Lead Shot Densities in a Site of Grit Ingestion for Greylag Geese Anser anser in Doñana ()

Rafael Mateo1*, Andreu Bonet2, Joan Carles Dolz3 & Raimon Guitart1 1Laboratory of Toxicology, Faculty of Veterinary, Autonomous University of Barcelona, 08193 Bellaterra, Spain. 2Department of Ecology, University of Alacant, P.O. Box 99, 03080 Alacant, Spain. 3Conselleria de Medi Ambient, Generalitat Valenciana, Churruca 29, 03071 Alacant, Spain. *e-mail: [email protected] SUMMARY Cerro de los Ánsares is a mobile dune in the Doñana National Park where wintering greylag geese (Anser anser) fly every morning to ingest grit. Hunting was allowed until 1983, generating accumulation of lead shot in the dune. We studied shot density in Cerro de los Ánsares by two methods, one on the surface by visual examination of squares of 1 m2 (n=100), and another by taking core samples of 20 cm of depth (n=100). Cores were sieved in two portions corresponding to depths of 0-10 cm and 10-20 cm. Shot density on the surface was about 5 shot/m2, but rose to 16.2 shot/m2 in the 20 cm column. Shot density was higher near the base of the dune than near the crest. This could be explained by the movement of the dune, which leaves behind shot pellets, or by a higher grit search and consequently shot ingestion by greylag geese near the crest of the dune because it is safer than the base. Grit found in the gizzard of greylag geese hunted in the Marshes around Doñana National Park had more particles >1.5 mm than the sand of Cerro de los Ánsares. This can be explained by an active selection by the geese of the largest particles, where shot are included, or by a higher retention of these particles in the gizzard. Lead shot (26 kg/ha) should be extracted from Cerro de los Ánsares, by manual or mechanical methods, to reduce lead poisoning in Greylag geese in Doñana. Keywords: Lead, shot, grit, greylag goose, Doñana. RESUMEN El Cerro de los Ánsares es una duna móvil del Parque Nacional de Doñana donde los ánsares comunes (Anser anser) invernantes acuden cada mañana para ingerir gastrolitos. La caza fue permitida allí hasta 1983, generando una acumulación de perdigones de plomo en la duna. Hemos estudiado la densidad de perdigones en el Cerro de los Ánsares mediante dos métodos, uno sobre la superficie por examen visual de cuadros de 1 m2 (n=100), y otro tomando columnas de arena de 20 cm de profundidad (n=100). Dichas columnas fueron cribadas en dos porciones correspondientes a profundidades de 0-10 cm y 10-20 cm. La densidad de perdigones sobre la superficie estuvo alrededor de 5 perdigones/m2, pero alcanzó los 16,2 perdigones/m2 en el total de los 20 cm de profundidad. La densidad de perdigones fue mayor cerca de la base de la duna que en las zonas cercanas a la cresta. Esto puede ser debido al movimiento de la duna, que deja los perdigones atrás, o a una mayor búsqueda de gastrolitos y consecuentemente de ingestión de perdigones por parte de los ánsares comunes cerca la cresta de la duna por ser un lugar más seguro que la base. Los gastrolitos encontrados en la molleja de ánsares comunes cazados en las Marismas del Guadalquivir, en el entorno del Parque Nacional de Doñana, correspondieron a partículas >1,5 mm en mayor proporción que en la arena presente en el Cerro de los Ánsares. Esto puede ser debido a una selección activa por parte de los ánsares comunes de estas partículas más grandes, donde se incluyen los perdigones, o a una mayor retención de estas partículas en la molleja. Una solución adecuada para la intoxicación por plomo en el ánsar común en Doñana sería la extracción de los perdigones de plomo (26 kg/ha) del Cerro de los Ánsares mediante métodos manuales o mecánicos. Palabras clave: Plomo, perdigón, gastrolitos, ánsar común, Doñana.

