Anda di halaman 1dari 11
343 DAILY ACTIVITY AND INTAKE RATE PATTERNS OF WINTERING COMMON CRANES Grus grus JUAN C. ALONSO! & JAVIER A. ALONSO? ABSTRACT Daily activity panemns of wintering Common Cranes Grus grus were studied at Gallocanta, NE Spain, and related with changes in foraging site, distance to roost and flocking behaviour. Per cent time spent feeding, feeding bout length, and food intake rate showed two peaks, in early morning . and late afternoon. Vigilance, preening and other activities followed an in- verse pattern, peaking at midday, when birds usually gathered at drinking sites, The feeding habitat and location, and the flock size changed between. y the morning and the evening foraging peaks. During the moming cranes showed a higher locomotor activity, foraging in smaller flocks, at higher dis- tances from the roost, and on grounds with higher food availability. On the contrary, during the afternoon birds showed lower locomotor activity, aggre- gated in larger flocks, and foraged closer to the roost, on grounds with lower food availability. We explain these daily changes as a consequence of the risk-sensitive foraging behaviour of the cranes. Birds shifted from a risk- prone foraging with higher intake rate in the morning to a risk-averse one with lower intake rate in the evening. \Museo Nacional de Ciencias Naturales (CSIC), José Gutiérrez Abascal 2, 28006 Madrid. Departamento de Biologia Animal, Facultad de Biologia, Universidad Complutense, 28040 Madrid, Spain. INTRODUCTION During the non-breeding season, birds must per- form few activities in order to survive, the most important being feeding, drinking, avoiding pre- dators and preening, However, since these activi- ties are usually incompatible, each individual has to make several decisions about timing, dur and sequence of these activities, to maximize fit- ness (Sibly & McFarland 1976, McCleery 1978, Caraco 1980, McNamara et al. 1987). Time alloca- tion decisions are related with the hunger state of } the individual through a feedback process, and therefore change throughout the day constituting the animal’s daily routine Individuals must also make decisions about the foraging patch, and in the case of gregarious spe- cies, about the flock size and distance to the roost. ‘These decisions are probably not independent from time allocation decisions and may also change throughout the day, According to the models of risk- sensitive foraging (Caraco et al, 1980, Stephens 1981, Stephens & Krebs 1986), birds should shift Received 18 February 1992, accepted 10 June 1992, throughout the day from a risk-prone behaviour during the morning, when they have negative ener- gy budgets, to a risk-averse behaviour during the evening, when they have positive energy budgets. Inthis paper we study the daily changes in time bud- get, feeding patch selection, flock size and food in- take rate of wintering Common Cranes Grus grus dispersing daily from a central roost to feed, We test if daily changes observed in this behaviour follow the predictions of the risk-sensitive foraging theory. Ifthe hypothesis of a progressive shift to alless risky foraging situation is correct, we should expect a de- crease in the distance to the roost, an increase in flock size, and a decrease in mean food availability of the sample of foraging patches used, throughout the day. METHODS ‘The study was carried out at Laguna de Galiocanta, NE Spain (40° 58” N, 1°30’ W), during the winter seasons 1980-81 to 1985-86, with the exception of ARDEA 80; 343-351 344 ARDEA 80 (3), 1992 Most cranes arriving from their breeding Northern Europe stage at Gallocanta some days or weeks in November-December on their way to SW Spain and a variable number remain there throughout the whole winter. During our stu- dy, yearly peak numbers of cranes staging during migration varied between 5900 (February 1983) and 21 000(February 1985), and average wintering, numbers varied between 3000 (January 1983) and 10 000 (December 1985-February 1986). We spent 1-2 days per week in the field and gathered data continuously between cranes” roost departure and roost entrance. Thus, our samples were evenly distributed throughout day and season. Flocks were located opportunely during regular circuits of the study area by vehicle. We did not re- peat the same circuit to avoid possible relationships between a particular site, habitat or