Ardeola 47(2), 2000, 171-181

SUCCESSION OF BREEDING BIRD COMMUNITIES AFTER THE ABANDONMENT OF AGRICULTURAL FIELDS IN SOUTH-EAST PORTUGAL Carlos Pedro SANTOS* SUMMARY.—Succession of breeding bird communities after the abandonment of agricultural fields in south-east Portugal. A study on the bird communities associated with a succession of vegetation in Baixo Alentejo, Portugal, was carried out. Five stages of vegetation were identified: fallow, fallow with shrubs, low shrubs, tall shrubs and mixed shrubs. The abandonment of the agricultural fields started this succession, which were then occupied by two species of the rock-rose family (Cistaceae), Cistus ladanifer and Cistus monspeliensis. The density of passerines increased with succession. In fallows, the most common species belonged to the family Alaudidae, while Sylviidae were the most frequent in the shrub stages. Generally, the numbers of passerine species and the total number of species increased with successional stage while the number of nonpasserine species decreased. Migratory and granivorous/insectivorous passerines were the most abundant in sparse fallow but were replaced by resident and insectivorous species in later stages. The proportion of passerines feeding on the ground or in the air decreased while the proportion of passerines feeding on vegetation increased with successional stage. Passerines nesting on the ground decreased with succession but passerines nesting in shrubs increased. Protecting the fields with the highest densities of non-passerine birds and letting the fields with steep slopes undergo a long-term vegetation succession seem to be the best measures for the management of these areas. Key words: Agricultural abandonment, Alentejo, bird communities, Cistus ladanifer, Cistus monspeliensis, old fields, succession. RESUMEN.—Sucesión de las comunidades de aves nidificantes tras el abandono de cultivos en el sureste de Portugal. Se ha llevado a cabo un estudio de las comunidades de aves asociadas con los diferentes estadíos de la sucesión vegetal tras el abandono de cultivos en el Bajo Alentejo (Portugal). Se identificaron cinco estadíos de vegetación: barbecho, barbecho con arbustos, matorrales bajos, matorrales puros altos y matorrales mixtos. La sucesión comienza con el abandono de los campos cultivados, que son ocupados por dos especies de jaras (Cistaceae), Cistus ladanifer y Cistus monspeliensis. La densidad de paseriformes aumentó con la sucesión. Las especies más comunes en los barbechos pertenecieron a la familia Alaudidae, mientras que la familia Sylviidae fué la más frecuente en las etapas de matorral. Por lo general, el número de especies de paseriformes y el número total de especies aumentaron con la etapa sucesional, mientras que el número de especies de aves no paseriformes disminuyó. Los paseriformes migradores y granívoros/insectívoros fueron más abundantes en barbechos ralos, siendo reemplazados por especies residentes e insectívoras en etapas sucesionales posteriores. La proporción de paseriformes que se alimentaron en el suelo o en el aire se redujo con la etapa sucesional, mientras que la proporción de paseriformes que se alimentan en la vegetación aumentó. Por otra parte, la proporción de paseriformes que anidan en el suelo disminuyó con la sucesión, mientras la de los que que anidan en los arbustos aumentó. La protección de los campos de cultivo que mantengan las mayores densidades de aves no paseriformes, junto con el abandono a largo plazo de los campos con pendientes acusadas, favoreciendo la sucesión de la vegetación, parecen ser las mejores acciones de manejo en estas áreas. Palabras clave: Abandono agrícola, Alentejo, barbechos, Cistus ladanifer, Cistus monspeliensis, comunidades de aves, sucesión.

INTRODUCTION When the agricultural fields in Baixo Alentejo, Portugal, are abandoned, a succession of vegetation takes place, dominated by two species from the rock-rose family (Cistaceae), Cistus ladanifer and Cistus monspeliensis. These shrubs now occupy great tracts of land and in

many cases are the only species present. The causes for this abandonment were related to soil exhaustion due to the wheat campaigns that began in 1929 (Mascarenhas, 1981). The great resistance of the above mentioned plants to drought and the fact that they segregate alelopathic substances, which suppress the growth of other plants, allowed them to occupy and

* Av. Espanha n.o 63-3.o T. P - 5200 Mogadouro, Portugal.

