Female Hooded Merganser Body Mass During Nesting ’

220 SHORT COMMUNICATIONS

The Condor 99:220-224 0 The Cooper Ornithological Society 1997

FEMALE HOODED MERGANSER BODY MASS DURING NESTING ’

MICHAEL C. ZICUS Minnesota Department of Natural Resources, Wetland Wildlife Populations and Research Group, 102 23rd St., Berni@, MN 56601

Abstract: Body mass of female Hooded Mergan- males in relation to the date incubation started, size sers (Lophodytescucullatus) nesting in widely dis- of the incubated clutch, and number of young hatched persed and newly erected wooden nest boxes in also was assessed. northcentral Minnesota was measured in 1982-1985. Median body mass during egg-laying was 635 g. Fe- METHODS male mass during incubation varied significantly STUDY AREA among years, but decreased monotonically at the same rate (1 .O g day-‘) each year. Mass at the end The study was conducted in northcentral Minnesota. The area contained lakes and wetlands in an east-west of incubation (519494 g) was 5.7 to 6.0% less than when incubation began. There was no indication that morainic ridge with a narrow sandy plain extending to females having the greatest body mass began nest- the north (see Zicus 1990 for a more detailed descrip- tion). Open-water ponds supporting little emergent ing earliest. However, females with the greatest body mass incubated the largest clutches and hatched the shoreline vegetation were the predominate wetland type. Wetland densities in the moraine averaged 7-8 most young. Comparison of the observed body mass- basins km-*. clutch size relationship with one assumed to exist in The hilly moraine was almost entirely the absence of intraspecific brood parasitism indi- forested by aspen (Pop&us spp.) with lesser amounts cated that more parasitic eggs were laid in nests in- of birch (Bet&a spp.) and basswood (Tilia ameri- cubated by heavier females. cana). Pines (Pinus spp.), particularly jack pine (P. banksiana), dominated the sand plain. Key words: Hooded Mergansers, Lophodytes cu- cullatus, incubation body mass, brood parasitism. PROCEDURE Twice-weekly checks of wooden nest boxes began in mid- to late March each year. When an egg was found, nests were reexamined weekly. At each visit, clutch INTRODUCTION size was noted and several eggs were candled (Weller Changes in the body mass of females during incuba- 1956). As hatch day approached, nests were visited tion have been studied for many waterfowl (see more frequently to determine exact hatch dates. Fe- review in Afton and Paulus 1992). However, within- males, when present during a nest-box check, were species variability of body mass change during incu- captured, weighed, and leg-banded. Body mass was bation is still poorly understood for most species, par- determined to the nearest 5 g and was not scaled to ticularly the degree to which body mass responds to account for variable female skeletal size. varying environmental conditions. Mass change of A variety of information was used to determine incubating Common Goldeneyes (Bucephala clan- nesting chronology. For nests that were visited more gula) differed substantially from Ontario to Minne- than once during laying, the dates the first and last sota (Mallory and Weatherhead 1993, Zicus and eggs were laid were estimated by assuming an average Riggs 1996) and even among lakes and years in the laying rate of 1 egg every 2 days (Dugger et al. 1994). same localitv (Zicus and Riggs 1996). Similar com- Intraspecific brood parasitism was common (Zicus parative studies are needed fiy other species. Hooded 1990), and nests often received > 1 Hooded Mergan- Mergansers (Lophodytescucullatus) nest from north- ser egg in 2 days. For these nests and those that were em Florida and New Mexico northward to at least examined once during laying, incubation starts were southern Canada (Dugger et al. 1994). Thus, they are estimated by backdating from known hatching dates. likely to encounter a wide range of environmental I assumed a 32-day incubation period (Zicus 1990, conditions that could influence foraging patterns, nu- Dugger et al. 1994) when candling eggs and back- trient acquisition, and ultimately incubation body dating nests. Clutch size included parasitic and non- mass change and reproduction. parasitic merganser eggs as there was no reliable way Body mass change was measured during nesting to distinguish the two. Eggs laid as a result of brood for female Hooded Mergansers in Minnesota. Yearly parasitism by Wood Ducks (Aix sponsa)and Common differences and the effect of incubation stage on fe- Goldeneyes were not included in the analysis. Repro- male body mass were examined. Body mass of fe- ductive stage for egg-laying females was defined rela- tive to the start of incubation, and that of incubating females was referenced to the departure of young ‘Received 8 March 1996. Accepted 27 August 1996. from nests. This differs from the convention often SHORT COMMUNICATIONS 221 used for incubating females, but it is preferable be 575 7 - 1982 cause it allows more meaningful comparison among corresponding days (Zicus et al. 1995). Many females were weighed more than once during incubation, so their mass change was investigated using a generalized linear mixed-model (GLMM) with- maximum- likelihood estimators (PROC MIXED: SAS Institute Inc. 1992). This aDDrOaCh allows measurements on subjects to be repeaied’ within and across years. Dependencies among repeated measures are modeled explicitly and ensuing tests are adjusted for this dependence based on the underlying 475 1 III 1, I I I I, I I I, I ( I I ( I II,, I covariance structure (Laird and Ware 1982, Ware 0 5 10 15 20 25 30 1985). A first-order autoregressive covariance struc- Days Before Departure From Nest ture was determined to be optimal for the models FIGURE. 