UBC Faculty Research and Publications

Winter body mass and over-ocean flocking as components of danger management by Pacific dunlins Ydenberg, Ronald C; Dekker, Dick; Kaiser, Gary; Shepherd, Philippa C; Ogden, Evans L; Rickards, Karen; Lank, David B Jan 21, 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-12898_2009_Article_137.pdf [ 1.18MB ]
Metadata
JSON: 52383-1.0223421.json
JSON-LD: 52383-1.0223421-ld.json
RDF/XML (Pretty): 52383-1.0223421-rdf.xml
RDF/JSON: 52383-1.0223421-rdf.json
Turtle: 52383-1.0223421-turtle.txt
N-Triples: 52383-1.0223421-rdf-ntriples.txt
Original Record: 52383-1.0223421-source.json
Full Text
52383-1.0223421-fulltext.txt
Citation
52383-1.0223421.ris

Full Text

RESEARCH ARTICLE Open AccessWinter body mass and over-ocean flocking ascomponents of danger management by PacificdunlinsRonald C Ydenberg1*, Dick Dekker2, Gary Kaiser3, Philippa CF Shepherd4, Lesley Evans Ogden5, Karen Rickards1,David B Lank1AbstractBackground: We compared records of the body mass and roosting behavior of Pacific dunlins (Calidris alpinapacifica) wintering on the Fraser River estuary in southwest British Columbia between the 1970s and the 1990s.‘Over-ocean flocking’ is a relatively safe but energetically-expensive alternative to roosting during the high tideperiod. Fat stores offer protection against starvation, but are a liability in escape performance, and increase flightcosts. Peregrine falcons (Falco peregrinus) were scarce on the Fraser River estuary in the 1970s, but their numbershave since recovered, and they prey heavily on dunlins. The increase has altered the balance between predationand starvation risks for dunlins, and thus how dunlins regulate roosting behavior and body mass to manage thedanger. We therefore predicted an increase in the frequency of over-ocean flocking as well as a decrease in theamount of fat carried by dunlins over these decades.Results: Historical observations indicate that over-ocean flocking of dunlins was rare prior to the mid-1990s andbecame common thereafter. Residual body masses of dunlins were higher in the 1970s, with the greatestdifference between the decades coinciding with peak peregrine abundance in October, and shrinking over thecourse of winter as falcon seasonal abundance declines. Whole-body fat content of dunlins was lower in the 1990s,and accounted for most of the change in body mass.Conclusions: Pacific dunlins appear to manage danger in a complex manner that involves adjustments both in fatreserves and roosting behavior. We discuss reasons why over-ocean flocking has apparently become morecommon on the Fraser estuary than at other dunlin wintering sites.BackgroundStarvation and predation risks are both important in theecology of wintering shorebirds [1-3], and stored bodyfat appears central to managing these risks. It providesinsurance against starvation during periods of foodshortage, but increases energetic flight expenditures andreduces predator escape performance [4,5]. Of coursethe fat reserve must also be built up in the first place,which requires extra foraging and hence predatorexposure.The trade-off theory of fat storage [6,7] holds that thesurvival-maximizing level is set by its contrasting effectson the probabilities of mortality by starvation and bydepredation. Many temperate shorebird species forexample, enlarge the fat reserve in mid-winter, increas-ing protection against the higher incidence of foodshortage at that time of year [8]. Decreases in theamount of mid-winter fat over recent decades, docu-mented for both shorebirds [9] and forest birds [10,11],have been linked to the population recovery of raptorssince DDT was banned in the 1970s. The interpretationis that increased raptor populations have heightenedpredation danger, so prey species have lowered the levelof winter fat reserves to increase their predator escapeability, at the expense of diminishing their ability towithstand starvation.Predation danger effects on roosting behavior havealso begun to be recognized. Shorebirds are sensitive to* Correspondence: ydenberg@sfu.ca1Centre for Wildlife Ecology, Simon Fraser University, Burnaby, BC V5A 1S6,CanadaYdenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1© 2010 Ydenberg et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.the safety features of roost sites, and may travel longdistances to utilize safe roosts [12-14]. They sometimesspend part of the high tide period in flight out overopen water, rather than roosting on the ground. Thishas been termed ‘roosting on the wing’ [15], ‘high tideflight’ [16], ‘airborne roosting’ [17], or ‘over-ocean flock-ing’ [18]. We use the latter term here. Over-ocean flock-ing has been described most extensively in BritishColumbia [[19], see especially [18]], but has also beenobserved in Britain [15], Washington State [16], Califor-nia [20], Germany [17] and The Netherlands [21].Though generally uncommon, over-ocean flockingappears to be a normal part of the behavioral repertoireof some shorebird species. A video sequence can bewatched at http://www.sfu.ca/biology/wildberg/species/wesa.htmlMost observers [16,17,20] attribute an antipredationfunction to over-ocean flocking. Prater [15] does notspecifically mention predators, but identifies disturbanceas the most important factor. Dekker [18] interprets itas a strategy to avoid stealth attacks by falcons andother raptors [22,23] to which shorebirds are vulnerablewhile roosting or foraging close to shorelines wherethere is cover from which raptors can attack. Dekker &Ydenberg [24] show that over-ocean flocking increasessafety from predators. It can be thought