The importance of pre-breeding areas for the arctic Barnacle Goose Branta leucopsis

The importance of pre-breeding areas for the arctic Barnacle Goose Branta leucopsis Christiane E. Hübner 1,2 Hübner C.E. 26. The importance of pre-breeding areas for the arctic Barnacle Goose Branta leucopsis. Ardea 94(3): 71 713. Before the final move to their breeding sites, many arctic-nesting geese visit staging areas close by, so called pre-breeding areas. The present study investigated the timing of migration by Barnacle Geese Branta leucopsis in relation to the dynamics in body stores during this period. The study area was located on the south-facing slopes of Ingeborgfjellet, Vårsolbukta, on the west coast of Spitsbergen, Svalbard (77 45'N, 14 24'E). During spring 23 and 24, staging geese were counted and body fat stores were estimated by scoring abdominal profiles on a daily basis. In total, almost one fifth of the Svalbard Barnacle Goose visited the area. Mean goose arrival was on 26 May in both years and peak numbers were recorded in the last week of May. Geese arrived at a time when fat deposition rates in their staging area in mid-norway dropped rapidly and geese staging in Vårsolbukta could achieve higher deposition rates. This supports the green wave hypothesis. An advantage of early arrival in the Arctic is the better predictability of the snow conditions at the breeding sites. Accordingly, geese departed earlier from Vårsolbukta in the year with advanced snow melt. Late arriving geese stayed shorter than early arriving geese and mean duration of staging was 3.9 d and 2.5 d in 23 and 24, respectively. Despite the short staging period, individuals gained considerable body fat stores. Males were apparently more sensitive to environmental conditions than females and deposited fat stores only in the year with early snow melt. This study suggests that Barnacle Geese can afford to arrive early in Svalbard without trading the benefit of being early against low body condition, provided that they stage in favourable pre-breeding areas before moving to the breeding sites. It also underlines the critical role of such pre-breeding areas, since they may function as a buffer between southern staging areas and the breeding site. Key words: pre-breeding areas, Branta leucopsis, Svalbard, Arctic, abdominal profiles, fat deposition, migration decisions 1 The University Centre in Svalbard, P.O. Box 156, 9171 Longyearbyen, Norway; 2 University of Tromsø, 937 Tromsø, Norway; (christiaane@npolar.no) INTRODUCTION It is generally accepted that performance before breeding plays an important role in the reproductive cycle of arctic-breeding birds (Ryder 197, Klaassen 23). Nevertheless, relatively little is known about the last step between spring staging and breeding, the pre-breeding period (Ganter & Cooke 1996, Tombre et al. 1996, Klaassen et al. 21). During this period, the geese may face a

72 ARDEA 94(3), 26 trade-off between the advantage of being early and the need to gather enough body stores, since both, the timing of nest initiation and the amount of body reserves before egg-laying, play an important role for the reproductive success (Ankney & MacInnes 1978, Ebbinge et al. 1982, Prop & de Vries 1993, Ebbinge & Spaans 1995, Dalhaug et al. 1996). Several goose species feed intensively before nest initiation (Fox & Madsen 1981, Fox & Ridgill 1985, Gauthier & Tardif 1991, Bromley & Jarvis 1993, Fox & Bergersen 25) and it has been suggested that daily energy requirements during the pre-breeding period may be covered mainly by food from the local area, rather than from body stores brought along (Bromley & Jarvis 1993, Ganter & Cooke 1996). However, at the time the geese arrive in the Arctic the ground is often still snow covered and food availability is low (Prop & de Vries 1993, Dalhaug et al. 1996). This study investigates the timing of migration in relation to the dynamics of body condition during pre-breeding for individual Barnacle Geese Branta leucopsis in Svalbard. Barnacle Geese arrive at the archipelago already from mid-may onwards. Before moving to the breeding sites, the birds spend some time at pre-breeding areas along the west coast of Svalbard (Mehlum 1998). Only a few High-arctic pre-breeding areas are known and knowledge about their role in the annual cycle of the geese is limited (but see Ganter & Cooke 1996). These areas are usually located on southfacing slopes and, thus, characterised by earlier snow melt compared to the breeding sites (Prop & de Vries 1993). Consequently, plant phenology is advanced and feeding in those areas may provide an opportunity to refill body stores before arriving at the breeding sites. The geese might in addition be able to predict snow conditions at the breeding sites and time their departure from the pre-breeding areas accordingly. The green wave hypothesis explains the timing of migration between staging areas with the timing of emergence of high quality food after snow melt (Owen 198, van der Graaf et al. 26). As spring commences, a green wave follows the snow melt and moves northwards (Drent et al. 1979, Owen 198, van der Graaf et al. 26). By following this green wave, geese benefit from high quality food that enables them to accumulate fat stores at high rates. The hypothesis predicts that a goose should leave an area when feeding conditions deteriorate and move to a staging area further north where higher fat deposition rates can be achieved. However, it still remains to be shown, whether the green wave hypothesis can be applied for the step between southern staging areas and the high-arctic. If so, the geese would arrive at the pre-breeding areas at a time when the local feeding conditions enable them to achieve higher fat deposition rates than further south at that date. The following