ORNIS HUNGARICA 12-13: 1-2 (2003) 297

Abundance of Four Lark Species in Relation to Portuguese Farming Systems

C. Stoate, The Game Conservancy Trust, Loddington House, Loddington, Leics. LE7 9XE, U.K. E-mail: [email protected]. R. Borralho, ERENA, Av. Visconde Valmor 11-3°, 1000 Lisbon, Portugal. M. Araújo, Centro de Ecologia Aplicada, Universidade de Évora, Portugal.

1. Introduction systems survive in some parts of southern Europe, including Portugal, where the main arable region is Alentejo in the south Throughout much of Europe, farmland of the country. However, such systems are birds have declined more than those of not currently economically viable. other habitats, because of the abandon- Larks (Alaudidae) represent a passerine ment of traditional farming systems and family that is strongly associated with the simplification of remaining agricultur- farmland landscapes. In Portugal, the al systems, including increased use of species present include Calandra Lark external inputs (Tucker & Heath 1994, Melanocorypha calandra, Woodlark Baldock 1991, Bignal & McCracken Lullula arborea, Short-toed Lark 1996). Traditional low-input agricultural Calandrella brachydactyla and Skylark Alauda arvensis. Of these, Woodlark and Tab. 1. Agricultural statistics for the three Short-toed Lark are widely distributed as land-use categories considered in Alentejo, Portugal (source: Cordovil 1993). breeding species in Portugal, while Skylark occurs mainly in the north of the country Intensi Extens Monta and Calandra Lark mainly in the southeast ve ive do Mean farm size (ha) 48 161 66 (Rufino 1989). Iberian populations of all (all farms) four species declined in numbers during the 1970s and 1980s and are currently the sub- Crop area (%) Total annual crops 81 42 28 ject of European conservation concern Winter cereals 45 40 21 (Tucker & Heath 1994). This study assess- Sunflower 20 0.3 0 Forage crops 6 2 7 es the abundance of these species in December and April in relation to three Fallow 15 52 66 arable systems in an agricultural landscape Perennial crops 11 0.8 2 of Baixo Alentejo, southern Portugal. Olive 10 0.8 2 Vines 1 0 0 2. Study area Land area per tractor 58 125 194 (ha)

Livestock (%) The study area included parts or all of five Sheep 68 78 83 administrative regions in Baixo Alentejo Cattle 8 11 7 Pigs 22 8 5 (Ferreira do Alentejo, Aljustrel, Castro Goats 2 3 5 Verde, Ourique and Almodôvar), an area 298 ORNIS HUNGARICA 12-13: 1-2 (2003)

70 250 60 200 50 Calandra Calandra 150 40 Wood Wood 100 30 Skylark Short-toed 20 50

10 density (kmBird -2) 0 Bird density (kmBird -2) 0 1995 1996 1995 1996 1995 1996 1994 1995 1996 1994 1995 1996 1994 1995 1996

Montado Extensive Intensive Montado Extensive Intensive

Fig 1. Densities (number km-2 with 95% con- Fig 2. Densities (number km-2 within 95% fidence limits) of Calandra Lark, Woodlark confidence limits) of Calandra Lark, Woodlark and Short-toed Lark in April in relation to and Skylark in December in relation to three three Alentejo farming systems. Alentejo farming systems. totalling 155 000 ha. Within this region, periods often last five years or more (Rio three land-use systems were recognised: Carvalho et al. 1995). Wheat yields are intensive agriculture, extensive agricul- 1.5-2.5 tonnes ha-1, yields at the lower end ture and Montado (Tab. 1). of this range being more common (P. Eden The intensive agriculture category is pers comm 1998). Triticale Triticum aes- characterised by a greater frequency tivum x Secale cereale and oats Avena sati- (>55%) of heavy soils, much of the area va are frequently grown in the extensive being irrigated. Wheat Triticum aestivum category and grazed or cut for silage. The and barley Hordeum distichum are the incorporation of a fallow period into the main cereal crops and silage grass Lolium rotation, and the relatively low potential sp, sunflower Helianthus annuus, sugar yields are associated with considerably beet Beta vulgaris and oilseed rape lower annual inputs than in the intensive Brassica napus are also grown. Wheat category. yields are 2.5-3.5 tonnes ha-1 without irri- Montado (equivalent to the Spanish gation but can be almost doubled with full Dehesa) is characterised by thin soils and irrigation (P. Eden pers comm 1998). tree cover, dominated by holm oak There are short rotations with little or no Quercus rotundifolia and cork oak Q. fallow (e.g. sunflower / 1st cereal / 2nd cere- suber at up to 20 trees ha-1 (mean=10.5 al). This system requires frequent use of trees ha-1, se=0.7). Like the extensive cate- -1 fertiliser (130 units N2 ha (P. Eden pers gory, there is no irrigation and the fallow comm 1998)) and herbicides, relative to area is high. A typical rotation is similar to the other land-use categories. With the that of the extensive category, although exception of some olive Olea europaea the fallow stage is often longer and forage groves, there is little tree cover. lupins Lupinus luteus may be included. The extensive agriculture category is Sheep Ovis aries, cattle Bos taurus and characterised by thin soils and by the pigs Sus scrofa are kept in all three land- largest average farm size of the three cate- use categories. Zero grazing is adopted on gories (Tab. 1). There is no irrigation and some farms in the intensive category but the fallow area is relatively high. A typical livestock normally graze fallows. Tab. 1 rotation takes the form: plough fallow / 1st lists the proportion of crops and livestock cereal / 2nd cereal / fallow / fallow. Fallow in each category. ORNIS HUNGARICA 12-13: 1-2 (2003) 299