1. Maximum likelihood estimates of lin- (Jennrich and Schluchter 1986). I modeled the effect ear trend in female Hooded Merganser body mass of year, linear, quadratic, and cubic effects of incu- during-incubation in northcentral Minnesota, 1982- bation day, and day of incubation interactions with 1985. p is the common slope for all years. year on female mass. When interactions were not sig- nificant (o = 0.05), a reduced model was used. Si- multaneous paired comparisons were made using a nest. Females were weighed repeatedly during nest- Bonferroni adjustment to pairwise differences in the ing from one to five times each year. This resulted time-adjusted means (Dobson 1990). in 100 weights at known points during incubation. Nest initiation date, date incubation started, incu- Two females were weighed in four years, one in three bated clutch size, and number of hatched young were years, eight in two years, and 30 in only one year. regressed on female body mass residuals (PROC As a result, the analysis included 57 within-year REG: SAS Institute Inc. 1990). In this analvsis. onlv time-series from 1982 to 1985. During the study, 13, the residual from the first measurement of a females 13, 24, and 7 females were weighed each year, re- body mass in each year was included. Females whose spectively. weights were associated with positive residuals were assumed to be heavier than the typical female and FEMALE BODY MASS those with a negative residual were assumed to be Egg-laying females were weighed from 1 to 20 days lighter. Years were pooled in the analysis because before they started incubating (median = 3 days). yearly nesting chronology was similar (Zicus 1990). Weights ranged from 570 to 720 g (median = 635 g). Confidence intervals (95%) for the percent of all The GLMM was fit to measurements of mass for incubated eggs and hatched ducklings that occurred incubating females that successfully incubated in nests incubated by females with positive body clutches or had their nests destroyed during incuba- mass residuals were computed using the normal ap- tion by predators. Neither the cubic nor the quadratic proximation for binomial proportions (Steel and Tor- model ftt the data. For the linear model, there was rie 1980). no day of incubation X year interaction (F,,, = I also compared the relationship between body mass 0.43, P = 0.74) indicating slopes were similarin’ all residuals and total clutch size with a presumed rela- years. However, both day of incubation (F, F4 = tionship that might have existed in the absence of 36.68, P < 0.001) and year (F3 54 = 4.53, P = -6.607) brood parasitism. Clutch size, in the absence of brood influenced bodv mass sianificantlv (Fig. 1). The parasitism, might be related to female mass in three model indicated that fernares lost approxrmately 5.7 ways: (1) number of eggs could be related positively to 6.0 percent of their body mass during incubation to mass, (2) number of eggs could be independent of (day 32 to day 1). mass, or (3) number of eggs could be related inversely The GLMM also estimated that females weighed to mass. Egg-laying potential in Mallards (Anasphty- from 493 to 518 g when they departed the nest rhynchos) is positively related to spring body mass with the brood (model intercepts) depending on the (Johnson et al. 1987); thus, I assumed a positive linear year. Unadjusted median body mass for females relationship between merganser clutch size and fe- ready to leave the nest with hatched young was male mass. The slope of the presumed relationship 480 g (range = 470-540). was established by further assuming that females with REPRODUCTIVE ATTRIBUTES the smallest observed body-mass residuals laid the smallest clutches (5 eggs) and that those with the larg- Model residuals for female body mass were not cor- est residuals laid the largest clutches (13 eggs). Five- related with the date nest initiation began (r = 0.11, egg clutches were the smallest incubated in the popu- IZ = 20, P = 0.65) and were only weakly and nega- lation and 13 eggs is likely near the maximum that a tively correlated with the date that a female started female can lay in one clutch (Zicus 1990). incubating a nest (r = -0.29, IZ = 58, P = 0.03). In addition, -incubated clutch size was not correlated with nest initiation date (r = -0.29. IZ = 20, P = RESULTS 0.22) and only weakly and negatively correlated with Ten females were weighed during egg-laying, 42 dur- the date incubation started (r = -0.29, n = 58, P = ing incubation, and seven with hatched young in the 0.03). Apparently, egg-laying did not begin earlier in SHORT COMMUNICATIONS

