Preserving frugivorous birds in agro-ecosystems: lessons from Spanish olive orchards

Authors


Correspondence author. E-mail: prey@ujaen.es

Summary

1. Frugivorous birds are a priority for conservation. They are experiencing the transformation of natural habitats to agro-ecosystems worldwide and some are taking advantage of agricultural production of fleshy-fruited plants. However, the mechanisms through which some birds are able to thrive in agricultural landscapes while others become extinct are poorly known.

2. This manuscript provides an overview of avian frugivory in olive orchards, one of the principal agro-ecosystems in the Mediterranean region and an important winter refuge for birds. The hypothesis that Mediterranean avian frugivores are pre-adapted to olive orchards is used to consider potential constraints to bird occupation of wider agro-ecosystems.

3. Agricultural practices and artificial selection of fruit cause habitat and landscape simplification and reduction of bird food resources in orchards, with resulting negative effects on bird diet and body condition, as well as on bird abundance and diversity. Some of these constraints can be partially overcome by the presence of small hedgerows and copses in the agricultural landscape.

4. Five pre-adaptive features determine the successful occurrence of a bird species in olive orchards: (1) frugivorism intensity; (2) ability to track variation in fruit availability; (3) diet plasticity to cope with low fruit diversity and unbalanced food; (4) fruit handling plasticity to cope with oversized fruits; and (5) ecomorphology and foraging niche conditions which increase the ability to respond to habitat simplification.

5.Synthesis and applications. Management practices have the potential to alleviate the constraints that habitat simplification puts on bird movement and diet. Two inter-related approaches to management are proposed: increasing landscape and habitat diversity by the occurrence of hedgerows and forest remnants; and increasing food availability through reducing the use of pesticides and promoting fruit diversity in hedges.

Introduction

The way in which animals and plants cope with new environments in a human-impacted world is of great interest for biodiversity management under global change scenarios (Pimm & Gittleman 1992; Benton, Vickery & Wilson 2003). The successful settlement of wildlife in croplands is of particular interest because major changes in these landscapes have led to the disappearance of many animal and plant species (Mas & Dietsch 2004). However, the mechanisms that permit some species to thrive in croplands remain largely unexplored.

Many plants depend on a mutualistic association with frugivorous animals, hence frugivores are a conservation priority (Cordeiro & Howe 2001; Tellería, Ramírez & Pérez-Tris 2005). The use of fleshy-fruited plants (Vitis, Rubus, Olea, Vaccinium, Prunus, Citrus, Coffea, etc.) by agriculture occurs worldwide. Vertebrates consume the fruits of the wild relatives of these crops and many of the agricultural croplands are used by frugivores. The successful establishment of frugivores in croplands can be interpreted in terms of pre-adaptation to these cropping conditions. Understanding these pre-adaptive processes is fundamental to understanding the requirements of these animals in new environments.

The use of olive orchards as wintering quarters by frugivorous birds in the Mediterranean region (Rey 1993) provides an example of such pre-adaptive processes and their implications for frugivore conservation. The cultivated olive tree Olea europaea var. europaea was developed over millennia by human selection of the wild olive Olea europaea var. sylvestris (Breton et al. 2006). Olive orchards occupy large areas of the Mediterranean Basin, with approximately 6 million hectares in Spain, Italy, Greece, Tunisia, France, Turkey, Israel, Morocco and Cyprus (Loussert & Brousse 1980; Junta de Andalucía 2003). There are also 0·5 millions hectares of olive orchards in Portugal in areas under Mediterranean climate. The cultivated olive tree exists in the same ecological environment as the wild plant. The occurrence of Mediterranean avian frugivores in olive orchards is thus expected to mirror the relationships between frugivores and native fleshy-fruited plants in natural Mediterranean habitats.

Population regulation during winter may be especially important for many temperate migrant birds that, although widespread during the breeding season, assemble in restricted geographic areas during the winter. The loss of natural habitats and food supply that affects local populations in wintering grounds can have disproportionate, large-scale effects on breeding population dynamics (Webster et al. 2002; Bibby 2003). The Mediterranean region has a long history of natural habitat loss, transformation to agricultural landscapes and reduction of fruit supplies. Olive orchards therefore have the potential to act as alternative habitats for the conservation of frugivores in the region.