of as an energe-tically-expensive roosting option, analogous to expend-ing energy to fly to a distant but safer roost site.In this paper we compile historical information onover-ocean flocking and winter body mass of Pacificdunlins (Calidris alpina pacifica) on the Fraser Riverestuary. At this site, the abundance of peregrine falcons(Falco peregrinus) during autumn and winter has risensteadily since the early 1980s [25], and they are impor-tant predators of dunlins [24]. We therefore predictedthat the winter weight and roosting behavior of Pacificdunlins had changed over this period, shifting theemphasis from starvation avoidance (i.e. ‘traditional’ground-based roosting, high winter fat reserves) to pre-dation avoidance (i.e. over-ocean flocking, low winter fatreserves).ResultsHigh tide behaviorWe located a total of six studies (see Table 1) that madeobservations on dunlins on the Fraser estuary prior to1994. Participants in three of these studies neverrecorded over-ocean flocking in their notes, or recalledobserving it when later interviewed by us. Observers inthe three other studies others saw it once, or on occa-sion. Over-ocean flocking was recorded much moreoften by later observers. All five of the post-1994 obser-vers report it as a regular or frequent occurrence.Two experienced long-term observers documentedtheir first observations of over-ocean flocking in the mid1990s. Richard Swanston has for many years made regu-lar birding visits to our study area, and works on theferry that sails along the study area. He first noted over-ocean flocking by Pacific dunlins in March, 1996 (onthe ship’s radar) and reports (pers. comm.) that it hassince become increasingly common. Dr. Robert Butler(pers. comm.; he also made one of the pre-1994 studies[26]) first observed over-ocean flocking on April 26,1994. Butler never observed over-ocean flocking in twowinters of regular surveys [26], see Table 1] or on manyother visits to the Fraser estuary prior to 1994.During the extended set of observations made by DDin January, 2006, over-ocean flocking was frequent, andprolonged (Fig. 1). It occurred on 15/17 observationdays, with the only two days on which it did not occur(January 11 and 24) being, respectively, the only wind-less day, and the only day with continual heavy rain.Table 1 Studies of Pacific dunlin on the Fraser River estuary made in winter, from 1971 to the present.Winter Observer Methods Over-ocean flocking?1971/72 R. Drent Surveys/observations [42] No1976 P. Major dunlin 3D flock structure [43] No1977/80 G. Kaiser frequent mist-netting [27] (pers. comm.) seen once1979/80 K. Fry regular high tide counts 27 surveys; [19] seen on occasion during fall migration1981/84 A. Farr dunlin foraging ecology [44] observed when tide and wind very high1989/90 1990/91 R. Butler regular high tide counts, 29 surveys [26] No1995/98 P. Shepherd Ph.D. study [31] See frequently in all yearsall years R. Swanston many birding visits, plus radar from ferry first noted March 1996, regular occurrence since (pers. comm.)all years R. Butler many visits first noted April 26, 1994 (pers. comm.)1997/2000 L. Evans Ogden Ph.D. study [39] seen frequently in all years1994/2003 D. Dekker 152 observation days [24] seen regularly in all years2006 (Jan) D. Dekker detailed observations see Fig. 1 Seen 15/17 d, averaging 2.75 h2005/2007 Y. Zharikov dunlin feeding dispersion [45] seen frequently; (pers. comm.)’Over-ocean flocking?’ asks whether the observer describes this phenomenon in the report, or, during later interview, recalled seeing it.Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 2 of 11The duration of over-ocean flocking ranged from 1.5 -6.5 h, and the mean duration over all 17 days was 2.8 h,which is similar to Shepherd’s (2001a) estimate made (inwinter 1995/96, and spring 1998) using radio telemetryof 3.0 h spent in flight each day, the majority duringhigh tide periods. These observations also indicated thatover-ocean flocking did not take place at night.Winter body massAs described in the Methods, of the three winters ofmist-netting effort in the late 1970s, full data are avail-able only for 1978/79 [see [27]]. Monthly summariesremain for the preceding (1977/78) and following (1979/80) years; these are depicted in Fig. 2 with an analogousmonthly summary prepared from data for 1978/79. Theseasonal progression of body mass is similar in all threewinters, and clearly shows the pattern identified by Kai-ser & Gillingham [27]. Body mass is low upon migratoryarrival in October, but rises steeply and remains highfrom November through January. Body mass falls inFebruary and remains low until late in March, when itclimbs steeply prior to spring migration. This compari-son shows that the 1978/79 data are representative ofother years at that time.The 1978/79 body mass data are compared with thosefrom the 1990s in Fig. 3a; and decade-specific third-degree splines are shown in Fig. 3b. The spline fitted tothe 1970s sample closely matches the pattern describedabove, but in the 1990s sample the November - Decem-ber peak in mass has all but disappeared. The bodymass difference between the decades is greatest (~4 g)during the autumn, and shrinks over the course of thewinter until it disappears in the pre-spring migrationperiod.Mean culmen and winglengths differ slightly but sig-nificantly between decades, though in opposite direc-tions (culmen 0.83 mm [2.15%] shorter in 1990s sample;wing length 1.33 mm [1.09%] longer in 1990s). Asexpected [28], culmen is a statistical predictor of bodymass for both decades (p < 0.0001 for main culmeneffect, culmen by decade