3. Methods Woodlark occurred only in Montado and Calandra Lark only in extensive farmland, in both April and December. Differences A total of 115 250 m transect counts, start- between land-use categories were signifi- ing at 1 km grid intersections and strati- cant for Woodlark in December fied by land-use categories, were conduct- (F2,5=210.4, P<0.001) but not in April, and ed along a random bearing (Montado for Calandra Lark in both December n=42, Extensive n=42, Intensive n=31). (F2,5=15.0, P<0.01) and April (F2,2=25.2, Transect counts were conducted by a sin- P<0.04). Short-toed Lark occurred in both gle observer in the first three hours after extensive and intensive farming systems, dawn, or the two hours before dusk in with a higher density in the former, December (1994, 1995 and 1996) and although this difference was not signifi- April (1995 and 1996). Perpendicular dis- cant. Skylark was absent in April and tances from the transect line to each occurred at similar densities in both exten- detected bird were estimated visually to sive and intensive systems in December. the nearest 10 m. Detectability was assumed to be the same in each of the 5. Discussion three land-use categories, even in Montado, because habitat structure is rela- tively open. The number of birds seen The international distribution of Woodlark together at an observation was recorded. is confined almost exclusively to Europe, Density estimates were calculated for between 12% and 34% of this population each land-use category using line transect occurring in Portugal (Tucker & Heath sampling and the computer program DIS- 1994). This species occurs in several TANCE (Buckland et al. 1993, Laake et al. Portuguese forest systems, but within 1993). Analyses used clusters (<1 individ- farming systems, the species is restricted ual) as analytical units, and untruncated per- to Montado. Large areas of Montado have pendicular distance data. A variety of rec- been cleared since the 1930s, and contin- ommended robust estimators implemented ued local clearance of trees, such as that by DISTANCE was used, the final model associated with the development of the selected in each case being the one with the Alqueva reservoir, is likely to have severe lowest Akaike's Information Criterion value consequences for Woodlark and other (Buckland et al. 1993). Differences in bird species associated with Montado in abundance between land-use categories Alentejo (Vieira & Eden 1995). Changes were tested using two-way ANOVA in scrub clearance frequency and prac- (year×land-use) and LSD post-hoc tests (at tices, grazing densities and livestock types P<0.05) log-transformed data. also influence tree health and habitat structure (Pinto-Correia & Mascarenhas 4. Results 1999). Financial support is currently avail- able for the re-establishment of Montado within Alentejo under EU Reg. 2080/92 so Densities of the four species are presented that the total area of Montado is currently at Figs 1 (April) & 2 (December). not experiencing substantial change. 300 ORNIS HUNGARICA 12-13: 1-2 (2003)