questions will be addressed in this paper. (1) How important are specific prebreeding areas for the Svalbard Barnacle Goose population? (2) What are the fat deposition rates for individual geese and how do they compare to those in southern staging areas? (3) What are the individual migratory decisions at, and prior to arrival at the pre-breeding area? Data on timing of migration and body condition, estimated as abdominal profiles, for individual Barnacle Geese were collected during two consecutive spring seasons. MATERIAL AND METHODS Study area The study was conducted on the southwest-facing slopes of the mountain Ingeborgfjellet, Vårsolbukta, on the west coast of Spitsbergen, Svalbard (77 45'N, 14 24'E) in 23 and 24 (Figs 1 and 2). High up in the mountain, the cliffs are covered by large colonies of Little Auks Alle alle (>2 pairs), Brünnich s Guillemots Uria lomvia (>1 pairs) and Kittiwakes Rissa tridactyla (>5 pairs; Pedersen et al. 2). The apparently high input of marine nutrients into the terrestrial system by these seabirds and the relatively early snow melt due to its aspect make Vårsolbukta presumably to an attractive pre-breeding area for geese. All three goose species breeding on Svalbard,

Hübner: PRE-BREEDING AREAS FOR ARCTIC GEESE 73 Figure 1. Vårsolbukta, a pre-breeding area for geese on the west coast of Spitsbergen, Svalbard. Most geese feed on the lower slopes of the mountain Ingeborgfjellet (photo C. Hübner). Barnacle Geese, Pink-footed Geese Anser brachyrhynchus, and Light-bellied Brent Geese Branta bernicla hrota, use the area during early spring (Lagerborg & Varpe 1999, Wrånes 21). The majority of the geese in Vårsolbukta feed in wet habitats, where the main vegetation type is dominated by the moss Calliergon richardsonii, grasses, like Tundra Grass Dupontia fisheri and Arctic Marsh grass Arctophila fulva, and the dicotyledonous Tundra Buttercup Ranunculus hyperboreus. During the pre-breeding period, grasses and forbs are still very scarce, thus the main food source for the geese is Calliergon (unpubl. data). During spring and summer the area is also heavily grazed by the Svalbard Reindeer Rangifer tarandus platyrhynchus. Study species Almost the entire Barnacle Goose population breeding on Svalbard winters at the Solway Firth, UK. The population reached critical low numbers in the 195s, but after full protection, numbers increased (Owen 198) and are now stabilizing at about 27 birds (Trinder et al. 25). During spring the geese leave the Solway Firth from mid- April onwards and migrate via staging areas off the Norwegian coast to Svalbard, where they arrive from mid-may onwards (Owen & Black 1999 and references therein, Fig. 2). About 1% of all birds in the population are marked with Darvic colour rings with alpha-numeric codes that can be read by telescope up to 25 m distance. Estimates of the total number of rings in the population are

74 ARDEA 94(3), 26 Kongsfjorden Nordenskiöldkysten Vårsolbukta 3. Vårsolbukta 2. Helgeland 1. Solway Firth Figure 2. Flyway of the Barnacle Goose population breeding on western Spitsbergen, Svalbard: 1. Solway Firth, UK: the main wintering area; 2. Helgeland, Norway: the main staging area during spring migration; 3. Vårsolbukta, Svalbard: a pre-breeding area. Inset: Spitsbergen with Kongsfjorden and Nordenskiöldkysten, two of the breeding sites on the west coast. 1956 and 265 rings in 23 and 24, respectively (J. Prop, pers. comm.). Temperatures, snow melt and vegetation phenology To monitor environmental conditions like temperature, snow melt and vegetation phenology, an 8x8 m square was defined in the wet habitat. Two temperature loggers (Gemini TinytagPlus) were randomly placed on the moss surface and their hourly measurements were used to calculate the daily mean temperature. Snow cover was assessed from daily digital photos, using the range filter plug-in in Adobe Photoshop Elements 2.. Weekly digital photos of randomly located plots (5x5 cm 2, three in 23 and two in 24) were analysed to estimate food availability for geese. A point quadrat method (Goodall 1952, Stampfli 1991) with 81 points was adapted to the computer screen and only points hitting vegetation were included in the analyses. Three categories were distinguished: green parts of grasses, forbs and other plant material (dead grass material, mosses, and lichens). When snow disappears, green moss emerges and covers almost the entire ground layer. Therefore, it was assumed that moss as a food source was only limited by snow cover. Goose staging Each year, geese were counted daily from 5 May until 1 June. The counting area covered a range of approximately 4.5 km 2 and included the majority of places used by the staging geese. During a count the area was divided in 2 sub-areas and

Hübner: PRE-BREEDING AREAS FOR ARCTIC GEESE 75 counted simultaneously by two observers. Flight activity of staging geese was low and the risk of counting individual geese twice was negligible. The counts lasted for.5 1.5 h, depending on the numbers of geese and weather conditions, and always took place between 1: 15: hrs. At least once per day, the area was searched for ringed geese and whenever possible the ring was identified. The risk that a ring was overlooked for more than one day was relatively small, i.e. 75% of all consecutive sightings of individual rings were made the next day or the day thereafter. The sex of ringed birds was determined in the field with help of behaviour and size of the birds and later confirmed by data from the winter area (L. Griffin, pers. comm.) and from catches during the moulting period (M.J.J.E. Loonen, pers. comm.). Arrival and departure dates for individual geese were defined as the date of first (date arr ) and last ring sighting (date dep ), respectively. Length of stay, i.e. the