among habitat types during December 1979 and February 1981 (P > 0.05). The densely covered areas of swamp timothy ponds received > 60% of the teal use at night on a monthly basis and barnyardgrass ponds received about 33%. In the entire study <5 pintails and no green-winged teal were ever flushed from open-water portions of the units at night. Nocturnal densities ranged from 0.0 birds/ha for green-winged teal in alkali bulrush ponds in February 1980 to 32.2 birds/ ha for pintails in barnyardgrass units in November 1979. We found fresh food material in the esophagi of 96% (N = 25) of the pintails and 93% (N = 101) of the green-winged teal collected at night.

Food Habits
We collected 262 pintails from the Kern NWR cropland marsh units. Diets were similar between study years, and sample sizes during nocturnal periods were too small for statistical tests, so all data were combined. When these data are lumped for all cropland types, pintails consumed 72.3% plant seeds and 27.7% animal matter (Table 1). By volume, the most abundant foods were swamp timothy caryopses (35.1%) and chironomid midge larvae (21.7%). These 2 foods occurred in 78.6 and 51.9% of all pintails collected, respectively. Foods of additional importance were barnyardgrass caryopses (11.8% ), sprangle-top (Leptochloa spp.) caryopses 5.7%, ammannia seeds (5.6%), and alkali bulrush achenes (5.2%). Together, these 6 foods formed 85.1% of the diet observed in sample birds (Table 1).

Generally, pintail esophagi collected from specific cropland habitat types contained large proportions of seeds from the major crop present in the habitat. Thus, pintail esophagi contained 41.5% barnyardgrass caryopses from barnyardgrass units, 50.8% swamp timothy caryopses from swamp timothy units, and 48.9% ammannia seeds from ammannia units. However, this trend did not persist in alkali bulrush units where pintails fed mostly on swamp timothy caryopses (37.2%) and chironomid midge larvae (27.3%). Alkali bulrush achenes formed only 13% of the diet observed for pintails in alkali bulrush units. Animal matter, mostly chironomids, formed nearly a third of the diet of pintails in all marsh types except in ammannia units where only 6.8% of the esophageal contents was animal matter.

The esophagi of 173 green-winged teal contained 62.3% plant seeds and 37.6% animal matter when data from all habitats were lumped (Table 1). Chironomid larvae and swamp timothy caryopses were the most commonly used foods. Respectively, these 2 foods formed 27.1 and 24.4% of the diet and occurred in 54.9 and 43.4% of the teal collected. Other commonly utilized foods were barnyardgrass caryopses (11.3%), ammannia seeds (5.7%), sprangle-top caryopses (4.1%), common fivehook (Bassia hyssopifolia) seeds (3.0%), and seed shrimps (Ostracoda) (2.7%). These 7 foods formed 78.3% of the observed diet in the green-winged teal sample (Table 1).

As with pintails, the food habits of green-winged teal generally reflected the dominant vegetation in the habitats, but the relationship was not as striking (Table 1). Green-winged teal esophagi collected from barnyardgrass units contained 36.2% barnyardgrass caryopses, those from swamp timothy units contained 38.0% swamp timothy caryopses, and those from ammannia units contained 35.6% ammannia seeds. Green-winged teal collected from alkali bulrush units, however, foraged mainly on chironomid larvae (46.2%) and swamp timothy caryopses (16.3%), and only 2.9% of their diet was composed of alkali bulrush achenes. As with pintails, the major animal food was chironomid larvae.

When all the cropland types are lumped, pintail and green-winged teal diets were composed of similar proportions of swamp timothy caryopses, chironomid midge larvae, and barnyardgrass caryopses (P > 0.05). However, pintails consumed relatively more plant foods and less animal foods (z = 2.33, P <0.03) than did green-winged teal.

Seasonal Trends
The food habits of pintails and green-winged teal showed a distinct seasonal shift from plant to animal foods during the wintering period (Figs. 1 and 2). Swamp timothy was the most important plant food during all months followed by barnyardgrass, ammannia, and alkali bulrush. Invertebrates formed major portions of pintail diets by December and by January for green-winged teal. Chironomid larvae formed the bulk of the invertebrate foods consumed by both duck species.

Within specific marsh types, the general trend from plant to animal foods observed for pintails was significant in barnyardgrass units (H = 29.45, P <0.001), swamp timothy units (H = 19.18, P <0.002), and in alkali bulrush units (H = 39.37, P <0.001). The seasonal shift towards animal matter in green-winged teal diets was significant in barnyardgrass units (H = 14.52, P <0.003), swamp timothy units (H = 13.10, P <0.015), and alkali bulrush units (H = 9.42, P <0.025). No statistical treatment was conducted on data from birds using ammannia units because of small sample sizes in some months although a similar trend was apparent there also.