daytime, and activity of the birds. All flocks were located on maps scale 1:50 000, to calculate the distance to the roost afterwards. We recorded date, time, flock size, foraging effort (measured as the percentage of birds head-down), percentage of juveniles, and type of ground (cereal sown field, cereal stubble field, sunflower stubble field, ploughed field, pas- ture, drinking place, others), We estimated the den- sity of each flock by means of the interindividual distance, measured in bird-lengths, along a transect line through the flock. Food availability was also measured in 1984-86 in stubble fields used by focal birds after finishing behavioural observations, as the mean and variance of twenty 25x25 cm random ‘samples of the number of cereal seeds found on the ground. From each flock we recorded the behaviour of 1-10 adults and 1-5 juveniles, ina number approxi- mately proportional to the relative flock size. Indi- viduals were selected randomly, but at more or less regular distance intervals along the flock’s longest diameter to avoid biases due to individual’s posi- tion in the flock (Inglis & Lazarus 1981). Birds were aged as adults or juveniles (first winter birds) on the basis of their head and neck colour (Cramp & Simmons 1980). We observed adults and juveniles in succession, in order to control for possible age differences in our sample due to flock size, food availability, ground type, or any other physical or social factor (see Alonso et al. 1987). From 1980 to 1983 focal individuals were ob- served for three minutes through 60-90x Questar telescopes. Time spent in the following activities was measured to the nearest second: feeding (head down), vigilance (head up), and preening. All other activities, including flying and resting (which in- cluded all head up postures lasting more than one minute), were also measured in the field but includ- ed as ‘others’ for analysis. During 1984-86 we re- duced observation time of focal birds to one min- ute, and recorded the rate of food intake, easily identified by the characteristic swallowing move- ment, and the number of paces. The mean dry weights of seeds were significantly different be- tween cereal stubble fields (0.027 g, n= 935 seeds from 50 different fields), cereal sown fields (0.033 8, =3000 seeds from {0 different fields), and sun- flower stubble fields (0.064 g, n = 836 seeds from 10 different fields). Therefore, we estimated food intake rate as the product of the number of seeds ingested per minute by the mean dry weight of the corresponding seed sample. We defined net intake rate as the dry weight ingested per minute spent actually feeding, i.e. head down, and absolute in- take rate as the dry weight ingested per minute of observation, We analysed the daily variation and age differences in per cent time devoted to each activity, mean duration of activity bouts in seconds, number of paces per minute, and net and absolute intake rates in grams ingested per minute, taking each one- or three-minute observation petiod as a data point. Later, we grouped data into hourly periods. Because foraging bouts frequently lasted _ more than one minute, we only used the sample of three-minute observation periods to calculate the mean foraging bout. We also calculated the accu- mulated daily food consumption multiplying each hourly value of absolute intake rate per minute by 60. Since onset and termination of cranes’ daily activity did not coincide with entire clock hours, in the first and last hourly periods the per minute rate values were multiplied by a mean fraction of those hourly periods, calculated for each month in- dependently as follows. We recorded the difference ‘Alonso & Alonso: WINTERING COMMON CRANES. between the mean roost departure and arrival times on 112 days evenly distributed throughout the six winter seasons, and subtracted 15 minutes to ac- count for the flying time from the roost to the first morning feeding sites, where we began observa- tions, plus 5 minutes for the flying time between the last evening feeding sites and the roost, esti- mated from data on individual radiotracking (own «* obs.), For a detailed description of roosting flight counts see Alonso et al. (1988). Sample sizes were, respectively for 1980-81 + through 1985-86, 427, 407, 48, 1659 and 1297 focal birds. There were no interannual differences in dai- ly or seasonal patterns of behaviour