172

SANTOS, C. P.

dominate the fields (Dias et al., 1989). So, the agricultural cycle of ploughed fields → cereal → fallow → ploughed fields, typical of this area, which allowed the restoration of soil fertility, was interrupted (Mascarenhas, 1981). According to Nuñez et al. (1986), the areas of rock-rose will develop into mixed shrubland with species like mastic tree Pistacia lentiscus, Phillyrea angustifolia, myrtle Myrtus communis, holm oak Quercus ilex and wild olive tree Olea europaea. Rivas Goday (1964) argues that succession in rock-rose areas tends towards Oleo-Ceratonion or Quercion rotundifoliae vegetation. The reform of the Common Agricultural Policy (CAP), which may cause further abandonment of agricultural fields or their conversion to other uses (Baldock, 1991), and the fact that a great number of threatened species depend on these types of fields (Santos, 1996), makes it highly desirable to study the effects of field abandonment on birds. The evaluation of the conservation importance of these fields and the consequences of their abandonment have been published by Santos (1996). The present paper gives quantitative data on the successional changes in avian species composition, together with some ecological features of the species concerned. STUDY AREA AND METHODS The study site was located mainly in Mértola district (SE Portugal) and in some fallow fields in Castro Verde district, near the border of Mértola district (7°25´ - 7°56´ W, 37°26´ 37°51´ N). The climate is Mediterranean with continental features, with hot and dry summers and mild winters with little rain. Almost all (99%) of the soils are poor, with no agricultural capacity, and most of them are exhausted and have problems of erosion. The majority of the area is covered with Mediterranean steppe and «subxerophitic» shrublands dominated by rockrose species. There are also areas of holm oak «dehesa» and pine Pinus sp. and eucalyptus Eucalyptus sp. plantations (Pena et al., 1985; Cancela, 1993). Five stages were studied in this succession: (1) fallow, (2) fallow with scattered shrubs, (3) low shrubs, (4) tall shrubs and (5) mixed shrubs. The succession studied is not a com-

plete one because no field with characteristics of a climax stage large enough to be studied was found in the area. In each of these stages, density and species richness of nesting passerines were assessed. The study focussed on passerines because they tend to have small territories that are closely associated with the small patches of shrubland. Density of non-passerines was not estimated because of the relatively large dimensions of their territories. Non-passerines actively using the vegetation patches of each stage, however, were registered (Reynolds et al., 1980). A point count method was used to estimate the density of birds. Following this method, we recorded all the birds detected, by sight as by sound, within a circle with a pre-determined radius (Bibby et al., 1992). Counts were made in the first 3 hours after sunrise on days without rain or strong wind. The time period used was 10 minutes (Fuller & Langslow, 1984), with a delay of 2 minutes after the arrival to the point, to allow the birds to settle. Two series of counts were made at each point. One between 21/4/94 and 10/5/94, for residents and early migrants, and another between 28/5/94 and 14/6/94 for later migrants. Twenty points were established in each successional stage. In stage 1 (fallows), the points had a 100 m radius. Stages 2 and 3 (fallow with shrubs and low shrubs), had points with a 75 m radius. In taller shrubs and mixed shrubs (stages 4 and 5) a radius of 50 m was used. Circles with different radii were used to allow for decreasing detectability of birds with vegetation growth (Reynolds et al., 1980). All points were marked with coloured ribbons, in the centre and in the limits in at least two directions. For the first three stages, the 50 m distance from the centre was also marked. A rangefinder was used to map all objects that could help to estimate the distance to the birds. Birds detected outside the circles but still in the same patch of vegetation were also recorded. The location of the circles was chosen by the following criteria: (1) Vegetation patches were characteristic of their successional stages: fallow had only herbaceous species; fallow with shrubs had patches of herbaceous species with 25% to 75% of the area covered by small patches of scattered rock-roses; low shrubs had more than 90% of the area covered with rockroses up to 1 m; tall shrubs had more than 90% of the area covered with rock-roses taller than

SUCCESSION OF BREEDING BIRD COMMUNITIES AFTER THE ABANDONMENT

1.5 m; mixed shrubs had more than 20% of the area covered with species other than rock-roses taller than 1.5 m and had also a tree layer (> 3 m). (2) Elements not representative of each stage, e.g. trees, patches of shrubs or clearings, should not exceed 10% of the area. (3) The vegetation patches in a nearby area, whether or not of the same type, should allow the establishment of at least 5 points, in order to optimise resources. (4) Points should be fairly accessible. (5) Points should be positioned inside a patch of vegetation 25 m wider than the circle limits (50 m in fallows). At each point the following measurements were made: slope, percentage of bare soil (fallows), percentage cover of the different plant species, percentage cover of trees, height of herbs and shrubs, stonyness and presence of stubble. Height of herbs in the fallows (stage 1) was measured each 10 m along a 50 m line from the center of the point. In fallows with shrubs (stage 2), measures were taken each 5 m for herbs as well as for shrubs, along a 25 m line from the center of the point. The same was done to measure shrub height in the other stages. Percentage of bare soil, vegetation cover and stonyness were estimated visually. To calculate the densities of the passerine species, the same procedure as Ferry & Frochot (1970) was used: 1 male singing, 1 pair, 1 nest or 1 family, means a pair; 1 bird calling or seen means 0.5 pairs. Colonial passerines like the Azure-Winged Magpie Cyanopica cyanus were not considered (Ehrlich et al., 1994), and neither were aerial feeders like swallows or swifts, because they do not have a close association with the vegetation below them when they are feeding (Southwood et al., 1986). Because of the difficulties in identifying Crested Larks Galerida cristata and Thekla Larks Galerida theklae, these species were combined as Crested Larks Galerida spp., although most of the birds seemed to be Thekla Larks. RESULTS AND DISCUSSION During the establishment of the points, differences in the density or even species composition of vegetation within each stage were noticed (Table 1). Although the objectives of this study and the definitions of the stages were not jeopardised by these differences, it was decided