However, this is a long-term project, and same area, found that Calandra Lark was the habitat is also threatened by the spread strongly associated with low vegetation on of a fungal disease that is causing wide- fallows within extensive farming systems. spread destruction of oak trees across the However, the ecological requirements of Iberian Peninsula (Brasier 1992). Calandra Lark in Portugal throughout the Large- and medium-sized invertebrates year are poorly understood and require such as grasshoppers (Orthoptera), spiders further investigation if appropriate man- (Araneae) and caterpillars (Lepidoptera agement is to be adopted for this species. and Symphyta larvae) form an important Short-toed Lark occurred in intensive nestling food source for each of the agricultural systems, but was present at species considered here (Cramp 1988). low densities and generally was associated Low pesticide inputs and continued incor- with non-crop habitats such as tracks and poration of fallow periods into arable rota- other uncultivated areas. Moreira (1999) tions are associated with higher densities reported an association between Short- of these invertebrates than in intensively toed Lark abundance and bare ground managed agricultural systems in Alentejo within fallows. Like Calandra Lark, this (Stoate et al. 2000). This might explain species appears to benefit from the main- the low breeding abundance of larks in tenance of extensive farming systems. intensively managed systems, for as This habitat supports more bird species of Ewald and Aebischer (1999) have shown, current conservation concern than does there is a relationship between pesticide any other (Tucker & Heath 1994, Santos use, invertebrate abundance and breeding 1996, Araújo et al. 1996). Some of these abundance of Grey Partridge Perdix perdix species (e.g. Lesser Kestrel Falco nau- and Corn Bunting Miliaria calandra in the manni, Great Bustard Otis tarda, Little U.K. Like larks, Corn Bunting is also pre- Bustard Tetrax tetrax and Roller Coracias sent at relatively low densities in inten- garrulus) are highly valued by birdwatch- sively managed agricultural systems in ers, and a small-scale tourist industry is Alentejo (Stoate et al. 2000). currently developing around the presence Skylark is absent from all farming sys- of these species on extensively managed tems during April, confirming the breed- farmland in Portugal. ing distribution of Rufino (1989). Farm tourism is also developing in However, in December this species occurs Spain (Garcia-Ramon et al. 1995) and pro- in both extensive and intensive landscape vides an opportunity for alternative categories. This may be explained by the income for management of extensive species using cereal crop leaves as a win- farming systems, either directly through ter food source, although Calandra Lark, accommodation on farms, or indirectly, as which also feeds on cereal leaves (Cramp in Alentejo, through EU Rural 1988), is absent from intensively managed Development support for maintaining such systems. About one third of the European systems. However, many steppe species Calandra Lark population occurs in the are susceptible to disturbance and access Iberian Peninsula where the species is requires careful management. This does declining in numbers (Tucker & Heath, not preclude the potential for combining 1994). Moreira (1999), working in the public and private support for extensive ORNIS HUNGARICA 12-13: 1-2 (2003) 301 farming systems that support larks and Ewald, J. & N. J. Aebischer. 1999. Pesticide use, avian food resources and bird densities in other species associated with extensively Sussex. – JNCC, Peterborough, U.K. (JNCC managed arable landscapes. Report No 296). Garcia-Ramon, D., Canoves, G. & N. Valdovinos. Acknowledgements. This study was part of a 1995. Farm tourism, gender and the environment wider project on biodiversity of European in Spain. – Annals of Tourism Research 22: 267- farming systems and was funded by ERENA 282. and the European Commission Environment Laake, J. L., Buckland, S. T., Anderson, D. R. & K. Programme (PL93-2239). We are also grateful P. Burnham. 1993. DISTANCE User's Guide. Version 2.0. Colorado Co-operative Fish and to Peter Eden for agricultural information and Wildlife Research Unit. Colorado State to Nicholas Aebischer comments on the text. University, Fort Collins, USA. Moreira, F. 1999. Relationships between vegetation structure and breeding bird densities in fallow References cereal steppes in Castro Verde, Portugal. – Bird Study 46: 309-318. Pinto-Correia, T. & J. Mascarenhas. 1999. Araújo, M., Borralho, R. & C. Stoate. 1996. Can bio- Contribution to the extensification/intensifica- diversity be measured using composite indices? tion debate: new trends in the Portuguese monta- pp. 124-125. In: Farinha, J. C., Almeida, J. & H. do. – Landscape and Urban Planning 46: 125- Costa. (Eds). I Congresso de Ornitologia. – 131. Sociedade Portuguesa para o Estudo das Aves, Rio Carvalho, C., Borralho, R., Bugalho, J. & A. Lisbon, Portugal. Barreto. 1995. A exploraçăo dos recursos bravios Baldock, D. 1991. Implications of EC farming and e a sua relaçăo com a economia agricola: per- countryside policies for conservation of lowland spectivas actuais. – Revista de Cięncias Agrarias dry grasslands. pp. 111-117. In: Goriup, P. D., 18: 11-22. Batten, L. A. & J. A. Norton. The Conservation Rufino, R. 1989. Atlas das Aves que Nidificam em of Lowland Dry Grassland Birds in Europe. – Portugal Continental. – CEMPA,. Lisboa, JNCC, Peterborough, U.K. Portugal. Bignal, E. M. & D. I. McCracken. 1996. Low-inten- Santos, C. P. D. 1996. O abandono dos campos agrí- sity farming systems in the conservation of the colas e suas implicaçőes nas communidades de countryside. – J. Appl. Ecol. 33: 413-424. aves nidificantes. – Cięncia e Natureza 2: 95- Brasier, C. M. 1992. Oak tree mortality in Iberia. – 102. Nature 360: 539. Stoate, C., Borralho, R. & M. Araújo. 2000. Factors Buckland, S. T., Anderson, D. R., Burnham, K. P. & affecting corn bunting Miliaria calandra abun- J. L. Laake. 1993. Distance Sampling: dance in a Portuguese agricultural landscape. – Estimating Abundance of Biological Agriculture, Ecosystems and Environment 77: Populations. – Chapman & Hall, London, U.K. 219-226. Cordovil, F. C. 1993. Tipos de Agricultura na Zona Tucker, G. M. & M. F. Heath. 1994. Birds in Europe; Agraria de Aljustrel por Subzonas em 1989. Their Conservation Status. – BirdLife Dinamia, Lisboa, Portugal. International, Cambridge, U.K. (BirdLife Cramp, S. 1988. Handbook of the Birds of Europe, Conservation Series No 3.). the Middle East and North Africa: The Birds of Vieira, L. M. & P. Eden. 1995. The Portuguese the Western Palearctic. Volume V. – Oxford Montados. pp. 99-102. In: McCracken, D. I., University Press, Oxford, U.K. Bignal, E. M. & S. E. Wenlock. (Eds). Farming on the Edge: the Nature of Traditional Farmland in Europe. – JNCC, Peterborough, U.K. 302 ORNIS HUNGARICA 12-13: 1-2 (2003)