The Condor 93582-590 0 The Cooper Ornithological Society 1991

FEEDING ECOLOGY OF WATERFOWL WINTERING ON EVAPORATION IN ’

NED H. EULISS, JR.~ AND ROBERT L. JARVIS Department of Fisheriesand Wildlife, Oregon State University,Corvallis, OR 97331

DAVID S. GILMER U.S. and Wildlif Service,Northern Prairie WildLyeResearch Center, 6924 Tremont Road, Dixon, CA 95620

Abstract. We examinedthe feedingecology of NorthernPintails (Anas acuta), Northern Shovelers(A. clypeata), and Ruddy Ducks (Oxyura jamaicensis) winteringon drainwater evaporationponds in California from 1982 through 1984. Pintails primarily consumed midges(Chironomidae) (39.3%) and widegeongrass(Ruppia maritima) nutlets (34.6%). Shovelersand RuddyDucks consumed 92.5% and 90.1%animal matter,respectively. Water boatmen(Corixidae) (5 1.6%),rotifers (Rotatoria) (20.4%), and copepods (Copepoda) (15.2%) werethe most importantShoveler foods, and midges(49.7%) and waterboatmen (36.0°@ werethe most important foods ofRuddy Ducks. All threespecies were opportunistic foragers, shiftingtheir dietsseasonally to the mostabundant foods given their behavioraland mor- phologicalattributes. Kev words: Aauatic invertebrates:evaooration ponds; feeding ecology;Northern Pintails; Northern Shovelek: Ruddy Ducks.

INTRODUCTION is by evaporation in shallow ponds (hereafter Agricultural development has led to a 94% loss called evaporation ponds). of historic in the Central Valley of Cal- There are presently about 3,000 ha of evap- ifornia (U.S. Fish and Wildlife Service 1978). In oration ponds in the Central Valley of California the and Buena Vista Lake basins, (Barnum and Gilmer 1988). This study was con- located in the southern , about ceived to evaluate the diets of several speciesof 250,000 ha of shallow wetlands have been con- waterfowl that used evaporation ponds. We se- verted to irrigated agriculture(Gilmer et al. 1982). lected Pintails (Anus acuta), Shovelers (A. cly- Historically, the region was a major concentra- peutu), and Ruddy Ducks (Oxyuru jumuicensis) tion area for waterfowl but present use is con- for sampling becausethey commonly used evap- fined to habitats provided by the Kern National oration ponds and each represented a distinct Wildlife Refuge (NWR), private duck clubs, wa- feeding mode (i.e., dabbling, filtering, and diving) ter storagebasins, flooded agricultural fields, and (Bellrose 1980). evaporation ponds. Traditional irrigation prac- STUDY AREA tices cause salts to concentrate in upper soil pro- files, frequently limiting plant growth. As a re- We studied waterfowl on drainwater evaporation sult, farmers have installed subsurfaceirrigation ponds operated by the Tulare Lake Drainage drainage systemsto remove salts from upper soil District (TLDD) in Kings and Kern counties, profiles of irrigated fields. This drainwater is sa- California. This region is characterized by long line and contains heavy metals and other envi- dry summerswith annual rainfall averagingabout ronmental contaminants (Presser and Barnes 15 cm (Kahrl1979). These ponds, built in 1980- 1985). Presently, the only economically accept- 1982, consisted of three separate evaporation able means of disposingof subsurfacedrainwater systems (EPS) that collectively comprised 18 separate ponds. Ponds ranged in size from 22- 104 ha (SZ= 65 ha). EPSs contained 4 or 10 interconnected ponds that allowed drainwater to ’ ’ Received17 September1990. Final acceptance11 flow through interconnected ponds to a terminal March 1991. 2Present address: U.S. Fish and Wildlife Service, cell. Ponds were generally < 1 m deep with flat Northern Prairie Wildlife ResearchCenter, Route 1, bottoms. Drainwater entering an EPS was about Box 96C, Jamestown,ND 58401-9736. 5-l 0 mS/cm electrical conductivity (EC) but in-