173

to separate the biotopes included in each stage aiming at a more accurate characterisation of the bird communities. Eight biotopes were thus obtained: sparse fallows, dense fallows, fallows with C. monspeliensis shrubs, fallows with C. ladanifer shrubs, low C. monspeliensis shrubland, low C. ladanifer shrubland, tall C. ladanifer shrubland and mixed shrubland. The relationship of the eight biotopes to the five successional stages is shown in Table 1. The average slope increased with successional stage (Table 1) which is due to the fact that fields located on steeper slopes were abandoned first. Many plain fields are still cultivated with wheat today. Substantial changes in bird species composition occurred with vegetation succession (Table 2). Some species were only present in the early stages, others in the late ones. Some species appear in almost all stages and others only in some specific biotope. The Spectacled Warbler is one example of this last case, as this species was only found in shrublands with C. monspeliensis, being relatively abundant there. The presence of tits (Parus spp.) at some points at low densities was due to the presence of trees nearby. For the most abundant species, significant differences between biotopes and between stages were found, while for the less abundant ones this was not the case (Table 3). Woodlark and Stonechat were the exceptions because significant differences between biotopes were found for the first species and among stages for the second one. Woodlarks have a higher abundance in biotope 2 C.m., but it becomes diluted when stages are compared. As for the Stonechat, the combining abundances of biotopes 3 C.m. and 3 C.l. allow for the existence of significant differences among stages. The total density of passerine birds increased with successional stage from fallows to mixed shrubland. The total density in fallows was less than half of that found in any other stage whilst the density in mixed shrubland was twice as great as in any other stage. The total density in fallows was only 6% of the total density found in the mixed shrubland stage. The pattern should be similar when based on densities for all nesting birds, including non-passerines. This agrees with the results obtained by Johnston & Odum (1956), Haapanen (1965),

174

TABLE 1 Characterisation of the vegetation of the different successional stages identified in abandoned fields in southeastern Portugal. Average values are presented (+ means presence; – means absence; 1measured in fallows only; 2measured in shrub patches only). C.m.: Cistus monspeliensis; C.l.: Cistus ladanifer; Q.i.: holm oak Quercus ilex; O.e.: olive tree Olea europea; L.: rosemary Lavandula sp.; P.l.: mastic tree Pistacia lentiscus; M.c.: myrtle Myrtus communis. [Caracterización de la vegetación de las diferentes etapas sucesionales identificadas en los campos abandonados del sureste de Portugal. Sólo se dan valores medios (+: presencia; –: ausencia; 1medido sólo en barbechos, 2medido sólo en manchas de matorral). C.m.: jaguarzo Cistus monspeliensis; C.l.: jara pringosa Cistus ladanifer; Q.i.: encina Quercus ilex; O.e.: olivo Olea europea; L.: lavanda Lavandula sp.; P.l.: lentisco Pistacia lentiscus; M.c.: mirto Myrtus communis.] Stage [Etapa] Biotope [Biotopo]

Vegetation cover [Cobertura vegetal]

Herbaceous level 2 [Estrato herbáceo] Tree level [Estrato arbóreo] Stubble [Rastrojo]

Fallow with shrubs [Barbecho con arbustos]

Low shrubland [Matorral bajo]

Tall shrubland [Matorral alto]

Mixed shrubland [Matorral mixto]

C. m.

C. l.

C. m.

C. l.

C. l.

14

6

20

20

8.5

8.5

16.6

21.5

8

12

11

9

57%

≈0

57%

43%

4

3

6.7

8.5

15.8 ± 7.1

16.2 ± 7.2

12.1 ± 10.2

17.4 ± 15.9

79.6 ± 9.8

141.3 ± 39.2

88.8 ± 6.1

126.5 ± 17.7

168.4 ± 18.6

156.8 ± 21.1

3% Q. i.

3% Q. i.

92% C. m.

0.2% Q. i.

0.5% O. e.

3% Q. i.

62% C. m. 29.5% herbs [herbáceas] 2% thistles [cardos]

61% C. l. 32% herbs [herbáceas] 1% thistles [carlos]

5% C. l. 3% herbs [herbáceas]

94% C. l. 5% C. m.

97% C. l. 2% C. m.

12% O. e. 66.5% C. l. + C. m.

0.8% herbs [herbáceas]

0.2% P. l. 0.3% L.