staging period for individual geese, was calculated by date dep date arr + 1. For pairs, only one set of data was used for calculating length of stay, arrival and departure dates as paired geese generally migrate together (pers. obs.). Abdominal profiles Scoring of abdominal profiles (AP) as a method to estimate the amount of body fat stores in geese is a well-accepted alternative to catching and other intrusive techniques (Owen 1981, Boyd et al. 1998, Zillich & Black 23). The shape of the abdomen is matched against a set of standard APs ranging from 1 7 (Black et al. 1991). In several goose species a close linear relationship between AP and body condition has been found (Fox et al. 1998, Fox 23, Zillich & Black 23, Madsen & Klaassen 26). This relationship was assumed to apply also for the Barnacle Goose and thus AP scores were used as continuous data. A modified version of the AP-drawings from Owen (1981) was used in the field to score APs of ringed birds. To calibrate observations between years and observers, digital photos taken of geese in feeding position were frequently discussed by the observers prior to and during the field seasons. APs were registered when the bird was standing in a lateral, head-down posture. AP at arrival (AP arr ) and departure (AP dep ) was defined as the AP at first and last sighting of the ringed bird, respectively. For individuals observed for several days, changes in fat stores were calculated as the difference in AP score between first and last sighting. Individual fat deposition rates were calculated by (AP dep AP arr )/length of stay, whereas daily fat deposition rates were averaged from individual AP changes for each date. Statistical analyses Statistical analyses were done using R, version 2.2.1 (R Development Core Team 25). Data on length of stay were positively skewed and should generally be considered as survival time data. Therefore, a Cox regression was used to test for differences in length of stay between years and arrival dates (Crawley 22). The relationship between arrival and departure dates was tested with a linear regression. General linear model analyses were used to identify variables affecting APs and individual fat deposition rates. In the models, year and sex were defined as class variables and non-significant variables were deleted by stepwise iteration. Results are based on the Type III Sum of Squares. A linear mixed effects model was applied to test for daily fat deposition rates, using the individual birds as random factor to account for repeated measurements. P-values smaller than.5 were considered statistically significant and all values are presented as means ± SE. RESULTS Temperatures, snow melt and vegetation phenology During the early period of goose staging, daily mean temperatures on the moss surface where slightly lower in 24 than in 23, but after 21 May temperatures were higher in 24 (Fig. 3A). The timing of snow melt differed considerably between the two years (Fig. 3A). In 24, the date of 5% snow clearance was two weeks earlier than

76 ARDEA 94(3), 26 snow cover (%) green grass (%) number of geese 1 8 6 4 2 2 15 1 5 2 15 1 5 23 24 23 24 23 24 15 2 25 3 4 9 May June in 23. The difference decreased towards the end of the melting period and the area was entirely snow free nine days earlier in 24 than in 23. However, the relative amount of green grass and forbs available in snow free patches was generally low and showed no difference between years A B C 6 4 2 2 4 temperature ( C) Figure 3. Environmental conditions and abundance of Barnacle Geese in Vårsolbukta, Svalbard. (A) Snow cover (continuous lines) and daily moss surface temperature (dotted lines); (B) Abundance estimates for green grass as percentage of hits during point framing (for details see methods); (C) Daily number of Barnacle Geese feeding in the area. (Fig. 3B). Thus, food availability was greater in 24 compared to 23 due to the larger amount of snow free area rather than due to more advanced plant development. Goose staging The daily number of Barnacle Geese staging in Vårsolbukta was similar among years during the early period, but after 25 May goose numbers in 23 exceeded the numbers in 24 (Fig. 3C). 26 May was the mean arrival date in both years, whereas the mean departure date differed by three days between years (3 May in 23 and 27 May in 24, t-test, t 331 = 4.97, P <.1). The area was also used by Pink-footed Geese and Light-bellied Brent Geese, although in much smaller numbers (peak counts, Pink-footed Geese: 126 (23) and 25 (24); Light-bellied Brent Geese: 162 (23) and 214 (24); compared to 1911 (23) and 1297 (24) for Barnacle Geese). A total of 412 and 336 individually ringed Barnacle Geese were sighted in 23 and 24, respectively. From the proportion of rings seen in relation to the total number of rings in the population, it follows that Vårsolbukta was used as a prebreeding area by 21.1% (23) and 16.3% (24) of the entire population. Average staging periods were slightly shorter in 24 than in 23 (23: 3.9 ±.27 d, n = 21; 24: 2.5 ±.19 d, n = 164; Cox regression, χ 2 = 23.27, P <.1). About half of the birds were seen once, and apparently stayed less than 24 h (47% in 23 and 58% in 24), whereas the rest of the birds stayed for 6.6 ±.36 d in 23 (n = 15) and for 4.6 ±.33 d in 24 (n = 69, difference between years: χ 2 = 23.12, P <.1). Excluding birds that were seen only once, the date of departure was correlated with the date of arrival in both years (linear regression, 23: R 2 =.35, P <.1, n =15; 24: R 2 =.62, P <.1, n = 69). Variation in length of stay among individuals was high, but decreased with progressing season (Fig. 4). Late arriving birds tended to make only a short stay in both years (Cox regression, 23: χ 2 = 18.55, P <.1, n = 21; 24: χ 2 = 5.31, P <.5, n = 164, Fig. 4).