- Observed clutch size - Observed hatched eggs vs. body mass , “8. body mass I r = 0.64 . r = 0.60 P = 0.000, I P = O.OWl n = 58 . /l=48 ...... p ...... I

I I ’ I ’ I I I I I -80 -60 -40 -20 0 20 40 60 80 -80 -60 -40 -20 0 20 40 60 80

Departure From Predicted Body Mass (g) Departure From Predicted Body Mass (g)

FIGURE. 2. Correlation of linear model residuals FIGURE. 3. Correlation of linear model residuals for body mass with number of Hooded Merganser for body mass with number of hatched Hooded Mer- eggs in the clutch in northcentral Minnesota, 1982- ganser eggs in the clutch in northcentral Minnesota, 1985. Shaded area represents the presumed range in 1982-1985. unparasitized clutch size. Dotted line represents the relationship between body mass and clutch size that was assumed to exist in the absence of brood para- body size, this was expected. Afton and Paulus sitism. (1992) predicted that females of small-bodied species would rely more on exogenous nutrient sources, ex- hibit lower incubation constancy, and lose propor- nests incubated by heavier females than it did in nests tionately less mass than would larger-bodied females. incubated by lighter females; however, heavier fe- Although they are small-bodied, Hooded Mergansers males may have begun incubation earlier. Further, may still rely, at least partially, on stored nutrients to there was no compelling evidence that nests with initiate nesting. Nesting starts in Minnesota at a time larger clutches were established earlier in the season when wetlands and small lakes are just beginning to than those with smaller clutches. become ice-free (Zicus 1990). This chronology dif- Body mass residuals of incubating females were fers from other small-bodied ducks such as Blue- correlated (r = 0.64, n = 58, P < 0.001) with the winged Teal (Anus discors) and Northern Shovelers total number of Hooded Merganser eggs in the clutch (A. clypeatu) that forage intensively after arrival on (Fig. 2) and the number of hatched merganser duck- the nesting grounds before egg-laying begins. Low lings (I = 0.60, n = 48, P < 0.001) (Fig. 3). These egg-laying rates by Hooded Mergansers (Zicus 1990, positive correlations suggested that heavier females Dugger et al. 1994) may be one factor that allows incubated larger clutches and hatched more young females to complete a clutch early in the season while than lighter females. When nests were pooled across relying on primarily exogenous nutrient sources. all years, females with positive residuals incubated Hooded Merganser mass loss during incubation 60.5 percent (95% confidence interval = 57.1 to might be greater in other locations although there are 63.9%) of all incubated merganser eggs and hatched too few data to know. Mallory et al. (1993) estimated 64.5 percent (95% confidence interval = 60.7 to that nine females lost >3 g day- ’ in Ontario based 68.3%) of all ducklings. on single measurements taken during incubation and The observed relationship between body mass re- pooled over a number of years. This rate suggests a siduals and incubated clutch size (Fig. 2) had an es- loss of about 17 percent of female body mass during timated slope of 0.10 (95% confidence interval = incubation, assuming an incubation period of 32 0.07 to 0.13). This was significantly greater than the days. Kennamer et al. (1988) reported on two South slope (0.06) of the relationship between female body Carolina females which were measured once in early mass and the number of eggs laid in a clutch that incubation and again at hatching; mass change aver- was assumed to exist in the absence of brood para- aged about 13 percent assuming a 32.day incubation sitism. This indicated that more intraspecific brood period (calculated from Kennamer et al. 1988). parasitism occurred in nests incubated by heavier fe- Many factors, including nesting chronology, females than in those incubated by lighter birds. male age, body condition, and access to food, have been related to the reproductive output of waterfowl. DISCUSSION For cavity nesting species, the extent and magnitude Female Hooded Mergansers lost little body mass in of intraspecific brood parasitism also can be impor- the course of incubation. Considering their small tant (Jones and Leopold 1967, Morse and Wight SHORT COMMUNICATIONS 223