interaction p = 0.690), butwing length is a significant predictor of body mass onlyin the 1970s, when birds were relatively heavy (wing bydecade interaction, p < 0.0001). To compare the wintermasses of Pacific dunlins shown in Fig. 3, we fitted a b-spline to the pooled data. Controlling for culmen andwing length, residual masses are strongly and signifi-cantly lower in 1970s than in the 1990s. The overallmean difference is 2.4 g (t = 19.3, df = 1708, p <0.0001), though as noted above, the difference is biggestin November, and falls to zero or near-zero by lateMarch.Fig. 4 compares the whole-body fat content of 1990sdunlins with that reported by McEwan & Whitehead[29] for the 1970s. The 1990s data are too few for adefinitive assessment, but show that all the birds mea-sured in the 1990s carried less fat than the meansreported for the 1970s, and that the absolute differenceis about 2-4 g, which suggests that the change in bodymass between the decades can be accounted for largelyby a decline in the amount of fat carried.Figure 1 Over-ocean flocking by Pacific dunlins. Graphical summary of over-ocean flocking observations by DD made in January, 2006.Columns are successive days, with time given on the y-axis. The horizontal line in each column gives the time of high tide, and the shadedportion indicates the duration of over-ocean flocking. Further details in the text.Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 3 of 11Peregrine seasonal abundance patternAs documented in Ydenberg et al. [[25]; see Methods],peregrines were uncommon on the Fraser estuary priorto the ban on DDT in 1973. Matching the continentalpattern, their numbers began to rise strongly in the late1970s or early 1980s and have climbed steadily since[25]. Peregrines have a marked seasonal pattern of abun-dance on the Fraser estuary (Fig. 5), rising steeply dur-ing August to the annual peak during October, beforedeclining during the remainder of the winter to anannual low just before spring migration resumes. Pere-grines are relatively abundant during the spring migra-tion period (mid-April - mid-May), but few breed locallyand they are virtually absent after this time until theybegin to reappear in August.This pattern seems to have emerged as peregrinenumbers recovered. Monthly raptor surveys carried outin and around our study area during the winters of1970-1979 by the Vancouver Natural History Societysighted only 51 peregrines in 416 surveys [30]. Thedetection rate was steady and low from October toMarch, indicating that seasonal pattern described byLank et al. [1] was absent or greatly reduced prior toperegrine population recovery.DiscussionThe data presented here show that Pacific dunlins win-tering on the Fraser estuary in British Columbia havechanged aspects of their winter ecology over the pastfew decades. In the late 1970s they showed classic tem-perate shorebird winter behavior, with a regular routineof feeding at low tide and ground-based roosting at hightide, and a mid-winter peak in body mass, attributableto fat storage. After the mid-1990s, the mid-winter peakFigure 2 Winter body mass of Pacific dunlins, by month. Summary by month of Pacific dunlin mean mass (g) measured on the Fraser Riverestuary during the winters of 1977/78, 1978/79 and 1979/80, for females (above) and males (below). Overall, the seasonal pattern of weights inthe three years is similar. Samples contain only gender-assigned individuals, and thus exclude about 30% of individuals. Exact dates of capturevary somewhat between months and years. Samples sizes for individual entries range from 14 - 130; total sample size = 1883. Error bars are 95%confidence intervals.Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 4 of 11in body mass disappeared, and dunlins began to spendseveral hours during the high tide period in over-oceanflocking. We interpret these changes as an adaptiveresponse to greatly increased peregrine falcon presenceduring the winter since the 1970s.Table 1 summarizes the evidence that over-oceanflocking has become more frequent. For three reasons wefeel that we can discount the possibility that over-oceanflocking occurred as frequently during the early (pre1994) studies, but was missed because early observerswere unaware of its occurrence, while later observers hadbeen alerted to this phenomenon. (A) Save Zharikov,none of the later observers was aware of over-oceanflocking when they began their studies (pers. comm.).This is not surprising, as the descriptions of over-oceanflocking available at the time [15,16,19] were not inmainstream publications, and none devotes more than afew lines to the description. (B) None of the 56 sets ofsystematic high tide surveys documented during the win-ters of 1979/80 (n = 27; [19]) and 1989/90 and 1990/91(n = 29; [26]) records a low number of roosting dunlins,as would be expected if the birds were engaged in unseenover-ocean flocking. (C) Two observers with extensiveexperience predating 1994 (see Table 1) both documen-ted their first observations of over-ocean flocking in themid-1990s. We thus feel confident in concluding that onthe Fraser estuary this phenomenon has greatly increasedin frequency, duration, or both, since the late 1970s.The data presented here also confirm that in the late1970s dunlins were heavier than in the late 1990s, theFigure 3 Winter body mass of Pacific dunlins. a) Masses and b) predicted masses generated by third degree b-splines of dunlins captured onthe Fraser River Delta during the winters of 1978/79 (in black) and 1994 - 2000 (in grey). See methods for details.Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 5 of 11difference being greatest (~4 g) in November, andshrinking steadily over the course of the winter until thepre-spring migration period, when masses are the same.This convergence rules out the possibility that the bodymass difference is attributable to some systematic bias(e.g. scale calibration). Our measurements of total bodyfat are consistent with all or most of the body mass dif-ference between the decades being attributable to adecline in fat reserves.The slight though statistically significant differences inmean culmen length and mean wing length between thedecades are likely attributable to differences in details ofthe measurement (e.g. personnel, instruments, calibra-tion) or sampling (e.g. habitat) procedures (cf. [28]). In asample of Pacific dunlins collected on the Fraser Riverestuary in 1992-1995 (i.e. between the samples reportedhere; see [31]), means for both culmen (2.18%, relativeto 1970s sample) and winglength (1.59%) are slightlyFigure 4 Whole body fat content of Pacific dunlins. Distributions of whole body fat content of dunlins collected in the late 1970s (blacksymbols; from McEwan & Whitehead [29]) versus the 1990s (grey symbols). Error bars are standard deviations. Females are indicated by a circleand males by a triangle. Two individuals from the 1990s dataset could not be gender-assigned and are designated by a star. Details of the datasets are given in the Methods.Figure 5 Seasonal change in peregrine abundance. Seasonal index of peregrine abundance, based on the mean number of sightings duringstandardized daily surveys, grouped into successive 5-day periods, 1986 - 2000. The annual peak is reached during October, and declines duringthe course of the winter. A second peak follows in mid-April. Pacific dunlins’ winter residence on the Fraser estuary begins just after the autumnpeak, and ends just before the spring peak of the peregrine index. The data were collected by John Ireland, manager of the George C. ReifelMigratory Bird Sanctuary, located on the Fraser estuary. Based on Lank et al. [1].Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 6 of 11larger than either sample reported on here. This indi-cates that the differences between samples do not repre-sent an ongoing change in the size composition of thepopulation, and supports the interpretation that thechanges are attributable to minor procedural differences.The seasonal pattern of mass decline between the dec-ades shows a striking correlation with the seasonal pat-tern of peregrine occurrence on the Fraser estuary. Thebiggest mass drop coincides with the autumn peak inperegrine abundance, with the difference shrinking asthe peregrine index declines during the winter. Mass inthe decades is the same in March, when peregrines areat their annual (near zero) low. This correlation sup-ports our hypothesis that Pacific dunlins have shiftedemphasis from protection against starvation to protec-tion against predation. Piersma et al. [9] reached a simi-lar conclusion in their analysis of changes in the mid-winter masses of golden plovers in The Netherlands. Asthe data are correlational, we must be cautious aboutinferring causation, or ruling out contributing roles forany of the many other factors that must have changedat this location between these decades.Another hypothesis to explain the body mass reduc-tion is that starvation risk has declined over recent dec-ades as climate change reduced the severity of winterweather. To evaluate this we obtained weather recordsfor the period 1970 - 2007 from the National ClimateData and Information Archive of Environment Canada.In Table 2 we summarize the rate of change in averagedaily temperature, total precipitation, and maximumwind gust on the study area over the winter months.These data provide weak support at best for the ideathat winter weather has become less severe. Though thechange in mean daily temperature is slightly positive foreach winter month over this period, ranging from 0.01to 0.07°C y-1, neither the change in total monthly preci-pitation nor gustiness are consistent in direction acrossmonths, and the low r2 values in Table 2 indicate thatall three metrics are very noisy. In fact, the total netchange over the 37 year record is much smaller thanmost of the year to year changes recorded.Neither does the seasonal pattern of change in bodymass between the decades match very closely the pat-tern in those climate measures that do show evidence ofchange. The biggest seasonal mass change between the1970s and 1990s occurs in the autumn, with the differ-ence shrinking until March, when masses are the same.In contrast January shows the strongest rate of increasein temperature since 1970, followed by March. Changesin precipitation and wind match even less well. Overall,we feel that the reduction in winter body mass of Pacificdunlins is not well-explained by an hypothesis based onreduced winter severity. Neither is this hypothesis ableto explain why over-ocean flocking has becomecommonplace.We regard the changes in roosting behavior and bodymass as adaptive adjustments to increased danger. Otherinterpretations are that the body mass decline is a non-strategic consequence of extra flight induced by harass-ment from numerous peregrines, or that it is a strategicreduction of body mass made to reduce flight costsbecause flight time has increased for a reason unrelatedto predation danger. Differentiating between these com-peting hypotheses could be undertaken with an analysisof when strategic over-ocean flocking (or more gener-ally, roosting behavior) ought to occur. The first studyto make