The size of the Common Crane (Grus grus) population wintering in Spain was estimated around 10000-15000 birds by Bernis (1960,1966). In 1979-80 the first census covering the whole wintering range of the species in Spain was carried out, with a maximum count of 14721 cranes on 8-10 February 1980 (Fernández et al. 1981). Araújo (1987) and Fernández et al. (1987) increased this figure to an estimated 17000 birds, adding a count at Gallocanta lake (NE Spain) plus the cranes already seen in France by that date. However, the latter estimation probably included wintering birds from Portugal and North Africa. Later, based on a count of 31945 cranes at Gallocanta during spring migration 1985 (Alonso et al. 1986a), the size of the crane population migrating through Western Europe was initially estimated around 40000-50000 birds (Alonso et al. 1986b, Alonso and Alonso 1988).

Methods
During the winter season 1987-88 we carried out 3 main censuses of the crane wintering population, on 27-28 November 1987, 8©9 January and 26-27 February 1988, surveying respectively 44, 61 and 36 crane wintering sites. These sites include all areas known since the first national survey (see Ferná ndez et al. 1981) plus a number of new areas discovered later. Additionally, several complementary counts were performed at most sites throughout the winter season. Cranes were counted while leaving or entering the roosts or, when this was not possible, on the feeding grounds. All observers knew the sites previously and had experience in counting cranes.

For the purposes of this study we defined a crane wintering site as an area traditionally used by cranes during the whole or part of the winter season, and where the birds showed clear daily routines between one or more roosts utilized regularly and the feeding grounds. This definition includes all wintering sites and those stopover sites which are regularly used each year during autumn and/or spring migrations, but excludes areas occupied irregularly during one or more days due to e.g. bad weather conditions (see Muntaner and Congost 1979, Rebollo 1985, Gómez and Díes 1987). The name assigned to each site was that of the nearest town, except in some cases where the name of the nearest river, lagoon or water reservoir was preferred. The sites "Lagunas de La Mancha" and "Navalvillar de Pela" include groups of cranes which use several roosts but forage on the same grounds and thus apparently behave as a single populational unit. In general, neighbouring sites are separated from each other by some geographical feature such as a mountain chain, an extent of unsuitable habitat or simply by a too large distance to be covered regularly by the birds, so that there is apparently no regular exchange of cranes between them.

During the winter 1988-89 we censused weekly the staging population at Gallocanta and considered simultaneous counts made in France and SW Spain.


Size of the population migrating through Western Europe

The maximum censuses of the crane wintering population were 54114 and 51466 birds, obtained respectively on 23 November 1989 and on 10 November 1988, during peak passage through Gallocanta. These figures are also, as far as we know, the maximum numbers of cranes ever counted simultaneously in the western migratory route of this species, thus representing the minimum possible size of the crane population using this route. The similarity between two figures obtained at Gallocanta in autumn 1988, with only four days interval (50768 and 51466 birds respectively on 6 and 10 November 1988), and that counted in autumn 1989, virtually excludes the possibility of a significant error in any of them. The extraordinary concentrations of cranes at Gallocanta were the result of an advanced and fast migration through Western Europe these years (Petit pers. com.).

Since some cranes had already departed from Gallocanta southwards when the peak figure was recorded, the size of the Common Crane population migrating through Western Europe and Northwest Africa could be estimated around 60000-70000 birds (see also next paragraph), i.e. a somewhat higher figure than that estimated by us previously (Alonso et al. 1986a,b, Alonso and Alonso 1988).

The difference between our last count and earlier censuses or estimations (Fernández et al. 1981, 1987, Araú jo 1987) could hardly be attributed to an increase in the total population size in such a short time, provided the low annual recruitment rate of this population (Alonso et al. 1990). It could be due to a shift in the migratory corridor from East to West, i.e. to a number of cranes having changed their earlier and traditional migratory route (possibly over Italy to Tunisia or Algeria) by a recent route passing through Gallocanta. However, the scarcity of crane migration observations over Italy and Tunisia and the few available data of wintering cranes in this last country (Urban et al. 1986 and pers. com., pers. obs. during a visit in January 1989) suggest that this route may have been of only marginal importance during at least the last decades and thus irrelevant in the differences observed in recent censuses in Spain. We rather think that these differences reflect a combination of  (a) an increased census accuracy,  (b) a betterÔ knowledge of the wintering areas, and  (c) a higher degree of

concentration of the cranes in time and space during the last migratory seasons.