INTRODUCTION The first report of lead poisoning in the Guadalquivir most affected by lead poisoning in the Palearctic flyway Marshes (Spain), also known as Doñana, was an epiz- (Pain, 1991; Mateo et al., 1997). ootic outbreak of greater flamingos (Phoenicopterus ruber) Lead shot densities in sediment of two zones of the that appeared in Lucio de Marilópez in 1991 (Ramo et Guadalquivir Marshes, Lucio de Caraviruelas and al., 1992). Later, ingestion of lead shot was detected in Hato Blanco rice fields were 14 shot/m2 and <1 shot/m2 10% of 20 greylag geese (Anser anser) shot by hunters in the upper 15 cm respectively (Mateo et al., 1998). (Mateo et al., 1998). Moreover, 28% of 46 greylag geese These densities were lower than those described in other found dead had ingested lead shot, and most of them Spanish such as the Natural Parks of the Ebro (20%) died by lead poisoning according to necropsy find- Delta, L’Albufera de València or El Fondo, where there ings. Lead poisoning was the second leading cause of are as many as 163-300 shot/m2 (Bonet et al., 1995; Mateo death in this sample, after hunting injuries (Mateo et al., et al., 1997, 1998). In another , Tablas de Daimiel 1998). The prevalence of lead shot ingestion in geese National Park, where hunting has been banned since species from North America is usually below 10% (San- 1965, approximately 100 shot/m2 remain in the upper 20 derson and Bellrose, 1986; DeStefano et al., 1991), and cm of sediment (Mateo et al., 1998). the prevalence in greylag geese in the United Kingdom Although lead shot densities as low as 0.7 shot/m2 is 7.1% (Mudge, 1983). We also found a lead poisoned were enough to produce massive mortalities in Canada Northern pintail (Anas acuta) in the Guadalquivir Geese in Colorado (Szymczak and Adrian, 1978), high marshes (Mateo et al., 1998), which is one of the species rates of lead shot ingestion in greylag geese in the

76 Guadalquivir Marshes could be better explained by the lometry of grit from the gizzards was done across sieves accumulation of shot in a dune called Cerro de los of 4, 3, 2, 1.5, 1 and 0.5 mm of mesh size. Ánsares (“Geese’s Hill”) rather than in flooded areas. We used ordinary kriging to map lead shot densities Large flocks of greylag geese wintering in the on the surface. This best linear unbiased estimator Guadalquivir Marshes fly to this dune to ingest grit every method uses spatial linear interpolation by means of morning (Sánchez et al., 1977). Hunters exploited this autocovariance matrix applied to the grid of data to mini- behaviour to shoot the geese, an activity that has re- mize the variance of the errors (Isaak and Srivastava, sulted in an accumulation of lead shot pellets at this site. 1989). Graphics have been softened with an algorithm The purpose of this study was to improve the knowledge that assumes linear variograms generated by the data available (Calderón et al., 1996) about density and obtained in two axes of coordenates (Bonet et al., 1995). distribution of lead shot pellets in this gritting site for We studied the distribution of shot densities on the geese. This information can be essential when it comes surface of the dune, with transects (n=10) and height to applying measures to alleviate lead poisoning in greylag of the dune (n=10) as class factors of a general linear geese in the Guadalquivir Marshes. model. In order to stabilize variances, shot number was transformed to the square root of its sum with 3/8 (Zar, MATERIAL AND METHODS 1996). Shot distribution in depth was also studied using Cerro de los Ánsares is a mobile dune located in the the data obtained in core sampling by including this fac- southern part of the Doñana National Park and is one of tor (0-10 cm and 10-20 cm) in the model. The Mann- the furthest inland eolian dunes of a system which cov- Whitney test was used to compare the weights of lead ers an area 30 km long and 3-5 km wide and migrates shot collected by the square and core methods. from the Atlantic Ocean towards the marsh (Llamas, 1990). These dunes have an average speed of 5-6 m/ RESULTS year and their shape is flattened on the side toward the Mean shot density on the surface obtained with both sea but more abrupt on the side toward the land (García- square and core methods was about 5 shot/m2. However, Novo et al., 1976). We studied lead shot densities on the most of the shot were buried (Table 1). According to the sea side or tail of the dune with two methods on Febru- square method, higher densities appeared in the eastern ary 1997, one on the surface and another up to 20 cm of half of the studied area, but differences among perpen- depth. Densities on the surface of the dune were studied dicular transects to the crest of the dune were not statis- by counting shot in 100 squares of 1 m2 which were lo- tically significant (Figure 1). However, there were sig- cated on a grid of 20x20 m and covered an area of 3.24 nificant differences related to the height (F =3.49, 9, 81 ha. Core samples of 20 cm of depth were collected with p=0.001), with the highest densities found on the base a PVC tube of 11.9 cm of diameter in the NE corner or tail of the dune (Figure 1). We also found this trend outside each square. The number of shot on the surface with the core sampling (Figure 2), but differences among of each core was noted. These samples were divided in heights were not significant (p=0.053). With this two parts of 0-10 cm and 10-20 cm of depth and were filtered with water across a sieve of 1 mm of mesh size. Figure 1: Map of shot densities (shot/m2) in Cerro de los Ánsares Shot were weighed. All the sand particles retained in made with the ordinary kriging method. the sieve were pooled, dried, and sorted with sieves of 4, North-Crest of the dune 3, 2 and 1.5 mm of mesh size. This was compared with 180 the size of grit in the gizzards of greylag geese (n=20) collected in a previous study (Mateo et al., 1998). Granu- 160