or intake rate. Thus, we combined the data from all six winter sea- sons to simplify the presentation of the results, a total of 3837 minutes for adults (2639 birds), and 1765 minutes for juveniles (1199 birds). Since all activities defined were not always per- formed by focal birds within the observation peri- od, many distributions were skewed or had a rela- tively large number of zero scores. No suitable transformation was available for distributions like these (see also Goss-Custard & Durell 1987). There- fore we analysed data by non-parametric methods (Mann-Whitney U-test or Kruskal-Wallis test), To further assess reliability of these results, subsets of these variables excluding zero scores (which were analysed separately) were also subjected to equiva- ent parametric tests (t-test and ANOVA) after loga- rithmic (bouts and intake rates) ot aresin transfi mation (per cent times), Two-tailed probability values are given for all analyses. RESULTS Time budget “The daily pattern of feeding activity showed two peaks both in adults and juveniles (Fig. 1; Kruskal-Wallis test, H = 250 in adults, H=71.6, in juveniles, both df= 10, P < 0.001). The pattern ‘was similar in all months considered independently (November-March), but more pronounced later in the season due to a more marked and prolonged through between both peaks. In adults, the two- 345 100} feeding oor Pe $ # ' type boat ¢ ° 60 + be ? + 40} + 2a} a aol vigilance E od = eet $ % % % % 4, eo % % % No °° 2% % 4o[. preening 20 + ‘, +e see eo" 4g. other activities Py) *% $ KP eho *%% 7 8 8 1 1 1 1 16 1 16 17 time of day Fig. 1, Daily variation in per cent time spent on differ ent activities by adult (filled dots) and juvenile (open dots) wintering Common Cranes. Vertical lines indicate 95% confidence intervals. n= 2639 adults and 1199 juve- niles, peak pattern was still highly significant when each month was considered separately. In juveniles, the decrease in per cent time feeding at midday was not significant in November (H = 11.4, df=9, 189 observed birds, P = 0.252), almost significant in December and January (H = 16.3, ¢f= 9, 269 birds, P = 0.060, and H = 16.7, df = 9. 216 birds, P = 0.054 respectively), and highly significant in Feb- ruaty and March (P < 0,001). In adult birds, the evening peak was higher than the morning peak in the whole winter sample and in the monthly sam- ples of February and March. Per cent time devoted to vigilance, preening and other activities increased throughout the mom- 346 ARDEA 80 (3), 1992 feeding bout a 1d 0.8) eo 2 at ' . | | t Bod +48 $ ta HET Ep hy 20 o2l Bo 1 oo 3 asf vigilance bout 1.9] 10) 4 ' + + “ peeg tt Tt les 44h y ot , oat 4 tye Tee meatday Oe Fig. 2. Daily variation in mean duration of feeding time of day bout (top, = 468) and vigilance bout (bottom, n = 1386) in adults. Vertical lines indicate 95% confidence inter- vals. ing and decreased again throughout the afternoon (Fig. 1;H =58.5, 146.3, and 38.9, 2639 adult birds, and H = 26.0, 55.3, and 28.7, 1199 juveniles, all df= 10, P < 0.001). In adult birds, the mean duration of feeding bouts showed two peaks, in early morning and late evening (Fig. 2; Fo, 45g = 2.13, P = 0.026), and the duration of vigilance bouts followed the inverse pattern (Fig. 2; Fo, 1375 = 5.22, P < 0.001). These daily variation patterns were absent in juveniles. ‘The duration of preening bouts showed no daily variation in adults or juveniles. Intake rate and pacing rate ‘The net food intake rate, i.e. intake rate while feeding, was relatively high and constant during the morning, decreased between midday and early afternoon, and later increased again to a second peak in late evening, the pattern being very similar in both age groups (Fig. 3; Fo,1450=3.48, P< 0.001 in adults, and Fy,726= 2.29, P =0.015 in juveniles). Fig. 3. Daily variation in net (top) and absolute (bot- tom) food intake rate, Filled dots = adults (n=respective- ly, 1460 and 1996), open dots = juveniles (n = 736 and 898). Vertical lines indicate 95% confidence intervals. ‘The daily pattem af absolute food intake rate was very similar to that of per cent time feeding in both adults and juveniles (Fig. 3; Kruskal-Wallis test, H=170, 1996 adult birds and H = 49.6, 898 juve- niles, both df = 10, P < 0.001). Morning net and absolute intake rates (average of first 4 hours) were significantly higher than evening intake rates (av- * erage of last 4 hours)(P <0.01 in both ages). The pattern of accumulated food ingested throughout the day increased almost linearly inNo- vember, reaching 350 g cereal in adults and 299 g in juveniles at the end of the day. The slope of the curve decreased progressively until March, when total food ingested was only 96 g in adults and 104 g, in juveniles (Fig. 4). ‘The pacing rate was studied excluding samples with feeding rate equal to zero, since it was intend- ed to represent an estimate of the feeding costs. In both age groups it was intermediate in early mom- ing, increasing until mid-morning and decreasing ‘Alonso & Alonso: WINTERING COMMON CRANES. 