[5821 WATERFOWL FEEDING ECOLOGY 583 creasedin successiveponds due to evaporation averagesof potential food items were con- to >300 mS/cm EC (TLDD, unpubl. data). sidered as sample replicates. We sorted food items collected from duck METHODS esophagi and pond samples into taxonomic We collected ducks by shooting after observing groups,and dried them to constant weight at 55- them feed for 2 10 min from September through 60°C for 24 hours. Martin and Barkley (196 l), March, 1982-1984. Shotguns were used to col- Grodhaus (1967), Pennak (1978), and Merritt lect birds within 40 m of shore and rifles were and Cummins (1984) were used to identify food used to collect birds foraging >40 m of shore. items. Water depth was recorded (f 0.1 cm) at the feed- We summarized food habits and standing ing location when birds were retrieved. At night, data as aggregatepercent (Swansonet al. we did not observe feeding behavior, but illu- 1974) dry mass. Statistical analyses were per- minated birds with a 12-volt floodlight and shot formed only with birds that contained I 5 mg of them as they flushed (Euliss 1984). Water depths food in their esophagi.We used aggregatepercent at feeding locations were not determined for birds dry masses of total invertebrates consumed as collected at night because exact foraging loca- the dependent variable in an analysis of variance tions were not observed. Observation of feeding (ANOVA) to evaluate the effect of treatments: behavior prior to collection is not necessaryto month, year, bird age, time of collection (diurnal obtain nocturnal food habits data (Euliss 1984) versus nocturnal), age of EPS, and all possible becauseducks mostly forage at night (Euliss and interactions. An arcsine transformation was re- Harris 1987, Tamisier 1978/79). We removed quired to stabilize the variance of aggregateper- their esophagi immediately and preserved their cent dry massesof food items. Overall differences contents in 80% ethanol (Swanson and Bartonek in use of specificfoods among duck specieswere 1970). Aquatic invertebrates lose dry mass when assessedwith ANOVA, and Student-Newman- stored in ethanol (Howmiller 1972). Hence, the Kuels (SNK) multiple comparison test was used proportions of invertebrates in waterfowl diets to locate differences. Orthogonal contrasts were reportedherein are conservative. Birds were sexed used to evaluate seasonal changesin waterfowl and aged using plumage characteristics(Carney diet. Water depths recorded at each site where 1964). Bursal examinations were also considered birds were collected required a squareroot trans- when plumage characteristics alone were insuf- formation to stabilize the variance. We tested ficient to positively classify age. transformed data for differences among duck Standing biomass of potential waterfowl food species using a SNK multiple comparison test items (aquatic invertebrates and seeds) present after the null hypothesis had been rejected using in evaporation ponds was determined from Oc- ANOVA. tober through March, 1982-1983 and from Sep- tember through March, 1983-l 984. We collect- RESULTS ed samplesfrom nine (18 pondstotal) evaporation ponds throughout the study; study ponds were POND BIOTA selectedacross a gradient (10 to > 70 mS/cm EC) Diversity of aquatic plants and invertebrates was of salt content. Equally spaced transects were low relative to that in surrounding freshwater establishedin each study pond and sampleswere wetlands, but the taxa present were often highly collected at random points along each transect. abundant (Euliss 1989). Widgeongrass (Ruppiu and benthic biota were collected maritima) was common in ponds having 40-75 with samplers modified after those described by mS/cm EC, and horned pondweed (Zannichellia Swanson (1978a, 1978b). Benthic and water col- palustris), occasionally observed in less saline umn sampleswere cleaned by sieving with a self- ponds, was not abundant. Midge larvae and wa- cleaning (0.5 mm mesh) screen(Swanson 1977). ter boatmen composed the bulk of the foods A benthic and a water column sample was col- available (Table 1). We recorded only two species lected from each of 10 transects in 1982-1983. ofmidge larvae; Tanypusgrodhausiwas the most To reduce within sample variance, we increased common. Similarly, the bulk of the water boat- to 20 transects and collected a benthic and a men biomass was formed by Trichocorixa reticu- water column sample per transect in 1983-l 984. lata although Corisella spp. was present during Each pond was sampled every three weeks and the spring. Additionally, copepods (Copepoda),