Table 1. Shot densities (shot/m2) estimated in the mobile dune 140 Cerro de los Ánsares using two methods. Squares of 1 m2 (n=100) to estimate surface density and cores of 120 11.9 cm of diameter (n=100) for the upper 20 cm of sand (including the surface) were studied. 100

D 'HSWK 0HWKRG 6KRWIRXQG 0HDQ±VH  5DQJH 80

6XUIDFH 6TXDUH  ±  60 6XUIDFH &RUH  ±  FPE &RUH  ±  40 FP &RUH  ±  20 6XUIDFHFP &RUH  ± 

D 0 $ULWKPHWLFÃPHDQñÃVWDQGDUGÃHUURUà 0 20 40 60 80 100 120 140 160 180 E ([FOXGLQJÃWKHÃVXUIDFHà South - Base of the dune

Ecotoxicology and Environmental Restoration 3(2) 2000 77 Figure 2: Depth of shot found at different distances from the Figure 3: Grit in the gizzard of hunted greylag geese (n=20) in base in Cerro de los Ánsares (0-10 cm does not the Guadalquivir Marshes, around Doñana National include shot found on the surface). Park.

method, shot densities were higher in 0-10 cm of depth the average speed of this dune system is about 5-6 m/ than in 10-20 cm (F =5.32, p=0.022). Moreover, year (García-Novo et al., 1976), it seems normal to find 1, 171 there was a significant interaction between depth and the highest densities more than 200 m behind the present height of the dune (F =2.28, p=0.020). In fact, shot crest. There can be another factor, such a higher grit 9, 171 present in the top 10 cm of sand were more abundant in and shot ingestion by greylag geese near the crest of the the lower half of the dune (92.3%) than in the upper dune, because it is safer than closer to the base. It could half (40%) (Figure 2). also explain the lower presence of shot on the surface of Lead shot found with the square method, on the sur- the higher half (Figure 2). Wintering census of greylag face, weighed less (n=398, median=0.138 g, geese in Doñana range from 40,000 to 80,000 (Sánchez range=0.052-1.280 g) than those found with the core et al., 1977; Troya and Bernués, 1990). If we assume that (n=12, median=0.185, range=0.114-0.343 g, Mann- 10% of them have one shot in the gizzard (Mateo et al., Whitney, Z=2.28, p=0.023). 1998), and considering the persistence of shot in the giz- Granulometric analysis showed that particles larger zard as 21 days (Bellrose, 1959), and the length of the than 1.5 mm were 4.2% of the weight of sand larger than wintering season of 4 months (Sánchez et al., 1977), 1 mm in the dune, while in the gizzards of greylag geese therefore we can estimate the ingestion of 23,000-46,000 they were about 62.0% of the weight of grit larger than 1 shot every year by the wintering greylag geese, which mm (Figure 3). may be a significant percentage of the shot on the sur- face at some parts of the dune. A better knowledge of DISCUSSION the number of geese and their frequency of visits to Cerro Shot distribution in the dune is similar to that re- de los Ánsares would be necessary. ported by Calderón et al. (1996), but densities described The total density in the upper 20 cm of Cerro de los by us were higher (Table 2). These authors also found Ánsares was lower than those found in other Spanish that more than 50% of shot were buried in the sand. wetlands (Guitart et al., 1994; Bonet et al., 1995; Mateo The higher shot density on the surface of the tail of et al., 1997, 1998), but was quite similar to the density the dune may have been caused by the movement of observed in a lagoon of the Guadalquivir Marshes (Mateo this eolic system of dunes. Considering that hunting in et al., 1998). However, in Cerro de los Ánsares shot are Cerro de los Ánsares has been banned since 1983 and easily available to greylag geese because their abundance on the surface. There could be an additional risk if the Table 2. Shot densities in Cerro de los Ánsares on the surface geese visually select grit particles. In fact, large grit was and in the upper 10 cm of sand (including the surfa- more abundant in the gizzard of geese than would be ce) at different heights of the dune. expected from its availability in the sand of the dune.   &DOGHUyQHWDO   3UHVHQWVWXG\D This could be a consequence of a visual selection of these largest particles, similar in size to shot, but it could be 'HSWK +HLJKW 6DPSOH 6KRWP 6DPSOH 6KRWP also a result of a higher retention of these particles in the gizzard. 6XUIDFH      One method recommended to reduce shot ingestion FP +LJK     is supplementation with suitable grit (Mudge, 1992), but experimental studies would be necessary to assess its use-  0HGLXP     fulness. In other cases of lead shot ingestion by upland  /RZ     game birds, a case similar to the geese in this site of grit  2YHUDOO     ingestion, the tillage of soil can reduce the availability of aData for the surface extracted from the sampling with the square and shot for birds (Kendall et al., 1996). However, this meas- data for the upper 10 cm from the core sampling. ure may not be appropriate in a mobile dune. 78 The weight (mean ± s.e.) of shot found was 0.161 ± REFERENCES 0.086 g, thus we can estimate a mean of 26.1 ± 13.9 kg Anonymous. 1991. Cartuchería española a examen. Trofeo of Pb/ha in the upper 20 cm of sand of the study zone. 254:1-25. The length of Cerro de los Ánsares is about 7 km, but Bellrose, F.C. 1959. Lead poisoning as a mortality factor in the length hunted was only 3 km. If we assume a disper- waterfowl populations. Illinois Natural History Survey Bul- sion of shot 250 m around the hunting blinds (Mudge, letin 27:235-288. 1984), the total hunting area of Cerro de los Ánsares Bonet, A., C. Olivares, M.L. Picó and S. Sales. 