Hübner: PRE-BREEDING AREAS FOR ARCTIC GEESE 77 length of stay (days) 16 14 12 1 8 6 4 2 23 24 19 22 25 28 31 3 May June date of arrival Figure 4. Length of stay in relation to the date of arrival for individual Barnacle Geese in Vårsolbukta, Svalbard. AP at arrival 4 3 2 1 males 23 females 23 males 24 females 24 18 22 26 3 3 May June date of arrival Figure 5. Abdominal profile (AP) at arrival in relation to the date of arrival for Barnacle Geese in Vårsolbukta, Svalbard. Lines show the predicted values from the model (Table 1A) and dots represent daily means ± SE. Fat deposition AP at arrival was generally higher in 24 than in 23. In both years, females had a higher AP than males and their AP increased with arrival date Table 1. Results of General Linear Models for abdominal profiles (AP) of Barnacle Geese in Vårsolbukta, Svalbard, during 23 and 24. With stepwise iteration non significant variables and interactions were removed from the models. (A) AP at arrival; (B) AP at departure; (C) Fat deposition rates (AP unit change d 1 for individual geese). Source df F P (A) Model 4 15.6 <.1 Year 1 5.74 <.1 Sex 1.49 ns Date of arrival 1 11.66 <.1 Date of arrival x sex 1 12.77 <.1 Residual error 365 (B) Model 4 117.93 <.1 Year 1 24.41 <.1 Sex 1 98.96 <.1 AP at arrival 1 33.1 <.1 Year x sex 1 5.55 <.5 Residual error 255 (C) Model 4 24.94 <.1 Year 1 19.16 <.1 Sex 1 64.8 <.1 AP at arrival 1 82.62 <.1 AP at arrival x year 1 6.88 <.1 Residual error 255 (Table 1A and Fig. 5). At departure, the AP was still higher in female geese than in males, with higher AP at departure in 24 than 23 (Table 1B and Fig. 6). AP at departure was linearly related to AP at arrival, but slopes were smaller than one, indicating a larger increase in body fat stores for birds arriving with little fat (Fig. 6). Although the data suggest a quadratic slope, adding a quadratic term to the model was not significant (F 1,166 =.65, ns). Individual females achieved on average a fat deposition rate of.8 ±.2 AP units d 1 in 23 (n = 77) and.1 ±.4 in 24 (n = 52).

78 ARDEA 94(3), 26 AP at departure 4 3 2 1 males 23 females 23 males 24 females 24 1 2 3 4 AP at arrival Figure 6. Abdominal profile (AP) at departure in relation to AP at arrival for Barnacle Geese in Vårsolbukta, Svalbard. Lines show the predicted values from the model (Table 1B) and dots represent daily means ± SE. For males, individual fat deposition rates were.1 ±.2 AP units d 1 and.9 ±.2 AP in 23 (n = 7) and 24 (n = 52), respectively (for statistics see Table 1C). These deposition rates were negatively correlated with AP at arrival, i.e. birds arriving with little fat stores accumulated fat at higher rates than birds that arrived with larger fat stores. Fat deposition rates decreased steeper with increasing AP at arrival in 24 than in 23 (significant interaction term, Table 1C). There was no trend in time for daily deposition rates in both sexes and years (linear mixed effects model, t 13 = 1.54, ns). DISCUSSION This study is the first to document individual foraging performance and dynamics of body stores of geese staging in an arctic pre-breeding area. Barnacle Geese arrived in Vårsolbukta from mid-may onwards, which corresponds well with the onset of migration from the staging areas in mid-norway (Prop et al. 23, P. Shimmings, pers. comm.). Female AP at arrival increased with progressing season, whereas in males, no seasonal pattern in fat stores at arrival was found. AP scores at arrival were lower in males than in females and remained so at departure. This difference between male and female stores is a common finding during spring (e.g. Raveling 1979, Boyd & Fox 1995, Boyd 2). Mean duration of staging was less than four days, late arriving geese stayed shorter than early arriving geese and variation between individuals was largest in the early period. Despite the short staging period, individuals gained considerable body fat stores during their stay in Vårsolbukta. However, there was a large variation among individuals and males only increased fat stores in the year with early snow melt. The green wave hypothesis The hypothesis of the green wave states that geese should move on to more profitable staging areas along their migration route as soon as foraging conditions deteriorate in the current area. In the present study area, the first geese arrived in Vårsolbukta during mid-may, a time when daily fat deposition rates for females in Helgeland, the main-staging area on the Norwegian mainland, drop rapidly below.1 AP units d 1 (Prop et al. 23). In Vårsolbukta, average daily fat deposition rates for females were constant at.8 and.1 AP units d 1 in 23 and 24, respectively. Consequently, after mid-may staging in Vårsolbukta is presumably more profitable than staging in Helgeland. These findings suggest that the green wave also influences the timing of movement between staging areas on the Norwegian mainland and the High-arctic pre-breeding areas in Svalbard. However, it should be noted that the increase in female AP may be affected by an enlargement of the reproductive tract before egg laying (Boyd & Fox 1995, Boyd et al. 1998, Boyd 2). It remains to be shown to which extent these physiological changes confound the fat store estimates by AP scoring. There are several reasons why geese should be able to deposit fat, i.e. have a positive energy balance, in spite of apparent low food availability during pre-breeding. First, at the high latitudes of