a strategic analysis of roost site choice is thatby Rogers et al. [13]. In their study area, distant roostsites required more travel than nearby roost sites, butbirds suffered fewer disturbances there. They accountedfor roost choice with a model that minimized totalenergy expenditure over the entire high water period,summing the energetic costs of both travel to the roostsite and time spent in flight while there. We agree withthis general approach, but feel that the choice is moreappropriately analyzed in terms of maximizing survivalthan in minimizing energy expenditure. Over-oceanflocking becomes worthwhile when it reduces the prob-ability of mortality [20], taking account of the extramortality that results from the extra foraging required.Several factors may dispose dunlins on the Fraserestuary to more prolonged and frequent over-oceanflocking than at other sites. First, the winter populationTable 2 Climate change during winter on the FraserestuaryMonth Daily mean temperature (°C) Total precipitation (mm)Mean Slope r2 Mean Slope r2November 6.07 0.03 0.06 180 0.97 0.02December 3.71 0.04 0.07 174 -0.80 0.02January 3.59 0.07 0.18 160 1.89 0.12February 4.78 0.01 0.01 112 -1.54 0.10March 6.64 0.03 0.09 114 0.33 0.01Month Maximum wind gust (km h-1)Mean Slope r2November 67.9 0.16 0.02December 69.5 0.21 0.02January 67.6 0.20 0.03February 62.6 -0.11 0.01March 68.5 -0.11 0.01Summary by month of daily mean temperature, total precipitation, andmaximum wind gust speed, during 1970 - 2007, from the National ClimateData and Information Archive of Environment Canada http://www.climate.weatheroffice.ec.gc.ca/, recorded at the Vancouver International Airport,located adjacent to the Fraser estuary. The table records for each wintermonth the overall mean, the slope of the regression on year (1970 to 2007),and the proportion of the variance explained (r2).Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 7 of 11of peregrines is high relative to other locales, and thedanger they pose to dunlins wintering there is particu-larly great, because kleptoparasitic competition fromBald Eagles (Haliaeetus leucocephalus) forces peregrinesto concentrate their foraging on dunlins instead ofducks, a favorite prey for peregrines elsewhere [32-34].Second, the nature of the Fraser River estuary is suchthat alternative coastal and inland roosting sites are alsodangerous, due to the presence of peregrines, merlins(Falco columbarius) and northern harriers (Circus cya-neus), all of which prey on dunlins [18]. Finally, the Fra-ser estuary is large and highly productive, and thewinter in southwest British Columbia generally mild,which makes it possible to finance the extra energeticexpenditure required. Dunlins cease over-ocean flockingduring extended periods of freezing weather [18], ondays with heavy rain (see Fig. 1) and on windless days,which suggests that increases in the energy requirement(freezing weather, rain) or the cost of the gliding andhovering mode of OOF flight (lack of wind) make it tooexpensive. Analyzing over-ocean flocking on the Fraserestuary and on other wintering areas in this way shouldlead to further understanding of when and why it isobserved.ConclusionsPacific dunlins (Calidris alpina pacifica) wintering onthe Fraser River estuary in southwest British Columbiaaltered aspects of their wintering ecology between the1970s and 1990s. We found that Pacific dunlins werelighter by several grams in the 1990s as compared withthe 1970s. The difference appears largely or entirelyattributable to a reduction in fat reserves. Historicalobservations indicate that over-ocean flocking in placeof roosting at high tide was rare prior to the mid-1990sand became common thereafter. We interpret thesechanges as a shift by dunlins toward increased protec-tion against predation, due to the increase since the1970s in the abundance of Peregrine falcons (Falco pere-grinus). Fat stores offer protection against starvation, butare a liability in escape performance, and increase flightcosts. ‘Over-ocean’ or ‘high tide’ flocking is a relativelysafe but energetically-expensive alternative to roostingduring the high tide period. Shifting use of these tacticsalters the balance of protection against predation vs.starvation risks.MethodsGeneralPacific dunlins winter at sites along the Pacific coastfrom southwest British Columbia to Baja California.Some 20-40,000 spend the winter period (November -March) on the Fraser River estuary in southwest BritishColumbia. As many as 100,000 or more may be presentin October during southward migration from Alaskanbreeding grounds [35,36], and in April during the returnnorthward migration. The tidal regime is semi-diurnaland winter low tides occur mostly at night. At high tide,dunlins roost along the foreshore, or on very high tidesin the adjacent saltmarsh and nearby fields. Roostingflocks may form almost anywhere on the nearly-contin-uous 25 km of foreshore along the estuary edge. Theobservations and counts of dunlins reported in the var-ious studies cited below were made from the dyke thatborders the estuary.Over-Ocean FlockingFlocks of shorebirds are well-known to respond to rap-tor attacks with spectacular flight maneuvers. The flightmode during over-ocean flocking is very different. Theearliest description we could find is in an unpublishedreport by Fry [19] on the wintering ecology of Pacificdunlins on the Fraser estuary. She writes that ‘ [over-ocean flocking] was.... noted during [autumn] migrationwhen flocks did not roost at high tide - instead theyformed extensive, widespread hovering clouds of birdshigh over the bay, occasionally roosting for short periodsbut displaying restlessness.’ Dekker’s [18] description,made at the same site, says ‘....the majority of the dunlinsflew out over the ocean, where they coursed back andforth. The flock drifted in a loose cloud on the wind orcoalesced into a dense undulating stream low over thewaves...’. These attributes can be clearly observed on thevideo sequence at http://www.sfu.ca/biology/wildberg/overoceanflocking.htmlOver-ocean flocking is also prolonged relative to pre-dator escape flights. Hötker [17] distinguished ‘shortflights caused by an actual attack of a raptor’ from ‘air-borne roosting’ by the length of flight (at least 30 min).Over-ocean flocking can be difficult to spot becausebirds fly far out over the ocean, may fly very high orvery low, and the flight involves ‘hovering’ or ‘drifting’rather than the conspicuous, flashing turns of shorebirdflocks evading predators. However, it may last through-out the high tide period (see below), and thus affordsample opportunity to be observed if it does occur.We located reports from all the field studies and obser-vations of wintering Pacific dunlins on the Fraser Riverestuary made since 1970, documented in published aswell as unpublished sources. All these studies involvedexperienced observers making frequent and regular visitsto the foreshore during high tide periods to census num-bers, to mist-net birds for banding or to collect otherdata, all of which created opportunities to see over-oceanflocking if it did occur. We interviewed all the investiga-tors we could trace about their recollections of this beha-vior, and also contacted other knowledgeable observerswho had made frequent visits to the estuary. These stu-dies and contacts are summarized in Table 1.Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 8 of 11Detailed observations of over-ocean flocking were car-ried out by one of us (DD) in the course of an observa-tional study of the hunting habits of peregrines. In eightwinters from 1994 to 2006, DD made 169 high-tide vis-its to the foreshore of the Fraser estuary, logging morethan 1000 observation hours [see [18,24,34]]. During the17 sequential days of observation (January 11-27, 2006)reported here, he monitored dunlin flocks from a van-tage point on the dyke. The onset and cessation of over-ocean flocking were recorded.Winter MassPacific dunlins were captured on the Fraser estuary dur-ing the winters of 1977/78, 1978/79 and 1979/80 by GK(n = 2680; see [27]). Captures were made using mistnetson mudflats at diurnal high tides. Each bird was weighed(± 1.0 g), and the exposed culmen (± 0.1 mm) and wing(± 1.0 mm) were measured. Sex was assigned based onculmen length (≤ 37.7 mm = male; 37.8 mm - 39.7 mm= unassigned; ≥ 39.8 mm = female) following Page [37].Note that by this method almost 30% individuals cannot be sex-assigned. Complete data for 1316 individualscaptured in winter 1978/1979 are in the archives of theCanadian Wildlife Service. Unfortunately, the originaldata for individuals captured in the preceding (1977/78)and subsequent winter (1979/80) could not be located,though a summary table of mean monthly weightsremains ([27]; Figs 1 and 2). We compare this to ananalogous summary prepared from the detailed data forwinter 1978/79, to help establish whether the latter yearwas perhaps unusual.Pacific dunlins were again captured on the Fraserestuary during the winters of 1994/95, 1995/96 and1997/98 for a radio-tracking study by PS ([31,38]; n =111), and during the winters of 1997/98, 1998/99 and1999/2000 for a feeding study by LEO ([39,40]; n =295). Capture methods differed from the 1970s in thatbirds were mist-netted not only on mudflats, but inadjacent fields on nocturnal high tides. Birds wereweighed and measured following the same protocol asin the 1970s (i.e. following [37]; see [28]). We comparemorphometrics of dunlins from the 1970s (GK’s 1978/79 sample) with those from the 1990s (birds caught byPS and by LEO). Some measures are missing for a fewindividuals, so sample sizes vary slightly in some of thecomparisons made below.We summarized seasonal patterns of body mass bygenerating date-specific predicted values using b-splineprocedures implemented by PROC TRANSREG in SAS®[41], generating a third degree equation with no knotsto allow for two inflection points over the winter. Wecalculated separate splines for each decade to comparepatterns visually.To test for statistical differences between decades, wefirst controlled for variation in body size by regressingmass against culmen length, winglength, and decade.Predicted mass values from this regression were used asinput for values of a seasonal spline calculated on datapooled over decades. We then tested for decade effectson date-specific residuals from this spline.McEwan & Whitehead [29] analyzed the whole-car-cass body composition (water content, fat-free dry mass,fat content) of 171 dunlins shot on the Fraser Riverestuary on six dates in the winter of 1979/80. (Thesebirds form a separate sample from that reported in Kai-ser & Gillingham [27].) We measured the amount of fatcarried by dunlins in the 1990s from birds collected inthe field by PS and LEO. All were either mist-net mor-talities, or found as fresh carcasses under a set of high-tension electrical lines along one of the jetties whichdunlin flocks routinely crossed. We presumed thesebirds were killed when they struck the wires in flight.Parts of many of these 64 carcasses were used in otherstudies, but the whole-body fat content could be mea-sured from a sample of 14. Following the methoddescribed by McEwan & Whitehead [29] permits