3.5% bare ground 1 [suelo desnudo]

2% bare ground 1 [suelo desnudo]

13% P. l. 1,5% M. c. 2% herbs [herbáceas]

69% herbs [herbáceas] 16% thistles [cardos] 15% bare ground 1 [suelo desnudo]

100% herbs [herbáceas]

–

–

–

–

–

–

0,2%

0.25%

–

–

1%

–

–

+

–

–

2% stones [piedras] Abundant [Abundante] 15%

SANTOS, C. P.

N.o point counts [N.o de estaciones de escucha] Stones and pebbles 1 [Piedras] Slope (grades) [Pendiente (grados)] Height herbs (cm) [Altura herbáceas] Height Shrubs (cm) [Altura arbustos]

Fallow [Barbecho] Sparse Fallow Dense Fallow [Barbecho [Barbecho poco denso] denso]

TABLE 2 Species density (pairs/ha) in the biotopes (SF: sparse fallows, DF: dense fallows, C.m.: Cistus monspeliensis, C.l.: Cistus ladanifer, MS: mixed shrubs) and the successional stages identified in abandoned fields in southeastern Portugal. Phenology classification of species according to Pena et al. (1984, 1985). Diet, feeding site and nesting site according to Ehrlich et al. (1994), Cramp (1988), Cramp (1992) and Cramp & Perrins (1993). [Densidad de aves (parejas/ha) en los biotopos (SF: barbechos poco densos, DF: barbechos densos, C.m.: Cistus monspeliensis, C.l.: Cistus ladanifer, MS: matorrales mixtos) y las etapas sucesionales identificadas en los campos abandonados del sureste de Portugal. La clasificación fenológica de las especies se realizó según Pena et al. (1984, 1985). La dieta, lugar de alimentación y lugar de nidificación se tomaron de Ehrlich et al. (1994), Cramp (1988), Cramp (1992) y Cramp & Perrins (1993).] Nesting Site [Lugar de nidificación]

Granivorous. [Granívoro]

Ground/air [Suelo/aire]

Ground [Suelo]

Resident [Residente]

Insectivorous [Insectívoro]

Ground/air

Ground

Galerida sp.

Resident

Insectivorous

Ground/air

Ground

0.08 ± 0.2

Miliaria calandra

Resident

Granivorous.

Ground/air

Ground

0.04 ± 0.1

Sylvia undata

Resident

Insectivorous

Vegetation [Vegetación]

Shrubs [Arbustos]

Sylvia melanocephala

Resident

Insectivorous

Vegetation

Shrubs

Lullula arborea

Resident

Insectivorous

Ground/air

Ground

Parus major

Resident

Insectivorous

Vegetation

Trees [Arboles]

0.05 ± 0.2

Parus caeruleus

Resident

Insectivorous

Vegetation

Trees

0.03 ± 0.09

Sylvia conspicillata

Migrant

Insectivorous

Vegetation

Shrubs

0.05 ± 0.2

Lanius senator

Migrant

Insectivorous

Ground/air

Trees

0.03 ± 0.08

Oenanthe hispanica

Migrant

Insectivorous

Ground/air

Ground

Carduelis cannabina

Resident

Insectivorous

Ground/air

Shrubs

0.09 ± 0.2

0.1 ± 0.3

Saxicola torquata

Resident

Insectivorous

Ground/air

Shrubs

0.02 ± 0.08

0.09 ± 0.2

Turdus merula

Resident

Insectivorous

Ground/air

Trees

Luscinia megarhynchos

Migrant

Insectivorous

Ground/air

Ground

Phenology

Diet

[Especies]

[Fenología]

[Dieta]

Calandrella brachydactyla

Migrant [Migrador]

Melanocorypha calandra

1

2

SF

DF

C. m.

0.6 ± 0.3

0.01 ± 0.05

0.1 ± 0.3

3

4

5

C. l.

C. m.

C. l.

C. l.

MS

0.8 ± 0.6

1.0 ± 0.3

0.5 ± 0.4

0.09 ± 0.2

0.03 ± 0.1

0.06 ± 0.3

0.2 ± 0.3

0.1 ± 0.4

0.4 ± 0.5

0.4 ± 0.4

0.4 ± 0.3

0.8 ± 0.5

0.2 ± 0.3

0.01 ± 0.04

0.4 ± 0.0

0.06 ± 0.2

0.8 ± 1.1

0.9 ± 0.7

1.1 ± 0.9

0.99 ± 1.1

0.5 ± 0.3

1.7 ± 1.3

3.04 ± 1.6

0.1 ± 0.2

0.08 ± 0.4 0.2 ± 0.4 0.06 ± 0.2

0.02 ± 0.08 0.3 ± 0.6

0.7 ± 0.9 0.2 ± 0.4

0.05 ± 0.3

0.2 ± 0.4 0.06 ± 0.3

1.02 ± 1.06

Troglodytes troglodytes

Resident

Insectivorous

Vegetation

Shrubs

0.2 ± 0.6

Carduelis chloris

Resident

Insectivorous

Ground/air

Trees

0.06 ± 0.3

Hippolais polyglotta

Migrant

Insectivorous

Vegetation

Shrubs

Total by biotopes [Total por biotopo]

0.1 ± 0.4 0.74

0.23 0.43

1.69

1.95 1.81

2.61

1.79

3.34

7.53

2.36

3.34

7.53

175

Total by stage [Total por etapa sucesional]

SUCCESSION OF BREEDING BIRD COMMUNITIES AFTER THE ABANDONMENT

Biotopes [Biotopos]

Feeding Site [Lugar de alimentación]

Species

176

SANTOS, C. P.