Discussion

Aerial census data were obtained from the Kern NWR to examine the spatial distributions of pintails and green-winged teal in the Tulare Lake Basin. About 60% of the wintering pintail population was observed on inundated grain fields and about 26% was observed on the Kern NWR. Over 76% of the green-winged teal were censused on the Kern NWR and the remainder were located mostly on inundated grain fields and agricultural water storage basins.

On the Kern NWR there was a distinct shift in habitat use between diurnal and nocturnal periods. Most pintails and green-winged teal used open water ponds during the daytime and used densely vegetated units at night. During daytime, swamp timothy units and the open water portions of alkali bulrush and barnyardgrass ponds received more duck use than densely covered areas. Essentially all ducks observed during the nocturnal surveys were flushed from areas of dense, emergent vegetation in all cropland types. In view of the technique used to accomplish nocturnal surveys, it could be argued that study birds swam to covered areas before flushing. However, observations made concurrent with this study indicated that ducks were extremely rare in the open water portions of the cropland study units at night. At dusk, we observed that birds feeding in open water zones would flush for no apparent reason and land in dense cover and resume feeding, often in the same unit. This phenomenon was observed frequently for both species. Moreover, we were unable to observe ducks with a night vision scope in ponds where birds were concentrated because emergent vegetation obscured the view.

In daylight the advantage of open water ponds to waterfowl may be related to raptor predation. Tamisier (1976) reported that the presence of avian predators, mainly marsh hawks (Circus cyaneus), caused green-winged teal to move into the most open areas and to bunch together in denser than normal groups in daytime. Pintails are reported to react to potential predators similarly (Tamisier 1976). Our personal observations were similar and we believe that birds of prey may play a major role influencing diurnal habitat selection by pintails and green-winged teal.

Nocturnal food habits data suggest that a larger proportion of both species foraged at night than during the day. We found fresh food material in the esophagi of 96% (N = 25) of the pintails and 93% (N = 101) of the green-winged teal collected at night, though these birds were collected without our observing feeding behavior. If this procedure had been followed during the day, the proportion of birds containing fresh food material in their esophagi may have been much lower because only about 50% of the pintails and about 60% of the green-winged teal we observed during the day were foraging. Tamisier (1976, 1978-79) reported similar findings. Our study area was intensively managed for waterfowl food plants, many of which were abundant in all habitat types, even densely vegetated areas. Therefore, we believe that the value of dense emergent vegetation in our study area lies more in the cover it provides than for the specific foods it contains. The nocturnal use of densely covered areas may be innate; the presence of similar foods in both open water areas and in areas of dense emergent cover on Kern NWR is fortuitous. Future investigations may provide more insight into this phenomenon.

During the day, ducks may be more secure in the open where they can see predators coming, but after dark, they feel more secure in dense cover (hiding) because of lowered visibility to potential predators. In daylight they can fly to escape predators, but they may be more reluctant to do so after dark.

Food Habits and Foraging Strategies
Pintails and green-winged teal had similar food habits. For each, the 3 most important foods were swamp timothy caryopses, chironomid midge larvae, and barnyardgrass caryopses. This similarity in food habits for pintail and green-winged teal has been recognized by other authors (Glasgow and Bardwell 1962, McGilvrey 1966, Tamisier 1976). Although the 3 most important foods were common to each dabbler, the proportions of other less important foods varied significantly between species (Euliss 1984). The statistical differences are probably more a result of food availability within the specific feeding zone each duck uses rather than actual differences in food preferences. On several occasions, foraging pintails and green-winged teal were collected simultaneously and their esophagi contained essentially identical ratios of the same foods. More often, however, these ducks oriented themselves according to water depth. Both species mainly dabbled for foods from pond bottoms but tipping also was observed. The low incidence of tip-up feeding probably was related to the shallow water depths (20 cm) in the study ponds. Thus, both species mainly dabbled for foods in water depths optimal for their body size. The mean feeding depth we recorded for pintails and green-winged teal was 17.1 and 11.9 cm, respectively. This difference was significant (z = 8.46, P < 0.0001) and probably represents a major habitat partitioning factor influencing the diets of these 2 ducks. The difference in the mean feeding depth suggests that each species foraged in slightly different areas and accounted for the differences found in the consumption of minor food items. Tamisier (1976) reported that these species often segregate when they cohabitate diurnal resting ponds during the winter in Louisiana.