adirect comparison.Peregrine AbundanceSince 1986, near-daily surveys for peregrines have beenmade by manager John Ireland at the George C. ReifelMigratory Bird Sanctuary, located on the Fraser estuary.In accord with the continental recovery of peregrinepopulations, the number of peregrines sighted has risensteadily. Also in accord with the continental pattern, theincrease began in the late 1970s or early 1980s [25],from a near-zero level. Peregrine abundance during thewinter and adjacent migration periods (October throughApril) was indexed by the daily number of peregrinessighted, calculated over 5 d periods, and averaged overall years (1986 - 2000). The data used here are pre-sented in Figure three of Lank et al. [1].List of AbbreviationsOOF: over-ocean flocking; defined in Methods.AcknowledgementsWe thank Anthea Farr, Richard Swanston, Yuri Zharikov, Moira Lemon, RobButler and especially the late Rudi Drent for sharing their knowledge withus, and helping us track down various studies and documents. John Irelandgenerously shared his data on peregrine numbers. Sarah Jamieson andKristin Gorman helped with fat analyses. Mike McKinlay made and edited thevideo of OOF available at http://www.sfu.ca/biology/wildberg/overoceanflocking.htmlToo many people assisted in the field to list here, but we acknowledge ourdebt to their efforts. Many landowners graciously permitted farmland access.Funding for this work was provided by Agriculture and Agri-Food Canada(Matching Investment Initiative), National Sciences and Engineering ResearchCouncil of Canada, Wildlife Habitat Canada, Environment Canada ScienceHorizons, Simon Fraser University, Delta Farmland and Wildlife Trust,American Museum of Natural History, American Ornithologists’ Union,Boundary Bay Conservation Committee, Pacific Field Corn Association,Canadian Wildlife Service, Cooper Ornithological Society, Government ofYdenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 9 of 11British Columbia First Job in Science and Technology program, John K.Cooper Award, and a Sigma-Xi Grant-in-Aid of Research. Permits for the1970s work were granted by the Canadian Wildlife Service, while the 1990swork was carried out with permits from the Canadian Wildlife Service andSimon Fraser University (SFU Animal Care Committee Project # 492B).Author details1Centre for Wildlife Ecology, Simon Fraser University, Burnaby, BC V5A 1S6,Canada. 23819-112A Street NW, Edmonton, AB T6J 1K4, Canada. 3402-3255Glasgow Avenue, Victoria, BC V8X 4S4, Canada. 4Western and NorthernService Centre, Parks Canada 300 - 300 West Georgia Street, Vancouver, BCV6B 6B4, Canada. 5Centre for Applied Conservation Research, Forest SciencesCentre, 2424 Main Mall, University of British Columbia, Vancouver, BC V6T1Z4, Canada.Authors’ contributionsRCY conceived and co-ordinated the study, assembled the diverse data sets,located and interviewed the authors of previous studies, and wrote themanuscript. DD carried out almost all the direct observation of OOF,including the data reported in Fig. 1. GK, PCFS and LEO collected andcontributed the data on Pacific dunlin body mass. KR carried out theanalyses of whole body fat. DBL managed all of the data, designed andperformed the statistical analyses, and helped to draft the manuscript. Allauthors read and approved the final manuscript.Received: 4 June 2009Accepted: 21 January 2010 Published: 21 January 2010References1. Lank DB, Butler RW, Ireland J, Ydenberg RC: Effects of predation dangeron migration strategies of sandpipers. Oikos 2003, 103:303-319.2. Lind J, Cresswell W: Anti-predation behaviour during bird migration; thebenefit of studying multiple behavioural dimensions. J Ornithol 2006,147:310-316.3. Ydenberg RC, Butler RW, Lank DB: Effects of predator landscapes on theevolutionary ecology of routing, timing and molt by long-distancemigrants. J Avian Biol 2007, 38:523-529.4. Hedenström A: Flight performance in relation to fuel load in birds. JTheor Biol 1992, 158:535-537.5. Burns JG, Ydenberg RC: The effects of wing loading and gender on theescape flights of least sandpipers (Calidris minutilla) and westernsandpipers (Calidris mauri). Beh Ecol Sociobiol 2002, 52:128-136.6. Houston AI, McNamara JM: A theoretical investigation of the fat reservesand mortality levels of small birds in winter. Ornis Scand 1993,24:205-219.7. Brodin A, Clark CW: Energy storage and expenditure. Foraging Chicago:University of Chicago PressStephens DW, Brown JS, Ydenberg RC 2007,221-269.8. Kelly JP, Warnock N, Page GW, Weathers WW: Effects of weather on dailybody mass regulation in wintering dunlin. J Exp Biol 2002, 205:109-120.9. Piersma T, Koolhaas A, Jukema J: Seasonal body mass changes in Eurasiangolden plovers Pluvialis apricaria staging in the Netherlands: decline inlate autumn mass peak correlates with increase in raptor numbers. Ibis2003, 145:565-571.10. Gosler AG, Greenwood JJD, Perrins C: Predation risk and the cost of beingfat. Nature 1995, 377:621-623.11. Gentle LK, Gosler AG: Fat reserves and perceived predation risk in thegreat tit, Parus major. Proc Roy Soc Lond B 2001, 268:487-491.12. Buchanan JB: A comparison of behavior and success rate of merlins andperegrine falcons when hunting dunlins in two coastal habitats. J RaptorRes 1996, 30:93-98.13. Rogers DI, Piersma T, Hassell CJ: Roost availability may constrain shorebirddistribution: exploring the energetic costs of roosting and disturbancearound a tropical bay. Biol Conserv 2006, 133:225-235.14. Rosa S, Encarnação AL, Granadeiro JP, Palmeirim JM: High water roostselection by waders: maximizing feeding opportunities or avoidingpredation?. Ibis 