TABLE 3 Kruskal-Wallis tests for the differences in abundance of species by biotopes and by stages. (*: P < 0.05). [Pruebas de Kruskal-Wallis para las diferencias en la abundancia de especies por biotopos y etapas de la sucesión (*: P < 0.05).] By biotopes [Por biotopos]

Species [Especies] Short-toed Lark [Terrera Común] Calandra Lark [Calandria] Crested Larks [Cogujadas] Corn Bunting [Triguero] Dartford Warbler [Curruca Rabilarga] Sardinian Warbler [Curruca Cabecinegra] Woodlark [Totovía] Great Tit [Carbonero Común] Blue Tit [Herrerillo Común] Spectacled Warbler [Curruca Tomillera] Woodchat Shrike [Alcaudón Común] Black-eared Wheatear [Collalba Rubia] Linnet [Pardillo Común] Stonechat [Tarabilla Común] Blackbird [Mirlo Común] Nightingale [Ruiseñor Común] Wren [Chochín] Greenfinch [Verderón Común] Melodious Warbler [Zarcero Común]

By stages [Por etapa]

H

P

H

P

79.07 46.29 63.52 17.99 36.36 79.76 16.35 8.09 8.09 76.41 9.91 7.24 20.46 12.16 11.75 48.79 8.08 4.00 8.08

0.000* 0.000* 0.000* 0.012* 0.000* 0.000* 0.022* 0.325 0.325 0.000* 0.194 0.404 0.005* 0.095 0.109 0.000* 0.326 0.780 0.326

33.38 25.25 55.86 13.81 36.25 72.23 8.08 4.00 4.00 48.97 8.92 3.03 20.43 11.51 8.92 48.78 8.08 4.00 8.08

0.000* 0.000* 0.000* 0.008* 0.000* 0.000* 0.089 0.406 0.406 0.000* 0.063 0.553 0.000* 0.021* 0.063 0.000* 0.089 0.406 0.089

Ferry & Frochot (1970), Jones (1972), Shugart & James (1973), Glowacinski (1975) and Lanyon (1981) for similar stages of succession to those examined here. These authors studied a more complete succession, from the initial stages to the climax forest vegetation, which is not the case here. Larks (Alaudidae) were the dominant family in the early stages, being gradually replaced by warblers (Sylviidae). Larks almost disappeared by stages 4 and 5, which is to be expected because these species are very dependent on bare ground (Cramp, 1988). Martínez & Purroy (1993), reviewing results from studies on birds in Spanish steppes, also found a dominance of larks. Warblers appeared in stage 2, their dominance increasing until stage 4 and then decreasing in stage 5. This family thus showed a great association with the shrub stages. Thrushes (Turdidae) only had some dominance in the last stage. Finches (Fringillidae) only appeared from stage 3 on. Buntings (Emberizidae) showed almost the same dominance

in all biotopes, while other families were relatively sporadic. For the families represented by the most abundant species significant differences were found, while for the families with less abundant species this was not the case (Table 4). Three of the families were only represented by one species (Emberizidae, Laniidae and Troglodytidae) so the test result is the same as for the corresponding species. The proportion of individuals belonging to migratory species showed no obvious trend with successional development (Fig. 1). By far the highest percentage of migrant passerines was registered in biotope 1SF. Apart from biotopes 1SF and 3C.m., the resident birds constitute more than 80% of the total of birds present. Considering successional stages only, fallows have the highest proportion of migrants (36%), followed by stage 3 (low shrubs) (20%) and stage 5 (mixed shrubs) (17%). These values are due to the high densities of only three species: Short-toed Lark, Spectacled Warbler and Nigh-

177

SUCCESSION OF BREEDING BIRD COMMUNITIES AFTER THE ABANDONMENT

TABLE 4 Kruskal-Wallis test for the differences in abundance of families by biotopes and by stages (* P < 0.05). [Pruebas de Kruskal-Wallis para las diferencias en la abundancia de familias por biotopos y etapas de la sucesión (* P < 0.05).] By biotopes [Por biotopos]

Families [Familias] Alaudidae Emberizidae Sylviidae Paridae Laniidae Turdidae Fringillidae Troglodytidae

By stages [Por etapa]