1969, Heusmann et al. 1980, Semel et al. 1988). Vari- LITERATURE CITED ous hypotheses about the significance of intraspecific brood parasitism have been proposed (see review in AFTON, A. D., AND S. L. PAULUS. 1992. Incubation Sayler 1992), but reasons why specific nests are para- and brood care, p. 62-108. In B. D. J. Batt, A. sitized remain poorly understood. D. Afton, M. G. Anderson, C. D. Ankney, D. H. Several mechanisms consistent with a number of Johnson, J. A. Kadlec, and G. L. Krapu reds.], hypotheses have been used to explain the selection Ecology and management of breeding water- of specific nests by females laying eggs parasiti- fowl. Univ. Minnesota Press, Minneapolis, MN. cally. Semel et al. (1988) suggested that highly CLAWSON,R. L., G. W. HARTMAN, AND L. H. FREDRICK- visible nest structures erected in high densities were SON. 1979. Dump nesting in a Missouri Wood parasitized more by Wood Ducks. During this study, Duck population. J. Wildl. Manage. 43:347-355. nest structures available to Hooded Mergansers DOBSON, A. J. 1990. An introduction to generalized were newly erected and widely spaced in the linear models. Chapman and Hall, London. landscape (Zicus 1990), likely minimizing the ef- DUGGER, B. D., K. M. DUGGER, AND L. H. FREDRICK- fects of structure placement on intraspecific brood SON. 1994. Hooded Merganser (Lophodytescu- parasitism. Nonetheless, brood parasitism was cullatus), No. 98. In A. Poole and F. Gill [eds.], prevalent in all years. Both Redheads (Aythya The Birds of North America. The Academy of americana) (Weller 1959, Sayler 1985) and Wood Natural Sciences, Philadelphia and The Ameri- Ducks (Grice and Rogers 196.5, Clawson et al. can Ornithologists ’ Union, Washington, DC. 1979) have been reported to lay more parasitic eggs GRICE, D., AND J. P. ROGERS. 1965. The Wood Duck early in the nesting season. As a result, nests in Massachusetts. Final Ren. Fed. Aid Proi. No. established earlier would be parasitized more and W-19-R. MA. Div. Fish an; Game, Bosto;. would likely have the largest clutch sizes. In my HEUSMANN, H. W., R. BELLVILLE, AND R. G. BURRELL. study, size of the incubated clutch was most 1980. Further observations on dump nesting by strongly correlated with the body mass of the Wood Ducks. J. Wildl. Manage. 44:908-915. incubating female and proportionately more para- JENNRICH,R. I., AND M. D. SCHLUCHTER.1986. Unbal- sitic eggs occurred in nests incubated by the heavy anced repeated-measures models with structured females. I observed no compelling evidence that covariance matrices. Biometrics 42:805-820. nests with more eggs or those incubated by the JOHNSON,D. H., D. W. SPARLING,AND L. M. COWAR- heaviest females were initiated earlier in the season DIN. 1987. A model of the productivity of the than other nests. Mallard Duck. Ecol. Modelling 38:257-275. The extent to which heavier females incubated pro- JONES,R. E., AND A. S. LEOPOLD. 1967. Nesting in- portionately more parasitic eggs was estimated con- terference in a dense population of Wood Dicks. servatively. The other two assumptions regarding the J. Wildl. Manage. 31:221-228. relationship between female body mass and clutch KENNAMER,R. A., w. F. HARVEY IV, AND G. R. HEPP. size in the absence of brood parasitism that could 1988. Notes on Hooded Merganser nests in the have been considered would have increased the dif- coastal slain of South Carolina. Wilson Bull. ference between the slopes of the observed and as- 100:6861-688 sumed relationships. This also would be true if the LAIRD, N. M., AND J. H. WARE. 1982. Random-effects minimum clutch produced by a single female was >5 models for longitudinal studies. Biometrics eggs or the maximum clutch that a single female can 38:963-974. lay was < 13 eggs. The observation that females with MORSE, T. E., AND H. M. WIGHT. 1969. Dump nest- the greatest body mass tended to incubate proportion- ing and its effect on production in Wood Ducks. ately more parasitic eggs is consistent with at least J. Wildl. Manage. 33:284-293. two hypotheses proposed to explain brood parasitism MALLORY, M. L., AND P. .I. WEATHERHEAD.1993. In- (Sayler 1992). Perhaps two or more females begin cubation rhythms and mass loss of Common laying eggs in the same nest, and the heaviest female Goldeneyes. Condor 95:849-859. displaces other females and incubates the clutch. Fe- MALLORY, M. L., H. G. LUMSDEN,AND R. A. WALTON. male mass and the ability to displace other females 1993. Nesting habits of Hooded Mergansers could be related to age, reproductive experience, Mergus cucullatusin northeastern Ontario. Wild- physical condition, or some combination. Alterna- fowl 44:101-107. tively, this could represent a situation wherein SAS INSTITUTEINC. 1990. SAS/STAT users’ guide, heavier females (i.e., mothers) were parasitized by version 6, 4th ed., vol. 2. SAS Institute Inc., lighter daughters. Clearly, much remains to be Cary, NC. learned about brood parasitism in Hooded Mergan- SAS INSTITUTE INC. 1992. SAS technical report sers. p-229: SAS/STAT software changes and en- hancements, release 6.07. SAS Institute Inc., I thank S. K. Hennes, S. J. Maxson, and T. L. Peter- Cary, NC. son of the Minnesota Department of Natural Re- SAYLER,R. D. 1985. Brood parasitism and reproduc- sources for help with field work. R. L. Naplin and tion of Canvasbacks and Redheads on the Delta T. A. Stursa erected some of the nest boxes included Marsh. Ph.D. diss. Univ. North Dakota, Grand in the study. Earlier drafts of the manuscript were im- Forks, ND. proved through careful review and thoughtful com- SAYLER,R. D. 1992. Ecology and evolution of brood ments by D. P. Rave, S. J. Maxson, and M. R. Riggs. parasitism in waterfowl, p. 290-322. In B. D. J. M. R. Riggs also assisted with the statistical analyses. Batt, A. D. Afton, M. G. Anderson, C. D. 224 SHORT COMMUNICATIONS

