Intensification of agriculture results in habitat destruction and has a negative effect on biodiversity. According to Pérez-Bote & Romero (2012), biodiversity in agricultural habitats is influenced by the surrounding landscape. A diverse plant community can influence beneficial arthropod populations by providing food or habitat resources that might not be found in a simple plant comunity (Costello & Daane, 1998). Clough et al. (2007) found that in arable land, biodiversity depends on recolonization from surrounding perennial habitats, which serve as overwintering habitats and contain alternative resources for arthropods. Beneficial organisms that are not resident in vineyards all year round must recolonize on a seasonal basis, most likely from natural vegetation in the surrounding landscape (Duelli & Obrist, 2003). Flowering plants provide nectar and pollen resources to insects during the growing season (Ambrosino et al., 2006; Blaauw & Isaacs, 2012). Roschewitz et al. (2005) and Gardiner et al. (2009) reported that vegetational diversity can also provide support for insect biological control at the local and landscape levels. Weeds play an important role in enhancing the abundance and diversity of arthropod predators and serve as a source of increased diversity in agroecosystems. In most agroecosystems, weeds are ever-present biological components within and around fields, adding to the complexity of interacing trophic levels which mediate a number of crop insect interactions with major effects on final yields (Nicholls & Altieri, 2012). However, weeds are traditionally viewed as plants that reduce yields by competing with crops or by harbouring pests and plant pathogens (Penagos et al., 2003). Increased diversity has been the rationale for enhancing biological control of arthropod pests through habitat management (Norris & Kogan, 2005). The same authors (Norris & Kogan, 2000) indicated that weed cover enhances the number and activity of spiders and ground beetles. In their research on dicotyledonous weeds, Wilson & Aebisher (1995) reported that the density of most arthropod species decreased significantly as distance from crop edge increased from 0 to 128 m. Some authors (Winkler, 2005; Bàrberi et al., 2010) have reviewed the importance of vegetation diversity for enhancing populations of beneficial arthropods in cropland. Wyss (1996), Simon et al. (2010) and Song et al. (2010) reported a positive effect of plant comunity diversification on beneficial arthropods in orchards. Non-crop habitats bordering agricultural fields in Europe have been found to have a favourable effect on a number of beneficials as spiders, ladybugs, and syrphids (Hillocks, 1998; Ernoult et al., 2013). Woodcock et al. (2008) showed the positive effects of composition and diversity of plants around the field margins on ground beetle diversity. Fields with a dense weed cover and high diversity usually have more predaceous and parasitic arthropods than weed-free fields (Speight & Lawton, 1976). Even now a lot of grape producers destroy weeds not only in fields but also in the surrounding landscape, such as field margins, field patches and non-cultivated areas. In the Zadar County (Croatia) vineyards, typical grape production is organized in monoculture, what leads to landscape simplification and decrease of predatory arthropod population. Vegetation surrounding the fields has a great impact on predator abundance in the vineyards. The main goal of this paper was to examine the influence of vegetation (weed) structure on predatory insects and spiders of three vineyards.