Hübner: PRE-BREEDING AREAS FOR ARCTIC GEESE 79 Svalbard geese experience 24 hours of light during spring and feeding days can thus be longer than in areas further south (Sanz et al. 2, Prop et al. 23). Secondly, the manuring effect of the seabird colony in Vårsolbukta may both stimulate plant growth and increase the plant nutrient content (Born & Böcher 21). Many vascular plants that grow below bird cliffs have been found to contain more nutrients than elsewhere (Summerhayes & Elton 1928, Wainright et al. 1998, Anderson 1999). However, little is known about mosses, the main food source for geese in Vårsolbukta, and further studies on the chemical composition of mosses under bird cliffs are needed. fat deposition (AP units) 4 3 2 1 4 3 prior to Vårsolbukta 23 at Vårsolbukta 23 prior to Vårsolbukta 24 at Vårsolbukta 24 Dynamics of fat stores Prior to arrival in Vårsolbukta, the geese deposit a certain amount of fat stores in previous staging areas, but also burn fat during the migration flight and other energy consuming activities. The amount of fat stores brought along to the prebreeding area is the result of this balance and the slope in Fig. 5 can be interpreted as the daily net change rate of fat stores prior to arrival in Vårsolbukta. For females this net change rate prior to arrival was.7 AP units d 1, whereas an average female staging in Vårsolbukta accumulated fat at a rate of.8 and.1 AP units d 1 in 23 and 24, respectively. This suggests that females can deposit a similar amount of fat either by arriving early in the pre-breeding area or by spending more time feeding in other areas prior to arrival in Vårsolbukta (Fig. 7). Therefore, it seems that individual females have several migratory options to increase their fitness according to their own performance and abilities. Males lost on average.1 AP units d 1 prior to arrival in Vårsolbukta and appeared to be more sensitive to the environmental variation in the prebreeding area than females, i.e. they only deposited extra body fat stores in Vårsolbukta when feeding conditions were good. Daily change rates in Vårsolbukta varied between.1 AP units d 1 in the year with late snow melt (23) and.9 AP units d 1 in the year with early snow melt (24). Consequently, in the year with early 2 1 17 21 25 29 2 May June Figure 7. Comparison of fat deposition rates in Barnacle Geese prior to arrival and during the stay in Vårsolbukta, Svalbard. The continuous lines are the same lines as in Fig. 5 and their slopes represent daily rates in fat deposition prior to Vårsolbukta. The slopes of the dotted lines show the mean fat deposition rates d 1 for geese staging in Vårsolbukta. snow melt, an early arriving male would gain a considerably higher amount of fat stores than a late arriving male would bring along (Fig. 7). The striking difference between sexes in deposition of body stores as estimated from AP scores points to a dichotomy in strategies at this phase just before egg laying. Males that move to the breeding sites with little fat stores may be able to deposit fat later during the incubation period. However, the amount of fat deposited by the female before egg-laying directly affects the reproductive success of a pair, e.g. through clutch size (Ankney & MacInnes 1978) and incubation constancy (Aldrich & Raveling 1983). Thus, during pre-breeding a male should primarily ensure good feeding conditions for his female, and only when conditions allow, refill his own body stores as well.

71 ARDEA 94(3), 26 Figure 8. Barnacle and Pink-footed Geese on snow-free patches during the pre-breeding period in Vårsolbukta, Svalbard (photo C. Hübner). APs at departure were positively related to the APs at arrival in Vårsolbukta, but geese with smaller fat stores at arrival gained relatively more fat than geese arriving with large fat stores. This suggests that pre-breeding areas may act as a buffer where individuals can compensate for their lower initial weight at arrival in the Arctic. Advantage of being early Geese might take the risk of unfavourable conditions at arrival in the Arctic to benefit from the advantages of being early. Many studies have shown a negative relationship between nest initiation date and reproductive success (Cooke et al. 1984, Sedinger & Flint 1991, Prop & de Vries 1993, Lindholm et al. 1994, Loonen 1997, Eichholz & Sedinger 1998). In the pre-breeding areas, the need to depart soon to the breeding sites should therefore generally increase with progressing season. Indeed, geese arriving late in Vårsolbukta showed shorter staging times and less variation in the length of their stay than earlier birds. Preliminary results also suggest that birds breeding further north in Svalbard, i.e. Kongsfjorden, tend to stay shorter in Vårsolbukta than birds breeding close by at Nordenskiöldkysten (Kongsfjorden: 3. ±.48 d, n = 33 (23), 1.9 ±.23 d, n =32 (24); Nordenskiöldkysten: 4.9 ±.49 d, n = 81 (23), 3.5 ±.45 d, n = 44 (24)). Moreover, the period between departure in Vårsolbukta and date of egg laying in the breeding colony differed considerably (Kongsfjorden, 23: 12.7 ± 2. d (n = 1); Nordenskiöldkysten, 24: 5.2 ±.53 d (n = 1)). This suggests that the time spent later