WATERFOWL FEEDING ECOLOGY 585

80-

20-

SEP OCT JAN FEB

MONTH

ta WATER BOATMEN EGGS ix FOG WEED SEEDS jgj IID WATER BOATMEN ALKALI WEED SEEDS

a MIDGES isi WIDGEONGRASS NUTLETS

OTHER ANIMAL 8 PEPPERGRASS SEEDS q RED BROME CARYOPSES q OTHER PLANTS

FIGURE 1. Seasonalfood habits of NorthernPintails collected from agriculturaldrainwater evaporation ponds in the San JoaquinValley, California,during Septemberthrough March 1982-1984.

March (t = -4.94, df = 5, P = 0.0001) repre- 4.9 cm for Shovelers,and 9.5 cm for Ruddy Ducks sented declines over previous months. Midges (df = 525, P < 0.05). were consumed most frequently during the later DISCUSSION half of the wintering period with increases over previous months occurring in February (t = 2.83, FOOD HABITS AND FORAGING df = 5, P = 0.0052) and March (t = 4.35, df = STRATEGIES 5, P = 0.0001). Average feeding depths we recorded for Pintails were similar to those reported by Euliss and Har- SELECTION OF FORAGING SITES ris (1987). Thus, feeding in the TLDD ponds (X Water depths at diurnal feeding sites were sig- 60-80 cm depth) was restricted to shallow areas nificantly different for each duck species.Mean along pond margins. Further, overall diurnal use depths at feeding sites were 2.8 cm for Pintails, by Pintails of TLDD evaporation ponds was low 586 N. H. EULISS, JR., R. L. JARVIS AND D. S. GILMER ...... a ...... d ...... 4 . .. . 4 . . . ti I.. .

SEP OCT NOV DEC JAN FEB MAR m (n=13) (n=13) (rjZ13) (nz11) (g=30) (!!=a (!!=17) i LL MONTH s 15 II

g$g Q ROTIFERS ALKALI WEED SEEDS q COPEPODS CLOVER FERN SPOROCARPS

II0 WATER BOATMEN n FESCUE CARYOPSES

MIDGES Ls WIDGEONGRASS NUTLETS

OTHER ANIMAL ta OTHERPLANTS

FIGURE 2. Seasonalfood habits of Northern Shovelerscollected from agriculturaldrainwater evaporation pondsin the San Joaquin Valley, California, during September through March 1982-1984.

relative to use of other available habitats, with (Euliss 1989) may have affecteduse of the ponds. most use occurring in September and again in Animal foods were used by Pintails during Sep- February and March (Coe 1990). Only 4% of the tember and then again in March, a seasonalpat- Pintails surveyed in the Tulare Lake Basin during tern that may relate to protein requirements of 198 1-1987 were on evaporation ponds (Barnum feather molt (Heitmeyer 1988) or reproduction and Euliss 1991). Moreover, Pintails forage ex- (Krapu 1979, 198 1). Thus, the importance of tensively on plant seeds during winter in Cali- TLDD ponds to Pintails may relate to the abun- fornia (Connelly and Chesemore 1980, Euliss and dance and availability of animal foods during Harris 1987, Miller 1987) and the low avail- time periods when their need for animal proteins ability of plant seedsin TLDD evaporation ponds is high (Euliss 1989). WATERFOWL FEEDING ECOLOGY 587

SEP OCT NOV DEC JAN FE0

5 MONTH 4. ii

q BRINE FLIES llllnuFILAREE lcllWATER BOATMEN a WIDGEONGRASS NUTLETS

i5zi MIDGES ia OTHER PLANTS

OTHER ANIMAL

FIGURE 3. Seasonalfood habits of Ruddy Ducks collected from agricultural drainwater evaporation ponds in the San Joaquin Valley, California, during September through March 1982-1984.