This study reviews frugivory in olive orchards in southern Spain and makes comparison with frugivory in natural habitats in the same region, where numerous studies have been conducted in recent years (reviewed in Herrera 1995). Southern Spain forms the focus because (i) it contains the largest amount of olive cultivation in the world; and (ii) it is one of the major wintering grounds and migratory pathways for frugivorous–insectivorous birds of the Palearctic (Tellería, Asensio & Díaz 1999). In addition, major insights into frugivory in the Mediterranean Basin have come from studies in southern Spain.

First, habitat structure and landscape heterogeneity is considered together with comparative information on abundance and diversity of frugivorous birds and fruits to formulate the hypothesis of Mediterranean avian frugivore pre-adaptation to olive orchards (Herrera 1983; Muñoz-Cobo 1987). Subsequently, the major pre-adaptive features which constrain the use of orchards by birds are identified and used to highlight mechanisms that limit the occupation of this cropland by different frugivores. Finally, the findings are used to suggest management practices to enhance frugivore diversity in this and other agro-ecosystems.

Physical characteristics and agricultural landscape structure of southern Spain

Although distributed throughout the Mediterranean, olives groves occupy more than 2 400 000 ha in Spain and around 1 500 000 ha in Andalusia (southern Spain), where they are mainly located in the centre and northeast of the Guadalquivir River Valley and on the Betic Mountains hillsides (frequently up to 800 m.a.s.l.). The Guadalquivir Valley is a triangular area of 35 000 km2 located between the natural boundaries of Sierra Morena in the north and the Betic Mountains in the south (Fig. S1, Supporting information). The valley has historically been used for intensive agriculture of olives, cereals, vineyards and sunflowers. The area is generally flat, rising from sea level to 700 m near Sierras de Cazorla, Segura y Las Villas Natural Park where the Guadalquirvir River surfaces. The climate is Mediterranean with hot and dry summers and cool, humid winters. Mean annual temperatures range from 15 to 18·5 °C, and annual precipitation is between 400 and 1020 mm. Although virtually eliminated from the area, the Mediterranean maquis associated with Quercus ilex (accompanied by Quercus suber in the southwest of the Valley and Q. faginea in wetter conditions) is the natural climax vegetation across the entire region (Aparicio 2008). The area under olive grove area retracted and expanded repeatedly during the twentieth century, but since 1995 it has expanded in southern Spain, with a total of 120 000 ha previously devoted to cereal and other crops now dedicated to olive plantation (Junta de Andalucía 2003).

Agricultural landscape structure

In the central and upper Guadalquivir Valley (Córdoba and Jaén provinces), olive groves form a monoculture that results in relatively homogeneous landscapes. Native vegetation represents <1% of the area of the Gudalquivir Valley (Aparicio 2008). Intensification practices eliminated hedgerows and other live fences between properties, and the rare hedges still present are small, and often undetectable in the landscape. While woodlots and copses are extremely rare in lowlands, they are better represented in the olive groves on mountain hillsides. Topography is thus one of the rare sources of heterogeneity in the landscape and contributes to the differentiation between olive orchards from the valley and mountains. Other sources of heterogeneity in this agricultural landscape include differences between groves in herbaceous cover treatment (Valera et al. 1997) and number of trees per hectare (between 70 in old plantations and 400 in some young plantations; Muñoz-Cobo & Purroy 1980). These variations seem to more greatly affect the breeding than the wintering communities, which are more influenced by food availability (Rey, Alcántara & Sánchez-Lafuente 1996; Valera et al. 1997; Muñoz-Cobo & Moreno-Montesino 2003). Finally, another source of variation is the difference between cultivars. Many olive cultivars are planted in Andalusia (Junta de Andalucía 2003) but the most abundant are regionally separated (Fig. S1). Cultivars differ in olive ripening phenology, harvesting timing, fruit crop and the extent of their inter-annual variation (Fig. S2, Supporting information). These differences generate a regional mosaic in olive availability with consequences for avian frugivore occurrence in the agricultural landscape.