345 between the mean roost departure and arrival times on 112 days evenly distributed throughout the six winter seasons, and subtracted 15 minutes to ac- count for the flying time from the roost to the first morning feeding sites, where we began observa- tions, plus 5 minutes for the flying time between the last evening feeding sites and the roost, esti- mated from data on individual radiotracking (own + obs.). For a detailed description of roosting flight counts see Alonso et al. (1985). Sample sizes were, respectively for 1980-81 *, through 1985-86, 427, 407, 48, 1659 and 1297 focal birds. There were no interannual differences in dai- ly or seasonal pattems of behaviour or intake rate. Thus, we combined the data from all six winter sea- sons to simplify the presentation of the results, a total of 3837 minutes for adults (2639 birds), and 1765 minutes for juveniles (1199 birds). Since all activities defined were not always per~ formed by focal birds within the observation peri- 4, many distributions were skewed or had a rela- tively large number of zero scores. No suitable transformation was available for distributions like these (see also Goss-Custard & Durell 1987). There- fore we analysed data by non-parametric methods (Mann-Whitney U-test or Kruskal-Wallis test). To further assess reliability of these results, subsets of, these variables exchiding zero scores (which were analysed separately) were also subjected to equiva- lent parametric tests (t-test and ANOVA) after loga- rithmic (bouts and intake rates) or arcsin transfor- mation (per cent times). Two-tailed probability values are given for all analyses. RESULTS Time budget ‘The daily pattern of feeding activity showed two peaks both in adults and juveniles (Fig. 1; Keuskail-Wallis test, H = 250 in adults, H = 71.6, in juveniles, both df= 10, P < 0.001). The pattern ‘was similar in all months considered independently (November-March), but more pronounced later in the season due to a more marked and prolonged through between both peaks. In adults, the two- 100f feeding 40} + 20 aol vigilance wot oe Hy % daytime « te % % % % % 4g}. Preening 20) eee em & o Ae Pte | al. other activities 20) ope Oe 4 ner ee ¢ wh 78 9 0H 13 1% 16 17 time of day Fig. 1. Daily variation in per cent time spent on differ- ent activities by adult (filled dots) and juvenile (open dots) wintering Common Cranes. Vertical lines indicate 95% confidence intervals. n=2639 adults and 1199 juve- niles. peak pattern was still highly significant when each month was considered separately. In juveniles, the decrease in per cent time feeding at midday was not significant in November (H = 11.4, df= 9, 189 observed birds, P = 0.252), almost significant in December and January (H = 16.3, df=9, 269 birds, P = 0,060, and H = 16.7, df = 9, 216 birds, P = 0.054 respectively), and highly significant in Feb- tuary and March (P < 0.001). In adult birds, the evening peak was higher than the morning peak in the whole winter sample and in the monthly sam- ples of February and March. Per cent time devoted to vigilance, preening and other activities increased throughout the mom- 346 ARDEA 80 (3), 1992 feeding bout " Hll gltyatid yy aa { tt aa fT yy ol t oa Teg tte tel) eta ge Soa PHT 4 4, ¢ . 