Shovelers and Ruddy Ducks consumed large ing by Shovelers allowed them to exploit small quantities of animal foods; both species con- foods, such as rotifers and copepods,that prob- sumed water boatmen with more being con- ably were not consumed by Pintails or Ruddy sumed early in the winter than during late winter Ducks. to early spring. For Ruddy Ducks, reduced con- Shovelers and Ruddy Ducks relied on water sumption of water boatmen was offset by in- boatmen. However, competition was seemingly creased consumption of midges. Shovelers in- avoided becausesimultaneous use occurred only creasedconsumption of rotifers in December and during periods when water boatmen were ex- February and copepods during March as con- tremely abundant; numbers of individuals often sumption ofwater boatmen declined. Filter feed- exceeded 200,000 individuals/m2 in certain lo- 588 N. H. EULISS, JR., R. L. JARVIS AND D. S. GILMER cations within ponds (Euliss 1989). While water different sized foods (Zweers 1980). Because of boatmen have been reported as important foods their small size, we did not attempt to quantify of Shovelers(Tietje and Teer 1988), Ruddy Ducks standing crops of rotifers in this study. However, are considered mostly as predators of midge lar- pumping water through extremely fine lamellae vae (Siegfried 1973, Hoppe et al. 1986). Thus, gapsto consume small foods would be costly and what may appear as competition for a food item would not be worthwhile unless they were suf- may instead be an opportunistic response by ficiently abundant to offsetthe energeticexpense. Ruddy Ducks to an extremely abundant food. The two most abundant macroinvertebrates in Filtering seems more adaptive and efficient in TLDD ponds were water boatmen and midges. capturing small swimming prey, including water Collectively, these taxa were observed to exceed boatmen, whereasdiving for foods appearsto be 400,000 individuals/m2 in certain locations more profitable in obtaining midges and other within ponds (Euliss 1989). These two insects lessmobile prey items along sediment interfaces. were readily available to foraging ducks as a re- sult of this abundance. FOOD USAGE Evaporation ponds in the San Joaquin Valley Food availability and nutritional need influence will continue to attract waterfowl because of food use given behavioral and morphological dif- abundant food production and the availability ferences of individual duck species (Euliss and of large areas that provide sanctuary. These in- Harris 1987, Miller 1987). Each specieswas op- vertebrate-rich wetlands are well suited to ducks portunistic and foraged on foods that were con- such as Shovelers and Ruddy Ducks that con- centrated. Wind was an important factor in con- sume large quantities of animal foods throughout centrating floating foods along windward shores their annual cycles. Evaporation ponds provide and making them readily available to feeding ready sourcesof protein-rich invertebrate foods ducks. Pintails responded to food availability on that are required by dabbling ducks to satisfy study ponds by feeding on widgeongrassnutlets protein requirements during reproduction and windrowed as the result of foraging activities of feather molt. However, the effect of environ- Redheads (Aythyu americana) and American mental contaminants on water birds, including Wigeon (Anus americana). waterfowl, needs to be considered. Drainwater Wind also altered the availability of several used to fill evaporation ponds is known to con- invertebrate taxa in the ponds. Pupae of midges, tain selenium (Presserand Barnes 1985), that has ephydrids, and other diptera float to the surface been associated with embryonic mortality and just prior to emergencewhere they are vulnerable deformity ofwater birds (Ohlendorfet al. 1986a, to predation. Pintails and Shovelers were ob- 1986b). Further, high concentrations of CaCO, served to take advantage of this phenomenon in TLDD evaporation ponds have been observed and foraged on these insects where concentra- to precipitate on and result in severe erosion of tions existed, for example along shores where Ruddy Duck rectrices@liss et al. 1989). In areas wind concentrated emerging diptera adults. One where environmental contaminants or excessive male Pintail we observed feeding in this manner salts pose threats to waterfowl, clean, food-rich, had consumed nearly 27,000 freshly emerged alternate freshwaterhabitats could be createdad- midges. We did not observe surface feeding by jacent to EPSs to reduce waterfowl exposure to Ruddy Ducks. contaminants. Further, water depths in contam- Feeding on several invertebrates was also en- inated EPSs could be held at ?2m to reduce hanced by numerical abundance. Rotifers con- attractiveness of the areas as foraging sites to sumed by Shovelersduring this study may be the shallow water feeding waterfowl and shorebirds. smallest foods consumed by any North Ameri- can waterfowl. Other filter-feeding anatids such ACKNOWLEDGMENTS as the Pink-eared Duck (Malacorhynchos mem- branaceus)can feed efficiently on foods as small We thank T. L. Barber and H. N. Coe for field assis- as 110 pm (Crome 198 5); rotifers (Keratellu) con- tance;A. E. Williamson for typing; J. M. Hicks and D. M. Mushet for nrenaring fiaures:T. L. Shaffer for sta- sumed by Shovelers in this study averaged only tistical analysis; and D.-A.-Barnum, W. L. Hohman, about 100 Km (Hutchinson 1967). Shovelers ad- G. L. Krapu, G. K. Rose, J. P. Skorupa, and G. A. just their lamellae spacing to facilitate straining Swansonfor manuscript review. This is Oregon State WATERFOWL FEEDING ECOLOGY 589