Census of cranes wintering in Spain

With respect to the number of cranes spending the whole winter season in Spain, only counts made between mid December and early February should be considered, since only during this period is the migratory activity of the species reduced to a minimum. The maximum figure was censused in January 1988, with a total 39579 birds. The other two censuses carried out that winter resulted in lower numbers  due mainly to the lower number of sites surveyed (respectively a 28% and a 41% less than in January).

Assuming a certain error in minus in the January count, the crane population wintering in Spain that year could be estimated around 45000 birds. An estimation of the total number of cranes wintering in the Western Mediterranean region would also include the 4847 cranes that remained in France that year (CROAP 1989), the 1500 cranes estimated for Portugal (Pinto and Almeida 1989) and the 2000 birds estimated for Morocco (Thé venot 1985 a,b, Thévenot and Salvi 1987), although both latter figures should be considered as minimum estimates, which should be increased by a percentage error difficult to state precisely but surely higher than that assumed for the Spanish census (see also Urban 1986, Urban et al. 1986, Alonso and Alonso 1988). The estimation for the crane population migrating through Western Europe and Northwest Africa would again be around 60000©70000 birds, the same as that made from the peak counts at Gallocanta on November 1988 and 1989 (see previous paragraph).

These data support the conclusion first stated by Bernis (1960, 1966) that most Common Cranes using the western route winter in Spain, in spite of most older references (Irby 1895, Verner 1909, Hartert 1910-22, Bedé 1926, Libbert 1936, 1938, Hortling 1939, Heim de Balsac et al. 1962, Moreau 1967) and even some relatively recent revisions (Makatsch 1970, Glutz et al. 1973, Walkinshaw 1973) in which Northwest Africa was assumed as the main wintering area of the species.

Particularly in 1987-88, around 75% of the Common Crane western population spent the winter in Spain. However, the number of cranes wintering in France that year may be considered as exceptional (Riols 1987, Petit pers. com.). Some of these birds probably winter in Spain in other years, thus resulting in the absolute number of birds wintering in Spain and the percentage with respect to the total population being somewhat higher than those given here.

Our estimation of the Common Crane population migrating through the western route is slightly higher than previous own (Alonso et al 1986a,b, Alonso and Alonso 1988) and other authors' estimations (48000-54000 cranes, based on a census of 45000 birds in the German Democratic Republic, Mewes and Prange 1988, although these figures do not completely exclude possible cranes from the central migration route).


Wintering distribution in Iberia

A total of 63 sites are currently known to be regularly used by wintering cranes in the Iberian Peninsula (59 in Spain, 2 in Portugal and 2 comprising areas in both countries). The highest density of sites corresponds to Badajoz and Cáceres provinces (Extremadura), respectively with 36.8% and 25% of all sites, followed by CÀ; rdoba province with 4.4%. Also Badajoz, Cáceres and Córdoba had the highest numbers of wintering cranes in 1988 (respectively 39%, 26% and 18% of the January count). The site used by a highest average number of cranes through the whole winter season is, however, Gallocanta lake (Zaragoza and Teruel provinces, in Aragón), where between 2000 and 10000 cranes have spent the winter during the last 10 years, and ca. 90% of all birds using the western route are thought to stopover during migration, having reached peak numbers of 54114 in autumn (23 November 1989) and 28660 in spring (8 March 1987).


Conservation conclusions

Eighty seven per cent of the crane wintering areas in Spain are legally unprotected, totalizing a 80% of the 795301ha occupied by the species in our country. In the remaining 13% of the areas protection does only affect part of their surfaces, although it includes the main roosts.

The main habitat is open Holm Oak (Quercus rotundifolia) wooded area (the Iberian "Dehesa") in 64% of the areas (51% of them with cereal crops as a substrate, and 49% with pastureland). In the remaining 36% areas the main habitat is cultivated farmland, either with cereal crops -mostly wheat and barley- (91% of them) or irrigated crops -mostly corn- (9%).

The mean number of cranes per wintering in an area is 850 birds, and the modal number, 200 birds. This is due to a 43% of the cranes gathering at only 4 areas with over 3000 birds each. This distribution pattern has probably resulted from a progressively higher use of cereal as food resource throughout  the last decades. Today, the diet of ca. 40% of the wintering cranes (and even more during migration periods) is based on cereal grain.

There is a direct correlation between the logarithm of the number of cranes wintering at a site and the logarithm of the area used by them. Five hundred cranes occupy ca. 10000ha.

Not disregarding the protection that should be ensured by the conservation authorities at those staging and wintering areas with the highest crane concentrations during migratory periods, it could be worth trying to restore and/or protect: the crane areas forming the primitive western migratory route in Sapin; wintering and stopover sites at the periphery of the wintering range in our country; and all traditional wintering areas on Holm Oak woods in SW Spain.