02 why? Fig. 2. Daily variation in mean duration of feeding bout (top, n= 468) and vigilance bout (bottom, » = 1386) in adults, Vertical lines indicate 95% confidence inter~ vals. ing and decreased again throughout the afternoon (Fig. 1; H=58.5, 146.3, and 38.9, 2639 adult birds, and .0, 55.3, and 28.7, 1199 juveniles, alt df= 10, P < 0.001). In adult birds, the mean duration of feeding bouts showed two peaks, in early morning and late evening (Fig. 2; Fo, 453 = 2.13, P = 0.026), and the duration of vigilance bouts followed the inverse patiern (Fig. 2; Pio, 1375 = 5.22, P < 0.001). These daily variation pattems were absent in juveniles. ‘The duration of preening bouts showed no daily variation in adults or juveniles. Intake rate and pacing rate The net food intake rate, i.e. intake rate while feeding, was relatively high and constant during the morning, decreased between midday and early afternoon, and later increased again to a second peak in late evening, the pattern being very similar in both age groups (Fig. 3: Fo, 4s= 3.18, P< 0.001 in adults, and Fo, 795= 2.29, P= 0.015 in juveniles). 7 8 9 0 1 12 15 14 15 16 17 time of day Fig. 3. Daily variation in net (top) and absolute (bot- tom) food intake rate. Filled dots =adults (n=respective- ly, 1460 and 1996), open dots = juveniles (n = 736 and 898). Vertical lines indicate 95% confidence intervals. The daily pattern of absolute food intake rate was very similar to that of per cent time feeding in both adults and juveniles (Fig. 3; Kruskall-Wallis test, H= 170, 1996 adult birds and H = 49.6, 898 juve- niles, both df = 10, P < 0.001). Morning net and absolute intake rates (average of first 4 hours) were significantly higher than evening intake rates (av- erage of last 4 hours)(P < 0.01 in both ages) The pattern of accumulated food ingested throughout the day increased almost linearly in No- vember, reaching 350 g cereal in adults and 299 g in juveniles at the end of the day. The slope of the curve decreased progressively until March, when total food ingested was only 96 gin adults and 104 g in juveniles (Fig. 4). ‘The pacing rate was studied excluding samples with feeding rate equal to zero, since it was intend- ed to represent an estimate of the feeding costs. In both age groups it was intermediate in early morn- ing, increasing until mid-morning and decreasing ‘Alonso & Alonso: WINTERING COMMON CRANES. 347 adults Aor 300) ‘ee ‘an 200) 100} total ingested (9) juveniles 300] Jyov Bee 200 Jan Feo 100 Siar OF 8 FT Is aS 6 7 8 time of day Fig, 4, Accumulated food intake throughout the day between November and March, Sample sizes were, re- spectively for November to March: 409, 592, 310, 443 and 242 for adults, and 168, 216, 128, 191 and 195 for juveniles, later to minimum values in late evening (Fig. 5; Kruskall-Wallis H = 38.4, in adults and 39.6 in ju- veniles, both df= 10, P < 0.001). Distance to roost, foraging ground and intake * rate ‘The mean distance between foraging flocks and the roost decreased regularly throughout the day (Fig. 6; Fo, 2s = 20.33, P< 0.001). Cranes used ce- real stubble and sown fields as their main feeding grounds, Use of sunflower stubbles was only resid- ual (2.1% of 513 foraging flocks, and restricted to November and December) (see also Alonso et al. 1984), The type of foraging ground changed signif- icantly throughout the day: cranes preferred grounds with food on surface during the morning and sown 25) ; 20} 8 fb ttha dy Bo tt hat ; ; Has © 5 ) 7° 8 9 10 11 12 13 14°15 16 17 time of day Fig. 5. Daily variation in number of paces per minute. Samples with food intake rate equal to zero were exclud- ed. Filled dots = adults (n= 1516), open dots = juveniles (n= 769). Vertical lines indicate 95% confidence inter- vals, grounds, in spite of having to dig up the seeds, in the evening (Fig. 7; x2 = 324.6, df= 9, P <0.001). ‘The flock feeding effort increased with the dis- tance to roost (Fy2434 = 3.78, P < 0.001). It was higher on sown grounds with no seeds on surface (85.2%, SE = 1.8, n = 125) than on grounds with food on surface (stubble fields plus sown grounds with seeds on surface, 76.5%, SE = 1.6, n = 236; 1=3.45, P <0.001). Flock size and interindividual distance ‘The mean flock size increased during the morn- ing, reaching a peak during midday, at drinking- ee ee ee _ a distance to roost (km) a time of day Fig. 6. Daily variation in the distance from foraging flocks to the roost. Vertical lines indicate 95% confi- dence intervals. n = 930 flocks. 