Hübner: PRE-BREEDING AREAS FOR ARCTIC GEESE 711 in other pre-breeding areas depends on the distance to the breeding site. The date of nest initiation, however, depends on the time when nest sites become snow free (Raveling 1978, Alsos et al. 1998), which is highly variable among years (Raveling 1978, Prop & de Vries 1993). Staging in pre-breeding areas close to the breeding sites enables the geese to better predict snow conditions at the breeding sites than if staging further away. They would be able to adjust departure dates and initiate nesting as soon as snow conditions permit. This is supported by data in the present study since geese arrived at the same time in the pre-breeding area in both years, while they departed earlier towards the breeding sites in the year with less snow cover. White-fronted Geese Anser albifrons breeding in Greenland show a similar pattern. They arrive early in May in Greenland and remain near sea level, where they find enough food until their nest sites in the uplands become snow-free (Fox & Madsen 1981, Fox & Ridgill 1985). Importance of pre-breeding areas For arctic-breeding geese, both the need to deposit sufficient fat stores during spring staging as well as the advantage of being early act together in forming the basis of migration decisions (Alerstam & Lindström 199). The present study suggests that Barnacle Geese can afford to arrive early in Svalbard, without trading the benefit of being early against low body condition, by staging in favourable pre-breeding areas before the final move to the breeding sites. This study also emphasises the important role of suitable pre-breeding areas in the Arctic. About one fifth of the Barnacle Goose population breeding in Svalbard was observed in a single pre-breeding area in spring and although staging times were short, individuals gained a considerable amount of body stores. Thus, the area may function as a critical buffer between southern staging areas and the breeding sites. The fact that also Pink-footed Geese and Brent Geese use the same location during this period (Fig. 8) further underlines the critical value of Vårsolbukta as a spring staging area for geese. ACKNOWLEDGEMENTS I am indebted to Katrine Bruntse and Eirin Bjørkvoll for their excellent help and commitment as field assistants during the six weeks in Vårsolbukta. The Governor of Svalbard kindly gave permission to use the cabin Camp Millar and the University Centre in Svalbard provided logistic support in the field. Maarten Loonen and Jouke Prop contributed with egg laying data from the breeding colonies in Ny Ålesund and Diabasøya, respectively. Larry Griffin from the Wildfowl & Wetlands Trust in UK provided information from the data base on ringed Svalbard Barnacle Geese. Thanks to Jouke Prop, Rudi Drent, Ingibjörg S. Jónsdóttir and Ingunn Tombre for comments on earlier drafts and thanks to Jouke Prop for inspiring discussions. This study is financed by the Norwegian Research Council, Roald Amundsens Senter for Arktisk Forskning, Tromsø and the Norwegian National Committee on Polar Research. REFERENCES Aldrich T.W. & Raveling D.G. 1983. Effects of experience and body weight on incubation behavior of Canada Geese. Auk 1: 67 679. Alerstam T. & Lindström Å. 199. Optimal bird migration: The relative importance of time, energy and safety. In: Gwinner E. (ed) Bird migration: physiology and ecophysiology: 331 251. Springer Verlag, Berlin. Alsos I.G., Elvebakk A. & Gabrielsen G.W. 1998. Vegetation exploitation by Barnacle geese Branta leucopsis during incubation on Svalbard. Polar Res. 17: 1 14. Anderson W.B., Polis G.A. 1999. Nutrient fluxes from water to land: seabirds affect plant nutrient status on Gulf of California islands. Oecologia 118: 324 332. Ankney C.D. & MacInnes C.D. 1978. Nutrient reserves and reproductive performance of female Lesser Snow geese. Auk 95: 459 471. Black J.M., Deerenberg C. & Owen M. 1991. Foraging behaviour and site selection of Barnacle Geese Branta leucopsis in a traditional and newly colonised spring staging habitat. Ardea 79: 349 358. Born E.W. & Böcher J. 21. The Ecology of Greenland. Ministry of Environment and Natural Resources, Nuuk. Boyd H. 2. Abdominal profiles of Barnacle Geese Branta leucopsis at staging areas in Iceland in May. Wildfowl 51: 33 47. Boyd H. & Fox A.D. 1995. Abdominal profiles of Icelandic Pink-footed Geese Anser brachyrhynchus in spring. Wildfowl 46: 161 175.