Ankney, D. H. Johnson, J. A. Kadlec, and G. L. WELLER, M. W. 1959. Parasitic egg-laying in Krapu [eds.], Ecology and management of the Redhead (Aythya americana) and other breeding waterfowl. Univ. Minnesota Press, North American Anatidae. Ecol. Monogr. Minneapolis, MN. 29:333-365. SEMEL, B., P. W. SHERMAN, AND S. M. BYERS. 1988. Zrcus, M. C. 1990. Nesting biology of Hooded Mer- Effects of brood parasitism and nest-box place- gansers using nest boxes. J. Wildl. Manage. ment on Wood Duck breeding ecology. Condor 54:637-643. 90:92&930. ZICUS, M. C., S. K. HENNES,AND M. R. RIGGS. 1995. STEEL, R. G. D., AND J. H. TORRIE. 1980. Principles Common Goldeneye nest attendance patterns. and procedures of statistics. McGraw-Hill, New Condor 97:461-472. York. ZICUS, M. C., AND M. R. RIGGS. 1996. Changes in WARE, J. H. 1985. Linear models for the analysis of body mass of female Common Goldeneyes dur- longitudinal studies. Am. Stat. 39:95-101. ing nesting and brood rearing. Wilson Bull. WELLER,M. W. 1956. A simple field candler for wa- 108:61-71. terfowl eggs. J. Wildl. Manage. 20:111-113.