1995. L’acumula- ció de perdigons de plom al Parc Natural del Fondó d’Elx was about 150 ha. We can estimate therefore an accu- (Alacant): distribució espacial i propostes d’actuació. mulation of 1.8-6 t of Pb in the upper 20 cm of sand. Butlletí de la Institució Catalana d’Història Natural 63:149- Lead on the surface of shot is transformed into a crust of 166. sulfates and carbonates, leading to the total disintegra- Calderón, J., C. Ramo, J.J. Chans and L. García. 1996. Plan de tion of shot in 30-300 years (Jørgensen and Willems, gestión cinegética para el ánsar común en las Marismas del 1987; Lin et al., 1995). The shot collected in Cerro de Guadalquivir. Estación Biológica de Doñana, Sevilla, Spain. los Ánsares showed a certain degree of erosion because DeStefano, S., C.J. Brand, D.H. Rusch, D.L. Finley and M.M. #4 shot employed to hunt greylag geese usually weighs Gillespie. 1991. Lead exposure in Canada geese of the East- 0.21-0.24 g (Anonymous, 1991), and collected shot ern Prairie population. Wildlife Society Bulletin 19:23-32. García-Novo, F., L. Ramírez-Díaz and A. Torres-Martínez. weighed about 0.16 g. Those shot found on the surface 1976. El sistema de dunas de Doñana. Instituto para la showed clear signs of erosion, possibly produced by sand Conservación de la Naturaleza, Naturalia Hispánica 5, moved by wind. Bioavailability of eroded lead depends Madrid, Spain. on acidic pH of soil, which increases the absorption of González, J.L. 1991. Historia natural del águila imperial ibérica lead by plants, and the presence of clay particles or or- (Aquila adalberti Brehm, 1861). Instituto Nacional para la ganic matter, which adsorb lead and reduce its uptake Conservación de la Naturaleza, Madrid, Spain. by plants (W.H.O., 1989; Pain, 1995). Although sandy Guitart, R., J. To-Figueras, R. Mateo, A. Bertolero, S. Cerradelo soil of dunes is not rich in organic matter and clay, the and A. Martínez-Vilalta. 1994. Lead poisoning in water- groundwater in these dunes is rich in calcium bicarbo- fowl from the Ebro delta, Spain: Calculation of lead expo- sure thresholds for mallards. Archives of Environmental Con- nates (Llamas, 1990) and it could limit the lead uptake tamination and Toxicology 27:289-293. by plants. However, greylag geese that have ingested lead Isaak, E.H. and R.M. Srivastava. 1989. An introduction of ap- shot or have been injured by hunters with lead ammuni- plied geostatistics. Oxford University Press, New York, U.S.A. tion are a risk to food chains in Doñana (Mateo et al., Jørgensen, S.S. and M. Willems. 1987. The fate of lead in soils: 2001), because this species is one of the main prey of the The transformation of lead pellets in shooting-range soils. endangered Spanish imperial eagle (Aquila adalberti) in Ambio 16:11-15. winter (González, 1991). Kendall, R.J., T.E. Lacher, C. Bunck, B. Daniel, C. Driver, C.E. Calderón et al. (1996) reported that three people co- Grue, F. Leighton, W. Stansley, P.G. Watanabe and M. llected 800 shot during 30 minutes in the part of the Whitworth. 1996. An ecological risk assessment of lead shot exposure in non-waterfowl avian species: upland game dune with highest density. A person could collect by hand birds and raptors. Environmental Toxicology and Chemistry 533 shot/h, thus 94 h/ha would be necessary to clean 15:4-20. only those shot accumulated on the surface (50,000 shot/ Lin, Z., B. Comet, U. Qvarfort and R. Herbert. 1995. The ha). However, most of the shot are buried, and special- chemical and mineralogical behaviour of Pb in shooting ized machinery designed to collect shot in shooting ranges range soils from central Sweden. Environmental Pollution should be employed to clean Cerro de los Ánsares. This 89:303-309. machinery is expensive, but it should not deter the pub- Llamas, M.R. 1990. Geohydrology of the eolian sands of the lic administrations and hunting or shooting associations Doñana National Park (Spain). Catena Supplement 18:145- to employ this machinery here or at similar sites where 154. Mateo, R., J. Belliure, J.C. Dolz, J.M. Aguilar-Serrano and R. accumulated lead shot represents a threat to wildlife or Guitart. 1998. High prevalences of lead poisoning in win- agriculture in Spain or elsewhere. tering waterfowl in Spain. Archives of Environmental Con- tamination and Toxicology 35:342-347. ACKNOWLEDGEMENTS Mateo, R., R. Cadenas, R. Máñez and R. Guitart. 2001. Lead We wish to thank the personnel of the National Park of shot ingestion in two raptor species from Doñana, Spain. Doñana, in special to Dr. Manuel Máñez and Antonio J. Gonzá- Ecotoxicology and Environmental Safety 48:6-10. lez, their help in the collection of samples, to Cristina Beans Mateo, R., R. Guitart and A. Martínez-Vilalta. 1997. Lead shot the revision of the English language and to Dr. Pedro Puig his pellets in the Ebro delta, Spain: Densities in sediments and help in the statistical analysis. We must also give thanks to prevalence of exposure in waterfowl. Environmental Pollu- Isidro Barroso from the Doñana Natural Park and to the tion 96:335-341. Doñana Biological Station for their logistic support. We thank Mudge, G.P. 1983. The incidence and significance of ingested Dr. Nelson Beyer his comments to the manuscript. lead pellet poisoning in British wildfowl. Biological Conser- vation 27:333-372. Mudge, G.P. 1984. Densities and settlement rates of spent shot- gun pellets in British wetland soils. Environmental Pollution