712 ARDEA 94(3), 26 Boyd H., Fox A.D., Kristiansen J.N., Stroud D.A., Walsh A.J. & Warren S.M. 1998. Changes in abdominal profiles of Greenland White-fronted geese during spring staging in Iceland. Wildfowl 49: 57 71. Bromley G. & Jarvis R.L. 1993. The energetics of migration and reproduction of Dusky Canada geese. Condor 95: 193 21. Cooke F., Findlay C.S. & Rockwell R.F. 1984. Recruitment and the timing of reproduction in Lesser Snow geese (Chen caerulescens caerulescens). Auk 11: 451 458. Crawley M.J. 22. Statistical computing: An introduction to data analysis using S-Plus. Wiley, West Sussex. Dalhaug L., Tombre I.M. & Erikstad K.E. 1996. Seasonal decline in clutch size of the Barnacle Goose in Svalbard. Condor 98: 42 47. Drent R.H., Ebbinge B. & Weijand B. 1979. Balancing the energy budgets of Arctic breeding geese throughout the annual cycle: a progress report. Verhandl. Ornith. Gesellsch. Bayern 23: 239 264. Ebbinge B., StJoseph A., Prokosch P. & Spaans B. 1982. The importance of spring staging areas for Arcticbreeding geese, wintering in Western Europe. Aquila 17: 249 258. Ebbinge B.S. & Spaans B. 1995. The importance of body reserves accumulated in spring staging areas in the temperate zone for breeding in Dark-Bellied Brent Geese Branta b. bernicla in the High Arctic. J. Avian Biol. 26: 15 113. Eichholz M.W. & Sedinger J.S. 1998. Factors affecting duration of incubation in Black Brant. Condor 1: 164 168. Fox A.D. 23. The Greenland White-fronted goose Anser albifrons flavirostris. The annual cycle of a migratory herbivore on the European continental fringe. Doctors dissertation (DSc). National Environmental Research Institute, Denmark. Fox A.D. & Bergersen E. 25. Lack of competition between barnacle geese Branta leucopsis and pink-footed geese Anser brachyrhynchus during the pre-breeding period in Svalbard. J. Avian Biol. 36: 173 178. Fox A.D., Kahlert J., Walsh A.J., Stroud D.A., Mitchell C., Kristiansen J.N. & Hansen E.B. 1998. Patterns of body mass change during moult in three different goose populations. Wildfowl 49: 45 56. Fox A.D. & Madsen J. 1981. The pre-nesting behaviour of the Greenland White-fronted goose. Wildfowl 32: 48 54. Fox A.D. & Ridgill S.C. 1985. Spring activity patterns of migrating Greenland White-fronted geese in West Greenland. Wildfowl 36: 21 28. Ganter B. & Cooke F. 1996. Pre-incubation feeding activities and energy budgets of Snow geese: Can food on the breeding grounds influence fecundity? Oecologia 16: 153 165. Gauthier G. & Tardif J. 1991. Female feeding and male vigilance during nesting in Greater Snow Geese. Condor 93: 71 711. Goodall D.W. 1952. Some considerations in the use of point quadrats for the analysis of vegetation. Aust. J. Sci. Res. Ser. B 5: 1 41. Klaassen M. 23. Relationships between migration and breeding strategies in Arctic breeding birds. In: Berthold P., Gwinner E. & Sonnenschein E. (eds) Avian migration: 237 249. Springer Verlag, Berlin. Klaassen M., Lindström A., Meltofte H. & Piersma T. 21. Arctic waders are not capital breeders. Nature 413: 794. Lagerborg M. & Varpe O. 1999. Registrering av ringgås ved Vårsolbukta. Report to the Governor of Svalbard. Lindholm A.G., Gauthier G. & Desrochers A. 1994. Effects of hatch date and food supply on gosling growth in Arctic-nesting Greater Snow geese. Condor 96: 898 98. Loonen M.J.J.E. 1997. Goose breeding ecology: Overcoming successive hurdles to raise goslings. PhD thesis, University Groningen. Madsen J. & Klaassen M. 26. Assessing body condition and energy budget components by scoring abdominal profiles in free-ranging geese. J. Avian Biol. 37: 283 287. Mehlum F. 1998. Areas in Svalbard important for geese during the pre-breeding, breeding and post-breeding periods. Norsk Polarinstutt Skrifter 2: 41 55. Owen M. 198. Wild geese of the world: their life history and ecology. Batsford Ltd, London. Owen M. 1981. Abdominal profile: a condition index for wild geese in the field. J. Wildl. Manage. 45: 227 23. Owen M. & Black J.M. 1999. Barnacle Goose Branta leucopsis: Svalbard. In: Madsen J., Cracknell G. & Fox T. (eds) Goose populations of the Western Palearctic: A review of status and distribution. National Environmental Research Institute, Denmark. Pedersen G., Christensen G.N. & Evenset A. 2. Sammenstilling av data om sjøfugl og sjøpattedyr i området Van Mijenfjorden, Svalbard. Akvaplan-Niva report, APN-421.1991. Prop J. & de Vries J. 1993. Impact of snow and food conditions on the reproductive performance of Barnacle geese Branta leucopsis. Ornis Scand. 24: 11 121. Prop J., Black J.M. & Shimmings P. 23. Travel schedules to the high-arctic: barnacle geese trade-off the timing of migration with accumulation of fat deposits. Oikos 13: 43 414. R Development Core Team 25. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.r-project.org.