Olive orchards as wintering grounds for European frugivores

The Mediterranean environments in southern Europe and northern Africa are important wintering grounds for many European birds, including frugivores (e.g. Tellería, Asensio & Díaz 1999; Tellería, Ramírez & Pérez-Tris 2005). Frugivores are more abundant in wild olive forests (58–90 birds/10 ha) and other lowland habitats dominated by fleshy-fruited plants (Fig. 1a). However, loss of natural habitats and the decrease in fruit resources caused by unselective clearing of forest undergrowth (a major threat for biodiversity in the region; Andrés & Ojeda 2002) has displaced frugivores to olive orchards during winter (Muñoz-Cobo & Purroy 1980; Suárez & Muñoz-Cobo 1984; Rey 1993). Frugivore abundance in orchards of Andalusia ranges between 23 and 55 birds/10 ha (Rey 1993; Muñoz-Cobo et al. 2001). Thus, 3·5–8 million frugivores overwinter in olive orchards in southern Spain.

Figure 1.

 Frugivore assemblages in three lowland habitats of southern Spain. Data from sclerophyllous scrublands are from Jordano (1985, 1987a), Arroyo & Tellería (1983), Tellería, Ramírez & Pérez-Tris (2005) and Alcántara et al. (1997). Data from wild olive scrublands are from Suárez & Muñoz-Cobo (1984) and Rey (1992, 1993). Data from olive orchards are from Suárez & Muñoz-Cobo (1984) and Rey (1992, 1993). Means and standard error bars are shown for: (a) data on frugivore assemblage abundance (birds/10 ha), percentage of frugivores in the bird community, and number of species; (b) abundance of the most abundant frugivores; (c) abundance of the minor frugivores.

In Andalusia, between three and five species of frugivorous birds commonly occur in lowlands olive orchards. Sylvia atricapilla Linnaeus, S. melanocephala Gmelin and Turdus philomelos Brehm regularly overwinter in this habitat, whereas T. iliacus Linnaeus and Erithacus rubecula Linnaeus mostly appear in low densities early in winter. T. merula Linnaeus does not occur in lowland olive orchards but frequently occurs in orchards in mountain areas containing small remnants of native vegetation (Muñoz-Cobo et al. 2001; P.J. Rey, pers. obs.). Up to 11 frugivorous species can occur in wild olive scrublands and other lowland habitats (Fig. 1a). Locally, flocks of Sturnus unicolor Temminck and S. vulgaris Linnaeus and the crows Corvus monedula Linnaeus, C. corone Linnaeus, C. corax Linnaeus, Pyrrhocorax pyrrhocorax Linnaeus and Cyanopica cyanus Pallas may also consume fruits in the three habitats (Jordano 1987a; Blanco, Fargallo & Cuevas 1994; Alcántara et al. 1997). Based on their inter-habitat variation (Fig. 1b,c; Table S1, Supporting Information), T. philomelos, T. iliacus and S. atricapilla are the least affected by habitat changes, whereas T. merula, T. viscivorus Linnaeus, E. rubecula, Phoenicurus ochruros Gmelin, S. melanocephala, S. undata Boddaert, C. cyanus and Garrulus glandarius Linnaeus are seriously affected (Rey 1993). Some of these birds, like T. philomelos, are declining in Northern Europe (Peach, Robinson & Murray 2004), and olive orchards appear to represent an important winter surrogate habitat for turdids.

Disparities and similarities between orchards and natural habitats

Habitat structure

Olive orchards are mono-specific stands with olive trees uniformly separated (6–12 m between trees). Tree cover ranges between 5% in young plantations and 30% in old plantations. Scrub layer is predominantly lacking and the herbaceous layer is ephemeral in most groves due to the use of herbicides and ploughing (Valera et al. 1997). Their spatial heterogeneity and habitat structure are extremely simple compared to native scrublands and forests (Rey, Valera & Sánchez-Lafuente 1997).

Fruit size

Cultivated olives are larger than the winter fruits of the Mediterranean region. The average width of the fruits in the scrublands of southern Spain is 6·9 mm for fleshy fruits, 7·6 mm for berries with one to three seeds and 10·7 mm for berries with more than three seeds (Jordano 1984). The mean width of wild olives is 8·4 mm, whereas cultivated olives range from 12·7 to 17·8 mm (Rey & Gutiérrez 1996; Rey et al. 1997). Therefore, the oversized fruits of cultivated trees do not fit the gape of Mediterranean avian frugivores and suitable fruits are scarce compared to wild olives or other fruits (Table 1).