This work explored the influence of ecological infrastructure and plant composition on arthropod population and diversity. Arthropod communities clearly differed between sites and were dominated by Aranea. In agricultural landscapes, spiders are generally most diverse in semi-natural habitat (Duelli et al., 1999). In this research, spiders were most abundant at the site OP, especially in zone C (pastures and meadows at wood vicinity) as suggested by Isaia et al. (2006). Bruggisser et al. (2010) reported that spider and plant richness was not higher in organic compared to conventional farming. The main reason for spider dominance is probably their broader diet and their ability to survive a long period without prey (Costello & Daane, 1999). Spiders also may benefit from higher weed populations (especially ED and BI sites), which provide higher structural complexity and potentially increase the availability of herbivore prey (Schmidt et al., 2005; Woodcock et al., 2013). We noticed that dry vegetation also affected to a high number of spiders and predaceous insects. Among Coleoptera the most abundant family was Coccinellidae presented mostly by Hippodamia spp. particularly in zone IB(C), possibly as a result of presence of their prey on plants such as Daucus carota (Burgio et al., 2006). In California vineyards, Daane et al. (2008) also found Hippodamia spp. and Scymnus spp. as important predators of mealbugs. Kopta et al. (2012) reported that plants like Amaranthus retroflexus L., Centaurea cyanus L. Daucus carotta L., and Foeniculum vulgare L. were present in almost all sites where ladybugs were found in high number. These plants provide alternative food sources such as pollen and nectar from extrafloral nectaries. As Raymond et al. (2000), we also found ladybugs (all developmental stages) on Chenopodium spp. Gaigher & Samways (2010) indicated that abundance of carabids was higher in organic than in the integrated vineyard. O’Rourke et al. (2008) and Eyre et al. (2012) reported that crop rotation and diverse crop habitats increased Carabidae activity, density and diversity. Their conclusions correspond with our results. The highest abundance of Carabidae found at site IB may lay in the fact that integrated vineyard was surrounded with arable land producing grains and vegetables with a lot of weed strips. Management of weed strips appears to increase the availability of food for carabids and results in enhanced reproduction (Zangger, 1994; Zangger et al., 1994). The other reason probably lays in the fact that ground surface of this vineyard (IB) was managed with cover crops and mulched several times during the vegetational period (Thomson & Hoffmann, 2007). As Burgio et al. (2006) reported, sampling by sweep net is able to collect only terrestial Carabidae present in the weed canopy, and not those on the soil. Green lacewing (Chrysopidae) showed preference to weed margins where plants such as Chenopodium album L., Convolvulus arvensis L. and Anthemis arvensis L. are found (Ruby et al., 2011). Beside, as food source, some weeds as Trifolium pratense L., Centaurea cyanus L., Papaver rhoeas L. and Vicia spp. seem suitable for lacewings as oviposition site. Our data agree with those reported by Eichenberger (1991), who also found lacewings on the plants above mentioned. Diptera from family Syrphidae were always more abundant in zone A of all sites where Apiaceae and Asteraceae dominated (Tooker et al., 2006). Syrphidae prefer grassy strips based on total botanical diversity (Speight, 2008). Plants from these families (Apiaceae and Asteraceae) are known as nectar and pollen producers in great amounts so they play an important role in Syrphidae attraction (Branquart & Hemptinne, 2000; Rebek et al., 2005; Morales & Kohler, 2008; van Rijn & Wäckers, 2010). As expected, the low number of syrphids at OP could be due to the low presence of flowering weeds. Costello & Daane (1999), Nicholls et al. (2000, 2008) and Altieri et al. (2005) also reported several dominant Heteroptera predators such as Nabis spp., Orius spp. and Geocoris spp. in the ecological infrastructure of vineyards. Insect predators such as Anthocoridae and Geocoridae prefer thrips, aphids, lepidopteran and hemipteran eggs and spider mites. High number of Anthocoridae was found at site IB (A) at the vicinity of apple orchard. One potential explanation is that these predators are strongly associated with aphids, mites and thrips which are always present in apple orchards (Rieux et al., 1999; Burgio at al., 2006; Gadino et al., 2012)Prischmann et al. (2005) found no significant differences in Heteroptera abundance between commercially managed and unmanaged vineyards. Daane et al. (2008) also found these bugs as not commonly present in large numbers in vineyards. Miridae were represented by Macrolophus spp. and Dicyphus spp. We noticed a relationship between Macrolophus spp. and the plant Ditrichia viscosa (L.). Greuter (Asteracea) as has also been documented in Greece (Perdikis et al., 2007). That can explain a higher number of these bugs in DE and OP, while in site IB D. viscosa was not present.

This study shows that ‘plant rich’ weedy margins (zone A) can enhance the abundance of some predators. Our results agree with some other studies showing that the introduction of flowering plants into agricultural settings leads to increased arthropod abundance (Rebek et al., 2005; Walton & Isaacs, 2011). Higher arthropod diversity was associated with greater compositional diversity of weed cover (Benton et al., 2003; Gaigher & Samways, 2010). Our results showed that the number of total insects was lower in meadows and pastures than in wildflower strips of field paths and weedy areas, what is similar to the results of Zurbrügg & Frank (2006). In almost all localities, vegetation structure was the best explanatory factor for insect distribution, abundance, and species richness. Our results are similar to the results reported in other studies of arthropods in organic and intergrated vineyards (Gaigher & Samways, 2010). The best plant families for preserving predators (Coccinellidae, Syrphidae and Chrysopidae) are Apiaceae, Asteraceae and Chenopodiaceae (Fiedler et al., 2008; Bertolaccini et al., 2011). Some studies also showed that organic management systems increased the arthropod abundance and richness but not the diversity (Clark, 1999). Overall, these results suggest that despite the assertion that organic fields contribute to increase biodiversity, arthropod abundance and diversity also depend on plant composition. Sites and zones with higher number of weed plants species especially those with attractive flowers obviously provide favourable conditions for natural enemies. Conservation and attendance of ecological infrastructure such as weedy margin can serve as habitat for beneficial fauna. Therefore it can play an important role in increasing the vineyard production quality.

In summary, this study shows the importance of non-cultivated areas on the abundance of predatory arthropods. Ecological infrastructure, especially weed margins, field paths and sorrounding wildflower areas, are important components of the vineyard sistem as they enhance the plant abundance and diversity. Despite the vineyard management, our conclusion is that plant community plays an important role in attracting and maintaining populations of arthropods. Regular approach to weed management in vineyards can increase biodiversity of beneficial organisms and achieve a more sustainable agroecosytem.