Hübner: PRE-BREEDING AREAS FOR ARCTIC GEESE 713 Raveling D.G. 1978. The timing of egg laying by northern geese. Auk 95: 294 33. Raveling D.G. 1979. The annual cycle of body composition of Canada geese with special reference to control of reproduction. Auk 96: 234 252. Ryder J.P. 197. A possible factor in the evolution of clutch size in Ross Goose. Wilson Bull. 82: 5 13. Sanz J.J., Tinbergen J.M., Moreno J., Orell M. & Verhulst S. 2. Latitudinal variation in parental energy expenditure during brood rearing in the Great Tit. Oecologia 122: 149 154. Sedinger J.S. & Flint P.L. 1991. Growth rate is negatively correlated with hatch date in Black Brant. Ecology 72: 496 52. Stampfli A. 1991. Accurate determination of vegetational change in meadows by successive point quadrat analysis. Vegetatio 96: 185 194. Summerhayes V.S. & Elton C.S. 1928. Further contributions to the ecology of Spitzbergen. J. Ecol. 16: 193 268. Tombre I.M., Erikstad K.E., Gabrielsen G.W., Strann K.B. & Black J.M. 1996. Body condition and spring migration in female high-arctic Barnacle geese Branta leucopsis. Wildl. Biol. 2: 247 251. Trinder M., Rowcliffe J.M., Pettifor R., Rees E.C. & Griffin L. 25. Status and population viability of the Svalbard barnacle goose. In: Trinder M., Rowcliffe J.M., Pettifor R., Rees E.C., Griffin L., Ogilvie M.A. & Percival S. (eds) Status and population viability analyses of geese in Scotland. Scottish Natural Heritage Commissioned Report No. 17. van der Graaf A.J., Stahl J., Klimkowska A., Bakker J.P. & Drent R.H. 26. Surfing on a green wave how plant growth drives spring migration in the Barnacle Goose Branta leucopsis. Ardea 94: 567 577. Wainright S.C., Haney J.C., Kerr C., Golovkin A.N. & Flint M.V. 1998. Utilization of nitrogen derived from seabird guano by terrestrial and marine plants at St. Paul, Pribilof islands, Bering sea, Alaska. Marine Biol. 131: 63 71. Wrånes E. 21. Registrering av ringgås og hvitkinngås ved Vårsolbukta, Bellsund. Report to the Governor of Svalbard. Zillich U. & Black J.M. 23. Body mass and abdominal profile index in captive Hawaiian geese. Wildfowl 53: 67 77. SAMENVATTING Voordat hoog-noordelijk broedende ganzen definitief naar hun nestplek gaan, verblijven ze enige tijd in speciale gebieden in de buurt van de broedplaatsen. In 23 en 24 werden aan de westkust van Spitsbergen bij Brandganzen Branta leucopsis de aankomst en verblijfsduur van de ganzen en de grootte van de lichaamsreserves gedurende deze periode onderzocht. Het onderzoeksgebied bevond zich op de hellingen van Ingeborgfjellet, een ruim 7 m hoge bergketen die uit een brede fjord oprijst. Op basis van ringwaarnemingen werd geschat dat een vijfde deel van de totale populatie die op Spitsbergen broedt, dit gebied aandeed. De gemiddelde aankomsttijd was in beide jaren 26 mei. Ook de piekaantallen werden steeds in de laatste week van mei geregistreerd. De aankomst van ganzen viel samen met het moment waarop de foerageeromstandigheden in de pleistergebieden langs de Noorse kust sterk achteruitgingen. Overeenkomstig de groenegolfhypothese waren de vogels in staat op Spitsbergen sneller vetvoorraden aan te leggen dan op de zuidelijker gelegen pleisterplaatsen als ze daar waren gebleven. De aanmaak van vetvoorraden werd geschat door de buikomvang van gemerkte vogels te scoren. Een voordeel van een vroege aankomst op Spitsbergen bleek dat de vogels beter in staat waren de sneeuwcondities op de broedplekken in te schatten. In een jaar waarin de dooi vroeg inviel, trokken de ganzen vroeger door naar de broedkolonies. De gemiddelde verblijftijd op de pleisterplaats op Spitsbergen was 3,9 (23) en 2,5 (24) dagen, waarbij individuen die laat arriveerden korter bleven dan vroege vogels. Ondanks de betrekkelijk korte verblijftijd sloegen de vogels hier een aanzienlijke hoeveelheid vet op. Mannetjes leken gevoeliger te zijn voor sneeuw- en voedselomstandigheden dan vrouwtjes, want in een laat jaar legden de mannetjes tijdens hun verblijf geen reserves aan. De waarnemingen suggereren dat een vroege aankomst in het broedgebied niet ten koste hoeft te gaan van de lichaamsconditie. Voorwaarde is wel dat de ganzen dan in staat zijn de schaarse plekken te vinden waar sneeuw eerder wegsmelt dan elders en de voedselomstandigheden het gunstigst zijn. Het onderzoeksgebied voldeed hieraan, omdat de berghellingen naar het zuiden waren gericht, terwijl bemesting door talrijke zeevogels voor een relatief hoogwaardig voedselaanbod zorgde. Received 31 December 24; accepted 2 April 26

ARDEA is the scientific journal of the Netherlands Ornithologists Union (NOU), published bi-annually in spring and autumn. Next to the regular issues, special issues are produced frequently. The NOU was founded in 191 as a non-profit ornithological society, composed of persons interested in field ornithology, ecology and biology of birds. All members of the NOU receive ARDEA and LIMOSA and are invited to attend scientific meetings held two or three times per year. Subscription ARDEA Separate subscription to Ardea is possible. The 27 subscription rates are 33 (The Netherlands), 37 (Europe), and 43 (rest of the world). See website for institutional subscription rates. Payments to Postbank account 125347, as indicated below under Membership NOU. Correspondence concerning subscription, and orders for back volumes to: J.J. de Vries, Oosterend 1 b, 8897 HZ Oosterend, The Netherlands (jacobird@xs4all.nl). Internet www.ardeajournal.nl NETHERLANDS ORNITHOLOGISTS UNION (NOU) Membership NOU The 27 membership fee is 41 (or 24 for persons <25 years old at the end of the year). Foreign membership amounts to 53 (Europe), or 63 (rest of the world). All payments to Postbank account 125347 in the name of Nederlandse Ornithologische Unie, Ir. van Stuivenbergweg 4, 6644 AB Ewijk, The Netherlands (BIC: PSTBNL21 IBAN: NL65 PSTB 125347). Correspondence concerning membership: J.J. de Vries, Oosterend 1 b, 8897 HZ Oosterend, The Netherlands (jacobird@xs4all.nl). Internet www.nou.nu