Table 1.   Body metrics of avian frugivores and associated fruit consumption and feeding behaviour. Predominant diet and % fruit in diet refer to frugivory in natural habitats. Fruit suitable to gape is stated in terms of percentage of fruits fitting the bird’s gape. Frequency of pecking is the percentage of total olive consumption obtained through pecking. Swallowing and pecking energetic is the energy rendered by each feeding behaviour. Data from Rey & Gutiérrez (1996, 1997); Rey et al. (1997)
 Sylvia atricapillaSylvia melanocephalaSylvia undataTurdus philomelosErithacus rubecula
Body mass (g)18·511·59·170·017·0
Gape width (mm)8·57·15·613·78·0
Predominant dietHeavily frugivorousOmnivorousInsectivorousHeavily frugivorousHeavily frugivorous
% Fruit in diet>70%50–70%<25%>70%>70%
Olives suitable to gape0–2%002–75%0
Wild olives suitable to gape60–88%13–58%<10%100%35–79%
Frequency of olive pecking>95%>95%No data<60%>95%
Pecking energetic (kJ s–1)0·03 ± 0·02 No dataNo data0·05 ± 0·02No data
Swallowing energetic (kJ s–1)0·12 ± 0·04 No dataNo data0·63 ± 0·08No data

Diversity of fleshy fruits

The diversity of fruit is poor in the olive orchard landscape because there are fewer hedges and copses next to streams supporting other fruit-bearing plants (Table S2, Supporting information).Wild olive-dominated scrublands are diverse and can support 6–10 different species of fruiting plants. Fruit diversity increases further in other Mediterranean scrublands in southern Spain (Herrera 1984a; Jordano 1984) with 16–21 fruit species per site (Table S2).

Availability of alternative foods for birds

Arthropod availability over the winter is higher in natural habitats than in olive orchards (Fig. 2), probably as a consequence of agricultural practices (pesticide, ploughing) and habitat simplification.

Figure 2.

 Monthly abundance of arthropods in lowland habitats of southern Spain: olive orchards (open circle), wild olive scrublands (black squares) and other scrublands (black triangles). To compare monthly abundances, the results of Friedman anova (ground) and Wilcoxon test (tree branches) are shown. Data from olive orchards and wild olive scrubland are from Rey (1992) and Rey & Valera (1999). Data from other scrublands are from Jordano (1989). Only arthropod abundance on the ground is available from other scrublands. Arthropods on the ground were sampled with pitfalls (0·38 dm2) in orchards and wild olive scrublands, and with sticky plastic sheets in other scrublands. Arthropods on branches were sampled by five-minute censuses.

Nutrient composition of the fruits

Cultivated and wild olives have very high lipid contents that range from 40 to 66% of dry mass (Jordano 1987a; Rey 1992). Among the Mediterranean fruits, only Pistacia lentiscus L. (58·8%), Pistacia terebithus L. (55·6%) and Laurus nobilis L. (54·3%) have such a high lipid content (Herrera 1987). Thus, olives are among the most energy-rich fruits in the Mediterranean Basin.

Fruit abundance

Production of fruit in the scrublands (Herrera 1984a; Jordano 1985, 1992) approaches the levels of orchards in terms of fruits ha–1, reaching a maximum of 3 × 106 (Muñoz-Cobo 1987). However, there is extreme inter-annual and among zones variation in fruit crop yield (Rey 1995; Fig. S2, Supporting information), which is related to supra-annual cycles of fruit production (Loussert & Brousse 1980). This is common in many scrublands in the Mediterranean region (Herrera 1984a, 1998), as well as in the wild olive fruit production (Jordano1987a; Alcántara et al. 1997).

Fruiting and ripening phenology

Cultivated and wild olives have winter ripening phenologies (Fig. S2 Supporting Information), common to many Mediterranean fruits (Herrera 1984a; Jordano 1984; Alcántara et al. 1997). Their maximum ripe fruit availability (December to February) matches the winter peak of frugivore abundance in the Mediterranean Basin (Jordano 1985; Rey 1995; Herrera 1998).

The pre-adaptation hypothesis

The olive cultivation zones in the Mediterranean Basin were originally occupied by Mediterranean scrublands or forests. Autumn and winter lipid-rich fruits, such as Pistacia lentiscus, P. terebinthus, Olea europaea L., Viburnun tinus L., Phillyrea angustifolia L, Jasminum fruticans L. and Hedera helix L., are well represented in these plant communities (Herrera 1984a, 1995; Jordano 1984). Because of their high reward, these fruits are major components in the winter diet of the Mediterranean avian frugivores (Herrera 1981, 1984a; Jordano & Herrera 1981; Jordano 1987a,b, 1988, 1989). In the orchards, olives account for the majority of the frugivore diet (Tejero, Camacho & Soler 1983; Tejero, Soler & Camacho 1984; Soler, Tejero & Camacho 1988; Soler et al. 1988; Rey & Valera 1999). Some frugivorous birds thus appear to be pre-adapted to olive culture due to structural and fruit-related similarities between olive orchards and the native scrublands that frugivores naturally inhabit (Muñoz-Cobo 1987).