University Agricultural Experiment Station technical weights of some common macrobenthic inverte- paper #9 159. brates. Trans. Am. Fish. Sot. 101:743-746. HUTCHMSON,G. E. 1967. A treatise on , Vol. 2: introduction to lake biology and limno- LITERATURE CITED olankton. John Wilev and Sons. New York. NY. BARNUM,D. A., ANDN. H. EULISS,JR. 1991. Impacts KAH~RL,W. L. [ED.] 1979: The California water’atlas. of changingirrigation practiceson waterfowl hab- California Governor’s Office of Planning and Re- itat use in the southern San Joaquin Valley, Cal- search,Los Altos, CA. ifornia. Calif. Fish Game 77:l l-22. KRAPU, G. L. 1979. Nutrition of female dabbling BARNUM,D. A., AND D. S. GILMER. 1988. Selenium ducks during reproduction, p. 59-70. In T. A. levels in biota from irrigation drainwater im- Bookhout [ed.], Waterfowl and wetlands-an in- poundmentsin the San JoaquinValley, California. tegrated review. Proc. 1977 Symp., Northcent. Lake Reservoir Manage. 4: 18l-l 86. Sect., The Wildl. Sot., Madison, WI. BELLROSE,F. C. 1980. Ducks, geese and swans of KRAPU,G. L. 1981. The role of nutrient reservesin North America, 3rd ed. StackpoleBooks, Harris- mallard reproduction. Auk 98:29-38. burg, PA. MARTIN, A. C., AND W. D. BARKLEY. 1961. Seed identification manual. Univ. Calif. Press, Berke- CARNEY,S. G. 1964. Preliminary keys to waterfowl ley, CA. age and sex identification by means of wing plum- MERRITT, R. W., AND K. W. CUMMINS. 1984. An age. U.S. Fish Wildl. Serv. Spec. Sci. Rep. Wildl. introduction to the aquatic insectsof North Amer- 82. ica, 2nd ed. Kendall-Hunt Publ. Co., Dubuque, COE,H. N. 1990. Use of wetlands in the Tulare Lake IA. Basinby wintering ducks.M.S. thesis.Oregon State Univ., Corvallis, OR. MILLER,M. R. 1987. Fall and winter foods of North- CONNELLY,D. P., ANDD. L. CHESEMORE.1980. Food em Pintails on three northern California refuges. J. Wildl. Mange. 51:403412. habits of Pintails, Anas acuta, wintering on sea- sonally flooded wetlands in the northern San Joa- OHLENDORF,H. M., D. J. HOFFMAN,M. K. SAIKI, AND quin Valley, California. Calif. Fish Game 66:233- T. W. ALDRICH. 1986a. Embryonic mortality and 237. abnormalities of aquatic birds: apparent impacts by selenium from irrigation drainwater. Sci. Tot. CROME,F.H.J. 1985. An experimental investigation Environ. 52~49-63. of filter feeding on by some special- OHLENWRF,H. M., R. L. HOTHEM,C. M. BUNCK,T. ized waterfowl. Aust. J. Zool. 33~849-862. W. ALDRICH.AND J. F. MOORE. 1986b. Rela- EULISS,N. H., JR. 1984. The feedingecology ofPintai1 tionships between selenium concentrations and and Green-winged Teal wintering on Kern Na- avian reproduction. Trans. N. Am. Wildl. Nat. tional Wildlife Refuge.M.S. thesis.Humboldt State Resour. Conf. 5 1:330-342. Univ., Arcata, CA. PENNAK,R. W. 1978. Fresh-water invertebrates of EULISS,N. H., JR. 1989. Assessment of drainwater the United States, 2nd ed. John Wiley and Sons, evaporation ponds as waterfowl habitat in the San New York, NY. JoaquinValley, California. Ph.D. diss.Oregon State PRESSER,T. S., ANDI. BARNES.1985. Dissolved con- Univ., Corvallis, OR. stituentsincluding selenium in water in the vicin- EULISS,N. H., JR., ANDS. W. HARRIS. 