H

P

H

P

66.56 17.99 70.92 8.09 9.91 41.96 20.46 8.08

0.000* 0.012* 0.000* 0.325 0.194 0.000* 0.005* 0.326

56.03 13.81 70.60 4.00 8.92 40.90 20.44 8.08

0.000* 0.008* 0.000* 0.406 0.063 0.000* 0.000* 0.089

FIG. 1.—Proportion of residents and migrants in the different biotopes (1SF: sparse fallow; 1DF: dense fallow; 2C.m.: fallow with shrubs of Cistus monspeliensis; 2C.l.: fallow with shrubs of Cistus ladanifer; 3C.m.: low shrubland of Cistus monspeliensis; 3C.l.: low shrubland of Cistus ladanifer; 4C.l.: tall shrubland of Cistus ladanifer; 5MS: mixed shrubland) Open bars: migrants, closed bars: residents. [Proporción de especies residentes y migrantes en los diferentes biotopos (1SF: barbecho poco denso; 1DF: barbecho denso; 2C.m.: barbecho con arbustos de Cistus monspeliensis; 2C.l.: barbecho con arbustos de Cistus ladanifer; 3C.m.: matorral bajo de Cistus monspeliensis; 3C.l.: matorral bajo de Cistus ladanifer; 4C.l.: matorral alto de Cistus ladanifer; 5MS: matorral mixto). Barras vacías: migrantes, barras rellenas: residentes.]

tingale, respectively. These results are not coincident with the ones reported by Ferry & Frochot (1970), who detected an increase in the percentage of migrants for similar stages of succession. One explanation may be the latitude difference between France and Portugal. Herrera (1978) found a significant, positive correla-

tion between percentage of migrants and latitude. In the Ferry & Frochot (1970) study, the number of migrant species found is more than twice the number of migrant species found in this study —only six species— and not the most abundant ones. This may explain the irregular pattern of migrants found in this succession.

178

SANTOS, C. P.

If we consider only long distance migrants (birds wintering south of the Sahara desert) the results are clearly different. In the present study only two species, the Woodchat Shrike and the Nightingale, fulfil the above definition. Thus, the only significant percentage of long distance migrants is 16% in stage 5. Helle & Fuller (1988), collecting data from successions in Europe that include the results of Ferry & Frochot (1970), verified that the proportion of long distance migrants is higher in stages with vegetation 1-2 m tall and then decreases in the following stages. Based on data from the northern hemisphere, Mönkkönen & Helle (1989) only found this pattern for successions in Europe, while in North America and Asia long distance migrants increased or did not decrease (Western North America) along the succession. Data obtained in this study seem to agree with the data from these authors, considering that the present succession is not a complete one. No obvious pattern was evident in the relative proportions of insectivorous and granivorous passerines as related to succession (Figure 2A). Granivorous birds dominated biotope 1 SF, while insectivorous birds made up more than 80% of the birds present in the other biotopes. Figure 2-B suggests a decrease in the proportion of birds that feed on the ground and/or in the air, which is 100% in the fallows, and an increase in the proportion of birds feeding on the vegetation. This is to be expected because the availability of bare ground decreases with successional stages. Nevertheless, in biotope 5 MS there is a slight increase of birds feeding on the ground, probably due to the height of the trees and death of shrubs which allows them access to the ground. Ferry & Frochot (1970) reported similar results. Figure 2-C suggests a similar pattern. There is a decrease in the proportion of birds nesting on the ground, and an increase of birds that nest on the vegetation. In biotope 5 MS there is a slight increase of birds nesting on the ground which may be explained by the reasons referred to above. In biotopes 2 C.m. and 3 C.l. the presence of tree-nesting birds is related to the presence of trees in the vicinity of point counts. Ferry & Frochot (1970) and Glowacinski (1975) reported similar results with respect to nest site selection in the same stages. Figure 3 shows the number of passerine species, non-passerine species and the total number of species, in the different stages of the

succession, inside and outside the point counts, in the two counts. The highest number of passerine species is found in the most heterogeneous biotopes, like fallow with shrubs and mixed shrubs. This last biotope has the highest number of passerine species. The number of nonpasserine species decreases from the fallows to the shrubland areas. The total number of species shows no clear trend with successional stage and is similar to that for passerines. A negative significant correlation was found between number of species of non-passerine birds and the successional stage (r = –0.9; P = 0.037), while no significant correlation was found for the passerine birds (r = 0.6; P = 0.285) or for all bird species (r = –0.103; P = 0.870). These results are not totally coincident with the ones obtained by Ferry & Frochot (1970), Jones (1972), Glowacinski (1975), Lanyon (1981), Southwood et al. (1986) and Valvo & Massa (1990). Their results showed an increase of the total number of species with successional development. Blondel (1981) found a positive relationship between the number of species and the structural complexity of the vegetation. The reason for this difference in the results is probably due to the fact that this succession corresponds only to two or three stages of the succession studied by the above authors. Probably, the differences found here would be smoothed if we could have included the next stages of the succession. Besides, if we consider stages 2 and 5 as edges, we may understand the higher number of species found there. Edges attract species from the two biotopes nearby which add to the species from the edge itself. Martin (1960) also found the highest values in forest edges. Southwood et a.l (1986) also found that larger species, usually non-passerine birds, were mainly associated with the open stages. Two hypotheses may explain this situation. Verner & Wilson (1966) argue that, in early stages, all primary productivity is concentrated in a narrow band just above the ground, while in the woodland areas it is spread throughout a larger volume of vegetation. Density of food is thus higher in the open areas and these species have a higher feeding efficiency. Lawton (1978, in Southwood et al., 1986) and Southwood et al. (1978, in Southwood et al., 1986) suggest that only small animals can explore the great variety of niches in woodland areas.