Pre-adaptive features that favour the settlement of birds in olive orchards

Five pre-adaptive features influence the successful occupation of olive orchards by a bird species.

1. High winter frugivorism

Frugivorism intensity (sensuHerrera 1984a,b; Jordano 1987c) of a bird species affects the successful occupation of olive orchards. The relative decrease in abundance between orchards and natural habitats (Table S1) is significantly related to percentage of fruit in the diet (r = −0·84, = 0·005) and the frugivorism index (r = −0·69, P = 0·038). Hence, birds with greater frugivorism in their natural habitat will experience fewer differences in their population abundance between natural habitats and orchards. In contrast, insectivorous birds will be likely to experience greater decreases in abundance (r = 0·80, P = 0·01). This pattern is found even within the same genus. Among the three overwintering Sylvia species, there is a size gradation directly related to the frugivorism intensity (Jordano 1987b) and to the successful occupation of the orchards. Sylvia atricapilla (18·5 g of body weight) is heavily frugivorous, S. melanocephala (11·5 g) is omnivorous and S. undata (9·1 g) is mostly insectivorous (Table 1). Thus, S. atricapilla is abundant in olive orchards and in scrublands, whereas S. melanocephala is less abundant and S. undata is extremely rare in orchards (Fig. 1; Table S1).

2. Fruit handling plasticity to cope with oversized fruits

The scarcity of fruits of suitable size for frugivores in olive orchards may limit their successful occupation of this habitat. Among those species overwintering in olive orchards, the largest species (T. philomelos, T. iliacus and S. atricapilla) survive better in orchards than smaller species (E. rubecula, S. melanocephala and S. undata) (Fig. 1; Table S1). Within the Sylvia genus, the larger species have greater gape width and are consistently more abundant in olive orchards. Olive orchards are clearly unsuitable winter habitats for S. undata and S. melanocephala, because of the limitations imposed by the size of the available fruit, but they are more appropriate for S. atricapilla.

Some birds develop an opportunistic feeding behaviour (pecking olive pulp) to cope with the low availability of suitable fruits. Pecking is particularly frequent for S. atricapilla, but larger birds such as T. philomelos, and smaller birds such as E. rubecula and S. melanocephala, also peck fruit. The feeding efficiency of this habit is less profitable than swallowing whole fruit (Table 1) but it is still profitable enough to defray the metabolic requirements of these birds (Rey & Gutiérrez 1996, 1997; Rey et al. 1997). These birds have also been reported to peck fruit in cherry orchards (Hernández 2008) and large wild olives and other fruits in natural habitats (Jordano 1987a; Alcántara et al. 1997). The capacity of some birds to peck as an alternative to swallowing the fruits is a pre-adaptive feature that allows frugivores to survive in croplands where fruit size has been increased by artificial selection.

3. Opportunism and diet plasticity to cope with low fruit diversity and unbalanced foods

Birds have very different diets in olive orchards and natural habitats (Fig. 3). In orchards, birds feed on fewer fruit species and the amount of fruit in the diet decreases in comparison to natural scrublands (Rey & Valera 1999). Consumption of arthropods is no lower in orchards relative to scrublands despite the decrease in arthropods in orchards (Fig. 3). It seems that frugivores actively search for arthropods to diversify their diet and to balance the intake of nutrients and energy. Most surprisingly, frugivores in olive orchards systematically include large amounts of leaves, flowers and seeds of weeds in their winter diets (Fig. 3). Consumption of this suboptimal food suggests feeding plasticity to balance nutritional requirements, especially the need for minerals and micronutrients (Rey & Valera 1999).

Figure 3.

