Effects of Traditional Orchard Abandonment and Landscape Context on the Beneficial Arthropod Community in a Mediterranean Agroecosystem

Abstract

Agricultural abandonment is one of the main land-use changes in Europe, and its consequences on biodiversity are context- and taxa-dependent. While several studies have worked on this topic, few have focused on traditional orchards, especially in different landscapes and under a Mediterranean climate. In this context, we aimed to determine the effects of almond orchard abandonment on the communities of three groups of beneficial arthropods and the role of the landscape context in modulating these effects. Between February and September 2019, four samplings were carried out in twelve almond orchards (three abandoned and three traditional (active orchards under traditional agricultural management) located in simple landscapes as well as three abandoned and three traditional in complex landscapes). Abandoned and traditional almond orchards harbor different arthropod communities and diversity metrics that are strongly conditioned by seasonality. Abandoned orchards can favor pollinators and natural enemies, providing alternative resources in simple landscapes. However, the role that abandoned orchards play in simple landscapes disappears as the percentage of semi-natural habitats in the landscape increases. Our results show that landscape simplification, through the loss of semi-natural habitats, has negative consequences on arthropod biodiversity, even in traditional farming landscapes with small fields and high crop diversity.

Keywords: abandonment; bee; landscape complexity; parasitoid; spider; traditional almond orchard.

Figure 1

Location of the study area in the Salamanca province (Western Spain) and the placement of the 6 landscape windows (yellow: low level of semi-natural habitats, green: high level of semi-natural habitats) and 12 almond orchards (blue triangles: abandoned almond orchards, orange circles: traditional almond orchards).

Figure 2

Sampling design. Black circles: pitfall traps; empty squares: 2 m × 2 m vacuuming quadrants; yellow, white, and blue circles: pan traps. Dashed arrows represent the minimum distance between traps or to the grove’s edge.

Figure 3

MDS for the spider, parasitoid, and bee communities sampled (Bray–Curtis index, square-root–transformed abundances). Polygons include the samples belonging to the same category of the variable system or level of semi-natural habitats for each sampling period. Only polygons for which significant pairwise comparisons were detected in the PERMANOVA are represented. SNH = semi-natural habitats.

Figure 4

Results of the linear models for spider abundance (a,b) and richness (c–e). Points with error bars represent estimated means and 95% confidence intervals from the models’ predictions, and dull dots represent the raw data. Significant differences (p < 0.05) are shown with asterisks. Differences in pairwise comparisons are shown with different letters. ns: not significant.

Figure 5

Results of the linear models for parasitoid abundance (a–c) and richness (d). Points with error bars represent estimated means and 95% confidence intervals from the models’ predictions, and dull dots represent the raw data. Significant (p < 0.05) differences are shown with asterisks and marginally significant (0.05 < p < 0.1) differences with dots. ns: not significant.

Figure 6

Results of the linear models for bee abundance (a,b) and richness (c,d). Points with error bars represent estimated means and 95% confidence intervals from the models’ predictions, and dull dots represent the raw data. Significant differences (p < 0.05) are shown with asterisks. Significant differences (p < 0.05) in pairwise comparisons are shown with different letters. ns: not significant.