Food Usage
Of the factors influencing food usage, availability appears to be the most important. Floating seeds are concentrated on the water's surface where they are highly available to feeding ducks. Both pintail and green-winged teal were highly opportunistic and foraged heavily on foods that were concentrated in this manner. During October 1980, copious amounts of ammannia seeds concentrated on the surface of some study units. Although large numbers of swamp timothy caryopses were present in these ponds, pintail esophagi contained 63.9% ammannia seeds during October (Fig. 1). By the beginning of November, ammannia seeds were less concentrated and pintails began feeding on swamp timothy caryopses (Fig. 1). However, green-winged teal collected from these units in November had fed mostly on ammannia seeds (Fig. 2).

Barnyardgrass caryopses also were heavily used when highly available. These caryopses floated and concentrated on pond surfaces until late December and were heavily utilized by pintails (Fig. 1) and green-winged teal (Fig. 2). Pintail use of barnyardgrass caryopses declined in January and increased again during February (Fig. 1). By February, barnyardgrass caryopses were concentrated on pond bottoms. Pintails responded to this change in availability by feeding off the pond bottoms. This change is apparent in their February diet (Fig. 1); our observations of increased tip-up feeding supports this conclusion. Green-winged teal did not respond accordingly, probably reflecting the disadvantage of their shorter necks and bodies when feeding in deep water.

The use of swamp timothy also was greatly influenced by its extreme availability. Pintails contained up to 117,400 swamp timothy caryopses/esophagus. These caryopses are roughly 0.5 × 1.0 mm in size and it is doubtful if pintails could forage efficiently on this food unless the caryopses were concentrated. The seed dispersal mechanism of swamp timothy involves the migration of caryopses to the periphery of the panicle (Euliss 1984). We assume that ducks feeding on swamp timothy place the panicles inside their mouths and brush the loosely attached caryopses free with the spines of their tongues. Thus, a large number of caryopses could be obtained in a short time. Esophagi collected during the fall contained caryopses that were very clean and free of chaff and other vegetative material. Later in the season as the plants began to decompose, chaff and other vegetative debris became numerous in duck esophagi. It is unlikely that these vegetative portions were of much nutritional value and probably were ingested incidentally when the birds fed on caryopses from deteriorating plants.

Use of chironomid midges also was influenced by their extreme availability. During February 1981, a barnyardgrass pond produced high concentrations of Cricotopus sp., an epiphytic chironomid. Two green-winged teal esophagi collected from this pond during February contained 6,053 and 4,925 midge larvae.

The seasonal diets of pintails and green-winged teal showed a distinct shift from plant to animal foods by the beginning of January. Invertebrates must re-establish their populations in seasonal marshes because of annual dessication. The mild fall and winter climate at Kern NWR appears to be conducive to the rapid development of initial invertebrate colonizations, particularly chironomids. However, the general cooling of the water during winter lengthens their life cycles and slows the rate of repopulation (Oliver 1971). Consequently, invertebrate populations do not become dense enough to provide a significant food base for waterfowl until late winter or early spring. In the fall, seeds were abundant and were used extensively by waterfowl until their availability decreased or the availability of invertebrates increased. This temporal shift in the availability of food resources provides protein-rich invertebrate foods by early spring just before they are required for the optimal reproduction of waterfowl.

Management Implications

Invertebrates, particularly chironomids, formed substantial portions of the diet for both pintails and green-winged teal. Similar findings have been observed at other California refuges (Beam and Gruenhagen 1980, Connelly and Chesemore 1980, Pederson and Pederson 1983). Based on this evidence, the value of the croplands depends to a great extent on invertebrates, which are incidental to current management practices. Future investigations addressing the habitat requirements and seasonal production of invertebrates may enable managers to enhance invertebrate populations thereby increasing the overall quality of wintering ground marshes.

Current management practices on Kern NWR appear to satisfy the needs of wintering waterfowl. However, management must intensify to counterbalance the effects of wetland loss and degradation elsewhere (Bellrose and Low 1978). Such practices as delayed dewatering in the spring will provide spring migrants with additional foods, mainly invertebrates, to ensure fitness during northward flights. The continued maintenance of shallow pond depths will enhance the availability of food items to waterfowl. Lastly, the development of units having dense, emergent vegetation for nighttime use is recommended. Due to constraints posed by high water costs over the last decade, waterfowl management in the entire San Joaquin Valley and elsewhere in California has been directed towards propagation of marsh plants with low irrigation requirements, particularly swamp timothy. These ponds receive great use during the day but negligible use at night. Ideally, wintering ground refuges should provide similar proportions of open water and densely vegetated marsh units to satisfy the 24-hour habitat requirements of ducks. It is clear that a continued shift to open-water marshes could lower the overall carrying capacity of California wetlands and should be discontinued until its effect is known.