The Condor 99~224-228 0 The Cooper Ornithological Society 1997

VARIATION IN INCUBATION PERIODS AND EGG METABOLISM IN MALLARDS: INTRINSIC MECHANISMS TO PROMOTE HATCH SYNCHRONY ’

MARGARETC. MACCLUSKIE Department of Biology and Wildlife, Institute of Arctic Biology, University of Alaska Fairbanks, Fairbanks, Alaska 99775, e-mail: [email protected]

PAUL L. FLINT National Biological Service, Alaska Science Center, IO11 East Tudor Road, Anchorage, Alaska 99503

JAMESS. SEDINGER Institute of Arctic Biology, Department of Biology and Wildlife, University of Alaska Fairbanks, Fairbanks, Alaska 99775

Abstract: We investigated factors affecting incu- chrony at hatching is less than that observed at the bation time and metabolic rates of Mallard (Anus end of laying (Caldwell and Comwell 1975, Afton platyrhynchos) eggs incubated under constant envi- 1979, Cargill and Cooke 198 l), suggesting that some ronmental conditions. Time required to reach the star- mechanisms exist to synchronize hatch (Afton 1979, pipped stage of hatch varied significantly among fe- Davies and Cooke 1983). Such mechanisms could in- males, but not with laying sequence or egg size. clude developmental retardation of eggs laid early in Metabolic rate of eggs varied positively with posi- the laying sequence or developmental acceleration of tion in the laying sequence and tended to vary among eggs laid after the onset of incubation (Davies and females. Metabolic rate did not vary with egg vol- Cooke 1983). Regardless of how such mechanisms ume or incubation length. Our results indicate meta- work, eggs within clutches must have variable incu- bolic rate may act as one synchronization mechanism bation lengths. for hatch. The role of maternal effects in develop- Synchronization mechanisms could be either ex- ment time should be considered in subsequent stud- trinsic to the egg or intrinsic to the egg. Extrinsic ies of incubation time in ducks. mechanisms could include behavior of the parent. For example, eggs in the center of the nest have higher Key words: Mallard, Anas platyrhynchos, me- temperatures (Caldwell and Comwell 1975), thus fe- tabolism, egg, incubation, incubation length. males may alter development time of specific eggs by regulating egg position in the nest. Also, vocal stimulation decreases time to hatching (Vince 1966, INTRODUCTION Orcutt and Orcutt 1976) and may accelerate devel- Waterfowl typically begin incubation before the opment of later laid eggs (Davies and Cooke 1983). clutch is complete (Caldwell and Comwell 1975, Intrinsic factors include inherent properties of the Cooper 1978, Afton and Paulus 1992). which results egg, such as egg composition, egg size, or physiol- in developmental asynchrony of eggswithin clutches ogy. Smaller eggs require less incubation, both at the end of laying (Caldwell and Comwell 1975, among and within species (Worth 1940, Rahn and Ar Afton 1979, Kennamer et al. 1990). However, asyn- 1974, Arnold 1993). Martin and Arnold (1991) found