SUCCESSION OF BREEDING BIRD COMMUNITIES AFTER THE ABANDONMENT

179

FIG. 2.—Proportion of passerines with A) different diets (open bars: granivorous, closed bars: insectivorous); B) different feeding sites (open bars: ground/air, closed bars: vegetation); C) with different nesting sites (open bars: ground; striped bars: shrubs; closed bars: trees) according to biotopes. See table 2 for the classifications of bird species. [Proporción de aves paseriformes con A) diferentes dietas (barras vacías: granívoros, barras rellenas: insectívoros); B) diferentes lugares de alimentación (barras vacías: tierra/aire; barras rellenas: vegetación); C) diferentes lugares de nidificación (barras vacías: tierra; barras rayadas: arbustos; barras rellenas: árboles), según los biotopos. Véase la tabla 2 para la clasificación de las especies de aves.]

180

SANTOS, C. P.

FIG. 3.—Number of species of passerines, non-passerines and total species found in the successional stages of the abandoned fields of southeastern Portugal (open bars: passerines; closed bars: non-passerines; striped bars: all birds). [Número de especies de paseriformes, de no paseriformes y especies totales encontrados en las diversas etapas sucesionales de los campos abandonados del sureste de Portugal (barras vacías: paseriformes; barras rellenas: no paseriformes; barras rayadas: todas las especies).]

FINAL REMARKS The wheat campaigns that started in 1929 in Alentejo resulted in the cultivation of large areas of land, mostly with no agricultural capacity. This fact led to an increase in soil erosion and to the exhaustion of soil fertility, making it almost sterile. The progressive abandonment by farmers and the invasion of rock-roses started a process of succession that will eventually transform the rock-rose shrubwoods into a Mediterranean forest. No one knows how long it will take, but many decades will certainly pass before we can see changes in some of these areas. The destruction of these shrublands should not occur, because they are the only protection of these soils, especially those on steep slopes. Birds are one of the vertebrate groups most affected by these changes of vegetation. This paper shows that a succession of bird communities follows the vegetation succession. An increase of passerine density and richness was found as the succession developed. In contrast, the richness and abundance of non-passerine birds decreased. This suggests that future management of this area should be based on two

aspects: a) a protection of the fields that support the highest densities of non-passerine birds, especially the most threatened ones (Santos, 1996), which are usually lowland fields; and b) to leave the fields with steep slopes to undergo long-term vegetation succession. Of course, this management should be done considering the management of other conservation interests, related to other species of fauna and flora, as well as the local population interests. ACKNOWLEDGEMENTS.— I would like to thank Prof. Jorge Palmeirim for all the support that he gave to this work. To the Associação de Defesa do Património de Mértola and Campo Arqueológico de Mértola for the facilities. To Dr. Susana Marques and Dr. Robert Fuller and Dr. Mario Díaz for their most valuable comments and corrections of this study.

BIBLIOGRAPHY BALDOCK, D. 1991. Implications of EC farming and countryside policies for conservation of lowland dry grasslands. In, P. D. Goriup, L. A. Batten & J. A. Norton (Eds.): The conservation of lowland dry grassland birds in Europe, pp 111-117. ICBP. Cambridge.