348, ARDEA 80 (3), 1992 100 % 75 50} 25 dl 8 9 10 1 12 13 4 15 16 17 time of day Fig. 7. Daily variation of the percentage tse by crane flocks of foraging grounds with (shaded bars) and with- out (white bars) food on surface. n = total number of fields sampled. Grounds with food on surface include cereal and sunflower stubbles and cereal sown grounds with seeds on surface; grounds without food on surface are sown fields. resting sites. Later, flocks decreased during early afternoon and increased again to reach highest values in late evening (Fig 8a). Mean interindivid- ual distance within flocks decreased throughout the day, following an inverse pattern to that of flock size (Fig. 8b). DISCUSSIO! ‘The mainactivity of cranes during winter was feed- ing, to which they devoted 62% of their daily time budget. The daily patterns of percentage time spent feeding, feeding bout length, and food intake rate showed two peaks, in early morning and late eve- ning. These double-peak patterns of feeding effort have probably evolved as an adaptation to compen- sate for the prolonged fasting imposed by winter nights and is now endogenous in many bird species (Beer 1961, Aschoff 1966, 1967, Meier & Russo 1985), Like in these species, hungry cranes filled theirempty crops as soon as possible in early momn- ing, and in late evening they refilled them to anti- cipate fasting during night. Water needs may have also been important in determining a double-peaked daily pattern in feed- ool @ { flock size 8 0 4 Fe? a 8 9 10 1 12:13 14 15 16 17 time of day Fig. 8. Daily variation in mean flock size (A, n = 930 flocks) and density (B, n = 510 flocks). Vertical lines in- dicate 95% confidence intervals, ing activity. Recent studies of water economy in granivorous birds have shown that even during winter a period of negative water balance usually occurs during midday (MacMillen 1990). In fact, the double-peaked average pattern observed re- sulted from most cranes stopping to feed on cereal fields at midday for some time, which they devoted to drinking, bathing and preening. Frequently they had to fly a considerable distance to reach one of - the few drinking sites available in our study area (see Alonso et al. 1984). ‘We think that the lower feeding effort around - midday was not due to a digestive bottleneck as proposed for some birds (e.g. Kenward & Sibly 1978, Zwarts & Dirksen 1990), since cranes were able to continue feeding at high rates during mid- day in November and December (see Fig. 4). Several authors have discussed the significance of the ‘bigeminus’ versus ‘alternans” patterns, ie. respectively with a more pronounced morning or evening peak. Some species living in northern lati- tudes show a higher feeding activity peak in eve- Alonso & Alonso: WINTERING COMMON CRANES 349 ning, which has been interpreted as an adaptation to reduce energy losses during the extremely cold morning temperatures (Gjerde & Wegge 1987, and references therein). These cases are probably ex- ceptioas, since most birds show higher morning ac- tivity peaks (Aschoff 1966, 1967). Wintering cranes had similar feeding efforts during the morning and evening periods, excepting in February and March. ‘The differing pattem in these two months was prob- ably determined by the arrival of large numbers of migrating cranes during the afternoon. After land- ing, cranes foraged intensively to compensate for the time they had spent flying, increasing the aver- age evening feeding peak. However, in spite of a globally greater feeding effort during the evening, net and absolute intake rates were higher during the morning. The main reason for the higher morning net intake rate was probably the higher average quality of the moming foraging grounds (stubble and sown grounds with food on surface) with respect to the evening fora- ging grounds (when the use of sown grounds with buried seeds increased). Comparison of daily curves of net and absolute intake rates suggests that birds maintained the net intake rate (ic. when searching food, head down), probably as high as possible, with the only constraint of the availability of food on the ground. On the evening sown grounds, although cranes spent less energy walk- ing, they had to dig up the cereal seeds from the earth, and thus spent more time and probably also more energy than during morning foraging. One benefit of these afternoon flocks was a reduced individual vigilance time due to the large flock size, and thus an increased time head down, searching food. But this compensated only in part for the low- er accessibility of buried seeds on sown grounds. ‘The daily changes in time allocation decisions were therefore related with those about feeding patch and flock size, and altogether with the pattern of daily intake rate. During