 Variation in winter diet of four avian frugivores in four lowland habitats of southern Spain: olive orchards, hedges in orchards, wild olive scrublands and other natural sclerophyllous scrublands. Numbers on each bar indicate number of fruit species per diet sample. NFPF refers to plant matter other than fruit. Data for olive orchards and wild olive scrublands are from Rey (1992), Rey, Alcántara & Sánchez-Lafuente (1996) and Rey & Valera (1999), complemented for some species with Soler et al. (1988), Soler, Tejero & Camacho (1988), Tejero, Camacho & Soler (1983) and Tejero, Soler & Camacho (1984). Data for other scrublands are from Jordano & Herrera (1981), Herrera (1984a), Jordano (1987a) for S. atricapilla and S. melanocephala, Herrera (1981) for E. rubecula, and Herrera (1984a) for T. philomelos.

4. Bird ecomorphology, foraging niche and the ability to cope with the structural simplicity of the orchard habitat

Birds feed on olives from branches or on the ground (Rey & Gutiérrez 1997) although this is limited by the birds’ external morphology. Thrushes and starlings are ground foragers (Smith 1974; Snow & Snow 1988) and consume olives on the ground; in contrast, Sylvia species are perching birds (Leisler 1980) and consume olives in the trees (Rey & Gutiérrez 1997). Differences in perching ability associated with hind limb morphology have been described for Sylvia (Leisler 1980; Leisler & Winkler 1985; Jordano 1987c). S. atricapilla forages frequently in trees and is better adapted to perch than its congeners S. melanocephala and S. undata, which forage in medium and small shrubs (Cuadrado 1987; Jordano 1987b). Erithacus rubecula (less adapted than the Sylvia species to forage using perches) forages on the ground among the vegetation or consumes fruits by hovering (Snow & Snow 1988). Olive orchards thus attract fruit gleaning-perching birds and ground foragers but cannot support strict scrub foragers. Compared to scrublands, the probability of successful occupation of orchards is 0·75 for perching frugivores, 0 for strict scrub foragers and 0·33 for other frugivores (log-likelihood = −8·24, d.f. = 2, P = 0·03; N = 12 species; bird categorisation in Table S1). It seems that lack of shrubs in orchards directly constrains bird occurrence. The lack (T. merula), rarity (S. undata) or pronounced decline (E. rubecula and S. melanocepahla) of some Mediterranean frugivores in orchards is probably related to structural simplification and lack of a shrub understorey in the orchards.

Ecomorphological and niche constraints to bird occupation of fruit orchards are not exclusive of olive landscapes. A paucity of understorey birds in fruit orchards, compared to remnants of native vegetation, has been reported in other regions and linked to insufficient shrub understorey (Little & Crowe 1994; Round, Gale & Brockelman 2006).

5. The ability of birds to track spatio-temporal variation in fruit availability

Olive groves offer a continuum of olive supply across spatial and temporal scales. This continuum of ripe fruit availability is a consequence of the variation between different regional zones in olive crops, ripening phenology and harvesting (Rey 1995), which are all linked to the peculiarities of the cultivars planted at each locality (Figs S1 and S2). By virtue of such a continuum, the olives become a predictable resource at a regional and landscape level each winter from November to March. In order to survive, birds must track the availability of ripe olives in space and time. The intrinsic variation (in phenology and location) between cultivars and the landscape continuity of the orchards are crucial for frugivore populations. Many studies have shown that frugivores are able to track spatio-temporal variations of fruit availability (Levey 1988; Mogenburg & Levey 2003) and that such variation becomes fundamental to survival in a region (see Tellería, Ramírez & Pérez-Tris 2005; for implications on avian frugivore conservation in the Mediterranean Basin). Tracking is possible because some birds exhibit nomadism and regional migration (Levey & Stiles 1992). Sylvia atricapilla and some Turdus species (especially T. philomelos) have winter nomadic behaviour (Simms 1978; Debussche & Isenmann 1984; Tellería & Pérez-Tris 2003) that allows them to respond to spatial and temporal changes in fruit availability, and they have been found tracking olive availability at both small and regional scales (Rey 1995). By contrast, S. melanocephala, E. rubecula and T. merula do not display winter nomadism (Debussche & Isenmann 1984; for E. rubecula; Lundberg 1985; for T. merula; and Cantos 1992, for S. melanocephala) and do not track olive abundance in orchards. Thus, olive orchards are suitable habitats for frugivores with high nomadic movement and regional migration and unsuitable for frugivores with low nomadic capacity.