SUCCESSION OF BREEDING BIRD COMMUNITIES AFTER THE ABANDONMENT

BIBBY, C. J., BURGESS, N. D. & HILL, D. A. 1992. Bird census techniques. Academic Press. London. BLONDEL, J. 1981. Structure and dynamics of bird communities in Mediterranean habitats. In, F. Di Castri, D. W. Goodall & R. L. Specht (Eds.): Maquis and Chaparrals, pp. 361-385. Elsevier. Amsterdam. CANCELA, J. F. 1993. Factores abióticos e culturais Caracterização paisagistica. In, L. Matos & J. Cancela (Eds.): Caracterização biofísica do troço médio do vale do Guadiana (região de Mértola), pp. 7-38. Associação Defesa Património. Mértola. CRAMP, S. (Ed.) 1988. The Birds of the Western Palearctic, vol V. Oxford University Press. Oxford. CRAMP, S. (Ed.) 1992. The Birds of the Western Palearctic, vol. VI. Oxford University Press. Oxford. CRAMP, S. & PERRINS, C. M. (Eds.) 1993. The Birds of the Western Palearctic, vol. VII. Oxford University Press. Oxford. DIAS, A. S., DIAS, L. S. & PEREIRA, I. P. 1989. Cistus ladanifer (esteva). Caracterização e perspectivas de utilização. Actas do II Congresso áreas protegidas, pp. 275-283. Lisboa, Dezembro 1989. EHRLICH, P. R., DOBKIN, D. S., WHEYE, D. & PIMM, S. L. 1994. The Birdwatcher’s Handbook. Oxford University Press. Oxford. FERRY, C. & FROCHOT, B. 1970. L’avifaune nidificatrice d’une foret de chenes pedonculés en Bourgogne: Etude de deux successions ecologiques. La Terre et la Vie, 2: 153-250. FULLER, R. J. & LANGSLOW, D. R. 1984. Estimating numbers of birds by point counts: how long should counts last?. Bird Study, 31: 195-202. GLOWACINSKI, Z. 1975. Succession of bird communities in the Niepolomice Forest (Southern Poland). Ekologia Polska, 23: 231-263. HAAPANEN, A. 1965. Bird fauna of the Finnish forests in relation to forest succession, I. Annales Zoologici Fennici, 2: 153-196. HELLE, P. & FULLER, R. 1988. Migrant Passerine birds in European forest successions in relation to vegetation height and geographical position. Journal of Animal Ecology, 57: 565-579. HERRERA, C. M. 1978. On the breeding distribution pattern of European migrant birds: MacArthur’s theme reexamined. Auk, 95: 496-509. INE 1993. Censos 91. Instituto Nacional de Estatística. Lisboa. JOHNSTON, D. & ODUM, E. 1956. Breeding bird populations in relation to plant succession on the Piedmont of Georgia. Ecology, 37: 50-62. JONES, P. 1972. Succession in breeding bird population of sample Welsh oakwoods. British Birds, 65: 291-299. LANYON, W. 1981. Breeding birds and old field suc-

181

cession on Fallow Long Island farmland. Bulletin of the American Museum of Natural History, 168: 1-60. MARTIN, N. D. 1960. An analysis of bird populations in relation to forest succession in Algoquin Provincial Park, Ontario. Ecology, 41: 126-140. MARTÍNEZ, F. J. & PURROY, F. J. 1993. Avifauna reproductora en los sistemas esteparizados ibéricos. Ecología, 7: 391-401. MASCARENHAS, J. M. P. B. 1981. L’evolution du paysage en Alentejo. Thése de Docteur Ingenieur. Université Paul Sabatier. Toulouse. MÖNKKÖNEN, M. & HELLE, P. 1989. Migratory habits of birds breeding in different stages of forest succession: A comparison between the Palearctic and the Nearctic. Annales Zoologici Fennici, 26: 323330. NÚÑEZ, E., CABEZAS, J. & ESCUDERO, J. C. 1986. Relación de matorrales en la formación de dehesas. Función de la jara, valorización de su biomasa y su utilización industrial. I Jornadas Técnicas sobre la conservación y desarrollo de las dehesas portuguesa y española. Badajoz, enero 1986. PENA, A., GOMES, L. & CABRAL, J. 1984. Estudo fitogeográfico e faunístico do concelho de Mértola. Câmara Municipal de Mértola. Mértola. PENA, A., GOMES, L. & CABRAL, J. 1985. Fauna e flora de Mértola. Câmara Municipal de Mértola. Mértola. REYNOLDS, R. T., SCOTT, J. M. & NUSSBAUM, R. A. 1980. A variable circular-plot method for estimating bird numbers. Condor, 82: 309-313. RIVAS GODAY, S. 1964. Vegetación y flórula de la cuenca Extremeña del Guadiana. Publicaciones Diputación Provincial de Badajoz. Madrid. SANTOS, C. P. 1996. O abandono dos campos agrícolas e suas implicações nas comunidades de aves nidificantes. Ciência e Natureza, 2: 95-102. SHUGART, H. & JAMES, D. 1973. Ecological succession of breeding bird populations in Northwestern Arkansas. Auk, 90: 62-77. SOUTHWOOD, T., BROWN, V., READER, P. & GREEN, E. 1986. The use of different stages of a secondary succession by birds. Bird Study, 33: 159-163. VALVO, M. & MASSA, B. 1990. Le comunitá di ucelli in gradienti vegetazionali delle isole Mediterranee e delle Canarie. Naturalista Siciliano, S. IV, XIV (Suppl): 95-105. VERNER, J. & WILSON, M. F. 1966. The influence of habitats on mating systems of North American passerine birds. Ecology, 47: 143-147. [Recibido: 17-6-98] [Aceptado: 23-12-99]