Ecotoxicology and Environmental Restoration 3(2) 2000 79 8:299-318. lem of lead poisoning in waterfowl. Illinois Natural History Mudge, G.P. 1992. Options for alleviating lead poisoning: a Survey Special Publication 4:1-34. review and assessement of alternatives to the use of non- Sánchez, A., J. Castroviejo and M. Delibes. 1977. On the win- toxic shot. In D.J. Pain, ed., Lead poisoning in waterfowl. tering of greylag geese in the Marismas of the Guadalquivir Proc. IWBR Workshop, Brussels, Belgium, 1991, IWRB Spec. (Southwestern Spain). Proceedings of the XIIIth Congress of Publs, No. 16. pp. 23-25. Game Biologists, pp. 65-76. Pain, D.J. 1991. L’intoxication Saturnine de l’avifaune: Une Szymczak, M.R. and W.J. Adrian. 1978. Lead poisoning in synthèse des travaux français. Gibier Faune Sauvage 8:79- Canada Geese in Southeast Colorado. Journal of Wildlife 92. Management 42:299-306. Pain, D.J. (1995) Lead in the environment. In D.J. Hoffman, Troya, A., and M. Bernués. 1990. Censo de aves acuáticas B.A. Rattner, G.A. Burton and J. Cairns, eds., Handbook of invernantes. Enero 1989. ICONA, Madrid. ecotoxicology, Lewis Publishers, Boca Raton, Florida, U.S.A.. W.H.O. 1989. Environmental Health Criteria 85: Lead - Envi- pp. 356-391. ronmental Aspects. World Health Organization, Geneva, Ramo, C., C. Sánchez and L. Hernández Saint-Aubin. 1992. Switzerland. Lead poisoning of Greater Flamingos Phoenicopterus ruber. Zar, J.H. 1996. Biostatistical analysis. Prentice Hall International, Wildfowl 43:220-222. London, U.K. Sanderson, G.C. and F.C. Bellrose. 1986. A review of the prob-

80