Managing frugivore biodiversity in Spanish olive orchards

Pre-adaptive features are fundamental to the successful occurrence of some birds in olive orchards during their winter stay in the Mediterranean region. Although olive orchards differ from natural habitats within the region, they still maintain structural and functional similarities. Interestingly, most of these pre-adaptive features are properties of the most abundant frugivores (S. atricapilla and T. philomelos) but are not as pronounced in less frequent frugivores. Orchard management for increasing frugivore diversity should encourage habitat diversity to facilitate bird movement, diet diversification, fruit foraging and niche requirements. It is suggested that this can be achieved by two inter-related approaches to management:

1. Management of the olive orchards structure: towards an increased heterogeneity

Heterogeneity should be managed at several spatial scales (Benton, Vickery & Wilson 2003). At small scales, the vertical and horizontal structural heterogeneity may be enhanced by extending the practice of separating olive orchards with hedges. This would provide a shrub layer for those frugivores for which this is a fundamental niche requirement (providing protection, foraging sites and movement facilities). Mist-netting in hedges demonstrates that 2–5 m-width hedges are enough to increase the local occurrence of some species (Fig. S3 Supporting information). At the landscape scale, hedgerows, rocky outcrops, copses, and stream vegetation belts should be promoted. All these elements diversify the olive orchard landscape, create heterogeneity and complexity, and connect the landscape for non-nomadic birds. At the regional scale, the range of distinct, locally predominant olive cultivars must be maintained, since differences in ripening of different cultivars increases the timeframe during which olives are available to birds. The homogenisation of orchards to only a few cultivars would compromise frugivore overwintering in the region.

2. Management of the food availability

Hedgerows should be planted with native species that produce fleshy fruits to diversify the fruit resources. The occurrence of wild fruits in remnant hedges around olive orchards makes these orchards more similar to natural habitats in fruit diversity and phenology. Hedgerows and other sources of fruit (e.g. groups of isolated fruiting trees) have been found to maintain frugivory in agricultural landscapes worldwide (Hinsley & Bellamy 2000; Croxton & Sparks 2004). Hedges in the agricultural landscape of southern Spain may contain not only winter but also summer and autumn fruits that would help support other migrant birds that display summer and autumn frugivory and use native habitats, but not olive orchards, as stopover sites. Hedges also increase the abundance of arthropods and other food resources, as demonstrated in many agricultural landscapes (Thomas & Marshall 1999; Pollard & Holland 2006).

The increase of organic agricultural practices in the region, especially reducing the use of pesticides, should be highly beneficial for winter frugivores by increasing arthropod availability as a food resource. A reduction in the use of pesticides has been shown to enhance bird diversity in fruit farmlands in the Mediterranean (Genghini, Gellini & Gustin 2006) and other regions (Chamberlain, Wilson & Fuller 1999; Freemark & Kirk 2001).

Generalizations to olive landscapes in the Mediterranean Basin

Much of the patterns of habitat simplification and homogenization described for olive landscapes in southern Spain are common to other olive landscapes in the Mediterranean. Agricultural intensification has been common in Mediterranean Europe since the 1950s (Kizos & Koulouri 2006),with increasing mechanization, removal of natural vegetation, fertilization and the use of pesticides, all resulting in an overall decrease in fruit supplies and other resources for birds (Genghini, Gellini & Gustin 2006). Other olive cultivation zones are also experiencing similar phenomena. For example, Santos & Cabral (2003) showed detrimental effects of olive cultivation intensification for several passerine guilds in Portugal. Thus, the recommendations to promote landscape heterogeneity and diversification of food resources are also applicable to most olive cultivation zones of the Mediterranean Basin. However, substantial differences exist in the extent and continuity of olive groves. In France, Italy and Greece, olive groves are frequently interspersed with other crops, producing agricultural mosaics. Live fences (hedgerows and vegetated stonewalls) and woodlots are better conserved in many of the agricultural landscapes in these countries compared to Mediterranean Spain (Genghini, Gellini & Gustin 2006; Kizos & Koulouri 2006) and, consequently, there is a greater diversity of food sources, including fruits. Ancient olive plantations in some regions of Italy and Greece have been maintained and provide refuge for biodiversity. There are action plans for the protection of this habitat (LIFE07 NAT/IT/000450), but no similar conservation strategies are being developed in Spain. The habitat simplification scenario is to some extent different for many olive cultivation zones of North Africa, where mechanization and use of pesticides is less common and the olive trees are frequently interspersed with other cultivated trees and some natural vegetation (P. J. Rey personal observation). No detailed information on bird abundance and frugivory in olive landscapes exists outwith Spain, but it would be expected that such scenarios would be favourable to frugivore conservation. However, the expansion and intensification of olive cultivation that has happened in North Africa in the last two decades is putting pressure on these heterogeneous landscapes and it would be expected that this would have adverse impacts on wintering populations of Mediterranean frugivores.