2006, 148:88-97.15. Prater AJ: Estuary birds of Britain and Ireland Calton: T & AD Poyser 1981.16. Brennan LA, Buchanan JB, Herman SG, Johnson TM: Interhabitatmovements of wintering dunlins in Western Washington. The Murrelet1985, 66:11-16.17. Hötker H: When do dunlins spend high tide in flight?. Waterbirds 2000,23:482-485.18. Dekker D: Over-ocean flocking by dunlins, Calidris alpina, and the effectof raptor predation at Boundary Bay, British Columbia. Can Field-Nat1998, 112:694-697.19. Fry K: Aspects of the winter ecology of the dunlin (Calidris alpina) on theFraser River delta. Technical report, Canadian Wildlife Service, Pacific andYukon region, Canada 1980.20. Conklin JR: Roost site-fidelity of dunlin (Calidris alpina pacifica) winteringon Humboldt Bay, California. MSc Thesis Humboldt State University, Arcata2005.21. Dekker D, Ferwerda A: Slechtvalken in Noard-Fryslan Butendyks. Twirre2008, 19:2-10.22. Dekker D: Hunting success rates, foraging habits, and prey selection ofPeregrine Falcons migrating through central Alberta. Can Field-Nat 1980,94:371-382.23. Cresswell W: Surprise as a winter hunting strategy in sparrowhawks,peregrines, and merlins. Ibis 1996, 138:684-692.24. Dekker D, Ydenberg RC: Raptor predation on wintering dunlins in relationto the tidal cycle. Condor 2004, 106:415-419.25. Ydenberg RC, Butler RW, Lank DB, Smith BD, Ireland J: Western sandpipershave altered migration tactics as peregrine falcon populations haverecovered. Proc Roy Soc Lond Ser B 2004, 271:1263-1269.26. Butler RW: Abundance, distribution and conservation of birds in thevicinity of Boundary Bay, British Columbia. Canadian Wildlife ServiceTechnical Report Pacific and Yukon region, Canada 1992, 155.27. Kaiser GW, Gillingham M: Some notes on the seasonal fluctuations inweight of dunlin (Calidris alpina) on the Fraser River delta, BritishColumbia. Wader Study Group Bull 1981, 31:46-48.28. Shepherd PCF, Lank DB, Smith BD, Warnock ND, Kaiser GW, Williams TD:Sex ratios of dunlin wintering at two latitudes on the Pacific coast.Condor 2001, 103:352-360.29. McEwan EH, Whitehead PM: Seasonal changes in body weight andcomposition of dunlin (Calidris alpina). Can J Zool 1984, 62:154-156.30. Arffman CM: A ten year study of wintering raptor populations in thelower Fraser Valley region of British Columbia. BSc thesis Simon FraserUniversity, Department of Biological Sciences 1987.31. Shepherd PCF: Space use, habitat preferences and time-activity budgetsof non-breeding dunlin (Calidris alpina pacifica) in the Fraser River delta,B.C. PhD thesis Simon Fraser University, Department of Biological Sciences2001.32. Dekker D: Peregrine Falcon predation on ducks in Alberta and BritishColumbia. J Wildl Manag 1987, 51:156-159.33. Dekker D: Prey capture by peregrine falcons wintering on southernVancouver Island, British Columbia. J Raptor Res 1995, 29:26-29.34. Dekker D: Peregrine predation on dunlins and ducks and kleptoparasiticinterference from bald eagles wintering at Boundary Bay, BritishColumbia. J Raptor Res 2003, 37:91-97.35. Warnock ND, Gill RE: Dunlin (Calidris alpina). The birds of North America. No203 Philadelphia: The Academy of Natural Sciences, Washington, DC: TheAmerican Ornithologists’ UnionPoole A, Gill F 1996.36. Shepherd PCF: Status and conservation of dunlin (Calidris alpina) inCanada. Bird Trends Ottawa: Canadian Wildlife Service Publications 2001, 8.37. Page G: Age, sex, molt and migration of dunlin at Bolinas Lagoon.Western Birds 1974, 5:1-12.38. Shepherd PCF, Lank DB: Marine and agricultural habitat preferences ofdunlin wintering in British Columbia. J Wildl Manag 2004, 68:61-73.39. Evans Ogden LJ: Non-breeding shorebirds in a coastal agriculturallandscape: winter habitat use and dietary sources. PhD thesis SimonFraser University, Department of Biological Sciences 2002.40. Evans Ogden LJ, Hobson KA, Lank DB, Bittman S: Stable isotope analysisreveals that agricultural habitat provides an important dietarycomponent for nonbreeding Dunlin. Avian Conserv Ecol - Écol conservoiseau 2005, 1:3http://www.ace-eco.org/vol1/iss1/art3/.41. SAS: SAS for Windows 9.1 Cary, NC: SAS Institute Inc 2002.42. Campbell RW, Shepard MG, Drent RH: Status of birds in the Vancouverarea in 1970. Syesis 1972, 5:137-167.43. Major PF, Dill LM: The three-dimensional structure of airborne bird flocks.Beh Ecol Sociobiol 1978, 4:111-122.Ydenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 10 of 1144. Farr A: Foraging strategies of dunlin in the intertidal areas betweenBoundary Bay and Roberts Bank. Delta, British Columbia: Canadian WildlifeService 1981.45. Zharikov Y, Elner RW, Shepherd PCF, Lank DB: Interplay between physicaland predator and landscapes affects transferability of shorebirddistribution models. Landscape Ecol 2009, 24:129-144.doi:10.1186/1472-6785-10-1Cite this article as: Ydenberg et al.: Winter body mass and over-oceanflocking as components of danger management by Pacific dunlins. BMCEcology 2010 10:1.Publish with BioMed Central   and  every scientist can read your work free of charge"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."Sir Paul Nurse, Cancer Research UKYour research papers will be:available free of charge to the entire biomedical communitypeer reviewed and published immediately upon acceptancecited in PubMed and archived on PubMed Central yours — you keep the copyrightSubmit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.aspBioMedcentralYdenberg et al. BMC Ecology 2010, 10:1http://www.biomedcentral.com/1472-6785/10/1Page 11 of 11

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0223421/manifest

Comment

Related Items