1987. Feeding itv of KestersonNational Wildlife Refugeand West ecology of Northern Pintails and Green-winged Grasslands, Fresno and Merced cot&ties, Cali- Teal wintering in California. J. Wildl. Mange. 5 1: fornia. U.S. GeologicalSurvey, Wat. Res. Inv. Rep. 724-732. 85-4220, Menlo Park, CA. EULISS,N. H., JR., R. L. JARVIS,AND D. S. GILMER. SIEGFRIED,W. 1973. Summer food and feeding of the 1989. Carbonate deposition on tail feathers of Ruddy Duck in Manitoba. Can. J. Zool. 5 1:1293- Ruddy Ducks using evaporation ponds. Condor 1297. 91:803-807. SWANSON,G. A. 1977. Self-cleaning screen for pro- GILMER,D. S., M. R. MILLER,R. D. BAUER,AND J. R. cessingbenthic samples.Prog. Fish-Cult. 39: 177- LEDONNE. 1982. California’s Central Valley 178. wintering waterfowl: concerns and challenges. SWANSON,G. A. 1987a. A simple lightweight core Trans. N. Am. Wildl. Nat. Resour. Conf. 47:44 l- samplerfor quantitatingwaterfowl foods. J. Wildl. 452. Manage. 42~426428. GRODHAUS, G. 1967. Identification of chironomid SWANSON,G. A. 1978b. A water column sampler for midges commonly associatedwith waste stabili- invertebrates in shallow wetlands. J. Wildl. Man- zation lagoonsin California. Calif. Vector Views age. 42~670-671. 14:1-12. SWANSON, G. A., AND J. C. BARTONEK. 1970. Bias HEITMEYER,M. E. 1988. Protein costsof the prebasic associatedwith food analysis in gizzards of blue- molt of female mallards. Condor 90:263-266. winged teal. J. Wildl. Manage. 34:739-746. HOPPE,R. T., L. M. SMITH,AND D. B. WESTER. 1986. SWANSON,G. A., G. L. tiu, J. C. BARTONEK,J. R. Foods of wintering diving ducks in South Caro- SERIE,AND D. H. JOHNSON. 1974. Advantages lina. J. Field Omithol. 57: 126-l 34. in mathematically weighting waterfowl food hab- HOWMILLER,R. P. 1972. Effects of preservatives on its data. J. Wildl. Manage. 38:302-307. 590 N. H. EULISS, JR., R. L. JARVIS AND D. S. GILMER

TAMISIER, A. 1978/79. The functional units of win- U.S. FISHAND WILDLIFE SERVICE.1978. Concept plan tering ducks:a spatial integration of their comfort for waterfowl wintering habitat preservation- and feeding requirements. Verh. Omithol. Ges. Central Valley, California. U.S. Fish Wildl. Serv., Bayem 23~229-238. Portland, OR. TIETJE,W. D., AND J. G. TEER. 1988. Winter body ZWEERS,G. 1980. Experimental functional analysis condition of Northern Shovelerson freshwaterand and formulation of causalmodels, p. 195-20 1. In saline habitats, p. 353-376. In M. W. Weller [ed.], R. Nijhring [ed.], Proc. XVII Int. Om. Congr., Waterfowl in winter. Univ. Minn. Press, Minne- Berlin. apolis.