Lessons from olive orchards. Comparison with other fruit croplands

A fundamental lesson from studies of frugivory in Spanish olive orchards is that the cultivation of fruit crops derived from native instead of exotic plant species will better preserve the original animal biodiversity of the region. Such agricultural landscapes maintain some of the structural and functional (the plant–animal interactions) properties of the natural habitats to which animals are adapted. On the other hand, it is important to acknowledge that different bird species have different pre-adaptive features that will enable them to thrive in agro-ecosystems. Most fruit croplands of the world are affected by intensification, landscape and habitat structural simplification and human selection of fruit size. As a result there are often food shortages for frugivores (e.g. Little & Crowe 1994; Nelson et al. 2000) similar to those described in olive orchards. It would therefore be expected that the pre-adaptive features influencing bird diversity in olive orchards will also be relevant in other fruit production systems.

This review demonstrates that modifications of the agricultural practices in a region can easily make agro-ecosystems more suitable to a greater number of bird species by incorporating features that will favour birds poorly adapted to croplands (Benton, Vickery & Wilson 2003). In particular, hedgerows are known to have positive effects on local bird diversity in other agro-ecosystems (e.g. Hinsley & Bellamy 2000; Herzon & O’Hara 2007). Likewise, the retention of natural elements in the agricultural landscape has been shown to increase bird diversity in many other regions (e.g. Haslem & Bennett 2008a,b). These natural elements connect the landscape (Hinsley & Bellamy 2000; Donald & Evans 2006) by acting as natural corridors and also provide additional food resources, which are fundamental to a diverse diet.

Finally, we can make tentative generalizations from the comparison of olive orchards with other fruit croplands claimed as important reservoirs for biodiversity. Rustic (shade) coffee plantations in Central America have repeatedly been proposed as functional surrogates of the tropical forest for biodiversity (reviewed in Philpott et al. 2008). Coffee plantations are exotic in these areas, but their function for biodiversity is achieved from the structural and taxonomical similarities with tropical forests due to the species that provide shade for coffee production in rustic plantations. Coffee plantations have become fundamental as winter refuges and stopover sites for Neotropical migrant birds because their structural complexity and taxonomical diversity provide suitable food sources and niche requirements. However, modern sun plantations are structurally and taxonomically simplified, mirroring to some extent some phenomena occurring in olive orchards. Most native plant species are removed leading to habitat homogenization, reductions of insects and fruits (food supplies for birds), and a concomitant reduction of bird biodiversity. Unlike olive cultivation zones, however, there is increasing awareness of the importance of bird conservation in agricultural landscapes of the Neotropics. The repercussions for biodiversity of different management regimes in coffee plantations are being thoroughly investigated in these systems (Philpott et al. 2008). There are rigorous certification protocols for defining biodiversity-friendly practices as stand-brand for the coffee market (Mas & Dietsch 2004). Similar certification programmes for olive production should be encouraged to conserve frugivorous/insectivorous European migrant birds in their winter Mediterranean quarters.

Future directions

Research programmes should be implemented to improve our knowledge of the role of different agricultural practices (organic vs. integrated and conventional farming) in maintaining functional biodiversity as well as landscape properties in olive orchards. A detailed knowledge is needed of landscape structure, the occurrence of natural elements in the landscape, the connectivity of these elements, their function and their relationship to biodiversity. Future studies should involve experimental trials combining different hedgerows settings and different amounts of herbaceous cover. These investigations should consider the effects of biodiversity-friendly practices on crops and the economics of crop reduction and subsidy inputs after achieving the eco-conditionality criteria of European Commission regulations. Avian studies in orchards need to be extended to olive cultivation zones in North Africa and the Eastern Mediterranean to allow a complete picture to be formed of the role of olive orchards as reservoirs for frugivorous/insectivorous European birds.

Acknowledgements

I thank comments from F. Valera, J. E. Gutiérrez and three anonymous referees. R. Zamora encouraged me to write this review. While writing this paper I was funded by project CGL2006-02848/BOS of the Spanish MEC and FEDER.

Ancillary