the moming cranes showed a higher locomotor activity, foraging in smaller flocks, at longer distances from the roost, on grounds with higher food availability, in which cereal seeds were easily obtained from the surface. Onthe contrary, during the afternoon, birds showed lower locomotor activity, aggregated in larger flocks, foraged closer to the roost and used more frequently sown grounds, which offered a lower but less variable food availability. These results suggest that the increasing degree of satiation with advancing daytime relaxed the dispersal trend of the birds, which foraged progressively closer to the roost and formed larger flocks on grounds with lower but more constant food availability, thus se- ducing other possible non-foraging risks. This dai- ly shift from a more risky but higher rewarding for- aging behaviour in the morning to a less risky and less rewarding one in the evening, as birds become satiated, agree with the predictions of risk-sensitive foraging models (Caraco et al. 1980, Stephens 1981, Stephens & Krebs 1986) These results also show how individual cranes can obtain similar food intake rates in two situa- tions differing in time budget, flock size, foraging ground type and distance to the dispersal center, i.e. during moming and evening feeding periods. This is relevant when studying the relationships be- tween intake rate and flock size or other social and behavioural variables, which are often assumed to be unimodal by simple foraging models (see Stephens & Krebs 1986). Finally, there was a steep seasonal decrease in feeding effort and daily food intake. Since the diet composition did not change throughout the season, we interpret this as a consequence of the increased energetic demands of birds recently arrived on mi- gration from the north, Cranes showed then a hy- perphagia similar to that observed in most migrato- ry birds just before migration departure (e.g. King & Farner 1965). Data (own obs.) on seasonal changes in body weight (lowest during autumn ané reaching highest values already in January) support the idea that cranes acquire fat reserves as soon as possible over the winter period and later they just feed to maintain appropriate intake levels. The ex- istence metabolism (EM) of a crane, considering 5000 g as its average weight (W), 0° C mean tem- perature, and a 10-hour photoperiod, is EM = 17.73W 03316 (Wiens & Dyer 1977) which is equal to 1.64 MJ bird-! day-1. Assuming a caloric vatue of 18.2 kJ g-! for cereal seeds (Cummins & Wuy- 350 ARDEA 80 (3), 1992 check 1971) and 77.7% assimilation efficiency (Castro ef al. 1989), a crane is expected to ingest around 116 g per day. This figure is much smaller than our field estimation of 350 g in November, and only somewhat higher than the estimated 96 g in March. This result also suggests that cranes progressively decrease fat deposition from highest values in November to a minimum in March, ‘The progressive fattening throughout the win- ter is probably the reason that we did not observe a premigratory increase in food intake such as that observed in many other species. The alternative explanation that the observed pattern could be due to seasonal food depletion is indeed excluded by the high percentage time spent in non-feeding activities in February and March (respectively 47.1 and 43.3%, compared to 27.1, 31.4 and 42.7% in November, December and January). ACKNOWLEDGEMENTS We thank A.C. Perdeck and S. Dirksen for their com- ments on an earlier version of the manuscript. Financial support was provided by Direccién General de Medio Ambiente, Instituto Nacional para la Conservacién de la Naturaleza, and Direccién General de Investigacin Cientifica y Técnica. This is a contribution to DGICYT- CSIC Project PB87-0389. REFERENCES Alonso, J.A.,J.C. Alonso & J.P. Veiga 1984. Winterfeed- ing ecology of the Crane in cereal farmland at Gallo- canta, Spain. Wildfowl 35:119-131. Alonso, J.A., J.C. Alonso & J.P. Veiga 1985. The influ- ence of moonlight on the timing of roosting flights in Common Cranes Grus grus. Ornis Scand. 16: 314-318. Alonso, J.C., J.A. Alonso & J.P. Veiga 1987. Flocking in wintering Common Cranes Grus grus: influence of population size, food abundance and habitat patchiness. Omnis Scand, 18:53-60. Aschoff, J. 1966. Circadian activity patterns with two peaks. Ecology 47:657-662. Aschoff, J. 1967. Circadian rhythms in birds. Proc. int. om. Congr. 14:81-105, Beer, J.R. 1961. Winter feeding patterns of the House Sparrow. Auk 78:63-71 Caraco, T. 1980. On foraging time allocation in a sto- chastic environment, Ecology 61:119-128. Caraco, T., $. Martindale & T.S. Whitham 1980. An empirical demonstration of risk-sensitive fora preferences. Anim, Behav. 28: 820-830 Castro, G., N. Stoyan & J.P. Myers 1989. Assimilation efficiency in birds: a function of taxon or food type? Comp. Biochem. Physiol. 92A:271-278 ‘Cummins, K.W. & J.C. Wuycheck 1971 Caloric equiva- lents for investigations in ecological energetics. Verh. int, Verein. theor. angew. Limnol. 18:1-152. Cramp, S. & K.E.L. Simmons (eds) 1980. The Birds of the Western Palearctic, Vol. 2. Oxford Univ. Press, Oxford, Gjerde, I. & P. Wegge 1987. Activity patterns of Caper- caille, Tetrao urogallus, during winter, Holarct. Ecol. 10:286-293. Goss-Custard, J.D. & S.E.A.Le V. Dit Durell 1987. Age- related effects in oystercatchers Haematopus ostralegus, feeding on mussels Mytilus edulis. U. Aggression. J. Anim. Ecol. 56:537-548. Inglis, LR. & J. Lazarus 1981. Vigilance and flock size in brent geese. The edge effect. Z. Tierpsychologie 57:193-200. Kenward, R.E. & R.M. Sibly 1978. Woodpigeon feeding, behaviour at Brassica sites. A field and laboratory investigation of Woodpigeon feeding behaviour during adoption and maintenance of a Brassica diet. Anim. Behav. 26:778-790. King, J.R. & DS. Famer 1965. Studies of fat deposition, in migratory birds. Ann, New York Acad. Sci. 131: 422-440, MacMillen, R.E. 1990, Water economy of granivorous birds: a predictive model. Condor 92:379-392. McCleery, RH. 1978. Optimal behaviour sequences and decision making. In: Krebs, J.R. & N.B. Davies (eds) Behavioural Ecology. An Evolutionary Ap- proach: 377-410. Blackwell, Oxford. McNamara, J.M., RH. Mace & A.1. Houston 1987. Optimal daily routines of singing and foraging in a bird singing to attract a mate. Behav. Ecol. Socio- biol. 20:399-40s. Meier, A.H. & A.C. Russo 1985. Circadian organization of the avian annual cycle. In: Johnston, R.F. (ed.), Current Ornithology, vol. 2: 301-343. Plenum Press, New York. Sibly, RM. & D.J. McFarland 1976. The fitness of be- haviour sequences, Am. Nat. 110:601-617. Stephens, D.W. 1981. The logic of risk-sensitive fora- ging preferences. Anim, Behav. 29: 628-629. Stephens, D.W. & JR. Krebs 1986. Foraging Theory. Princeton Univ. Press, Princeton. Wiens, ILA. & M.L. Dyer 1977. Assessing the Povental impact of graniverous birds in ecosystems. I Pinowski & S.C. Kendeigh (eds) Granivorous Birds Alonso & Alonso: WINTERING COMMON CRANES. 351 in Ecosystems: 301-340. Cambridge Univ. Press, Cambridge. ‘Zwarts, L.Z. & S. Dirksen 1990, Digestive bottleneck Jimits the increase in food intake of Whimbrels pre- paring for spring migration from the Bane d” Arguin, Mauritania. Ardea 78:257-278. SAMENVATTING In Gallocanta, noordoostelijk Spanje, werden overwin- terende Kraanvogels bestudeerd. Dit artikel gaat over de. dagelijkse activiteitspatronen van Kraanvogels in relatie tot kenmerken van de fourageerplaats, de afstand tot de slaapplaats en de groepsgrooite. De hoeveelheid tijd be- steed aan voedselzoeken, de duur van ‘vlagen’ van voed- selzoek-gedrag en de voedselopname-snelhieid vertoon- de twee pieken per dag: én in de vroege ochtend en een tweede in de late namiddag (Figs 1-3). De tijd bested aan waken, poetsen en andere activiteiten vertoonde én piek rond het middaguur (Figs 1 & 2), wanneer de vogels, zich gewoonlijk ophielden bij de drinkpteatsen, Het pa- troon van deze activiteiten was dus tegengesteld aan dat van het patroon van de fourageergedragingen. De mobiliteit van de kraanvogels was vooral ’s och- tends hoog (Fig. 5). Er werd dan ver weg van de roest- plaats (Fig. 6) gefourageerd op plaatsen met een hoog voedselaanbod (Fig. 7) in betrekkelijk kleine groepen (Fig. 8). In de late namiddag was het beeld omgekeerd. De opnamesnelheid verandert daarbij nauwelijks. Dit patroon werd beschouwd als een verschuiving van ris- kante naar minder riskante fourageerstrategieén in de Joop van de dag, naarmate de honger afneemt. De waar- genomen sterke afname in de totale voedselopname per vogel per dag van november tot maart (Fig. 4) werd toe~ geschreven aan de zeer grote voedselbehoefte van pas aangekomen migranten die in een zo vroeg mogelijk sta- dium reserves proberen op te bouwen.

Anda mungkin juga menyukai