Is the Abandonment of Organic Grassland a Threat to Alpine Insect Diversity?

Abstract

Land abandonment is a multifaceted, nonlinear, worldwide phenomenon that is influenced by a variety of factors and opinions. The goal of this study was to understand the significance of land abandonment for true bugs and syrphids in three grassland management regimes that includes abandoned, intensive, and extensive alpine organic grasslands. In 2021 and 2022, we sampled true bugs and syrphids by applying observation plot and sweep netting sampling methods. Extensive grasslands had significantly higher true bug and syrphid abundance compared to abandoned grasslands. However, no difference of species richness was found in studied grassland regimes. Large numbers of unique species (25.5% true bugs and 21.5% syrphids) only occurred in the abandoned grasslands but not in intensive and extensive grasslands. Similarly, true bug assemblages in abandoned grasslands differed significantly from assemblages in intensive and extensive grasslands. We found that extensive grassland can manage to increase true bugs and syrphid abundance. Likewise, undisturbed abandoned grassland is not a threat to insect diversity, and supports the survival of more unique true bug and syrphid species. A mosaic landscape consisting of abandoned grassland along with grassland having different, mainly extensive, management intensity could be an ideal arrangement for alpine biodiversity conservation.

1. Introduction

In recent decades, agricultural systems and land usage in Europe have seen a tremendous transition due to interdependent changes in the social, financial, technical, and cultural context [1,2]. Such a transformation in agricultural areas caused landscape and biological alterations that resulted in land abandonment [3,4], which is associated with a loss of insect and plant biodiversity [5,6]. However, the literature also showed that land abandonment could enhance overall heterogeneity and biodiversity at different scales [7]. Therefore, two distinct land abandonment scenarios are possible: on the one hand, land abandonment may assist with passive landscape restoration [8], which will help to support biodiversity while minimizing the direct human effect on the environment. On the other hand, farmland biodiversity may be threatened by the land abandonment of agricultural areas [9]. The space allocated for agricultural areas has been decreasing especially in the European traditional alpine grassland “EUROSTAT, http://www.europa.eu/ (accessed on 2 February 2023)”. Among the various habitats, permanent alpine grassland provides a substantial contribution to the insect biodiversity of alpine agroecosystems [10,11]. However, between 1980 and 2000, 40% of grassland areas in the Alps were abandoned, which resulted in a reduction of 17% in livestock units [12]. Alpine grassland contributes greatly to the insect biodiversity of mountain agro-ecosystems while also providing a broad range of ecosystem services with socioeconomic value to human society [5]. Due to intensification, insect diversity in remaining agricultural areas, particularly in alpine grassland of high conservation value, is coming under more and more pressure. There is a consensus that intensive grassland management practices like frequent mowing or intense grazing have a negative impact on the diversity of insects [13,14]. To avoid such insect decline, a change in grassland management (reduced mown frequency, i.e., extensive management) was recommended to possibly halt insect decline [15,16,17]. Even though several studies have concentrated on how grassland management regimes affect insect biodiversity at the field scale [18,19,20], detailed studies considering three important grassland management regimes (intensive, extensive, and abandoned grassland) are still lacking.

In this study, we contribute to filling this gap by quantifying the impact of grassland management regimes on insect biodiversity in an Austrian alpine region. As important biotic elements of grassland ecosystems, we considered true bugs and syrphids as contrasting ecological groups inhabiting the grasslands. True bugs are a relatively sedentary insect group with a low level of food specialization [21], while syrphids are more mobile and display higher habitat specialization [22]. We selected three grassland types representing different management regimes including intensive grassland (three or more cuts per year), extensive grassland (one or two cuts per year), and abandoned grassland (not mown for more than 10 to 20 years), all of which are certified organic grassland.

In this study, we address the following questions: (a) Do true bug and syrphid abundance and species richness in abandoned grassland differ from low and high-intensity farming practices? (b) How do true bugs and syrphid species coexist among studied grassland management regimes? We hypothesized that (1) extensive grassland promotes a higher abundance of true bugs and syrphids compared to abandoned grassland; (2) there are more unique true bugs and syrphid species in abandoned grassland compared to extensive and intensive grassland because abandoned grassland is undisturbed, thus providing stable conditions, and further has woody vegetation offering additional structural complexity [10,11]; and (3) true bugs species assemblages in abandoned grassland are different from those in intensive and extensive grassland. Our findings might help policymakers to understand the influence of the present organic grassland structural shift from low- to high-intensity farming practices on insect biodiversity, as well as evaluate the impacts of abandonment in alpine agricultural systems.

2. Materials and Methods

2.1. Study Area and Site Characteristics

The study area is between 47°97′ N and 48°14′ N in the province of Lower Austria covering a total area of 19,186 km². The yearly temperature is 10.36 °C (50.65 F) and it is 1.54% higher than Austria’s average temperature. The study area has 128.92 rainy days (35.32% of the time) and typically receives about 51.55 mm of precipitation annually. The study area is prevalently alpine with an average height from 633 m to 1345 m, and it is divided into several valleys with a dense hydrographical network [23]. Around 60% of the land is used for livestock farming, which dominates the agriculture sector.

The field experiment was carried out from May to August 2021 and 2022. Two distinct certified organic grassland types (intensive and extensive) were selected, which were close to each other and nearby abandoned grassland. In total, we surveyed 10 replicates of each extensive, intensive, and abandoned grassland (n = 30) located on south-facing slopes. Extensive grasslands did not receive inorganic fertilizers and were mown once or twice (two late mowings) per year. Intensive grasslands were mown a maximum four times per year. Both grassland types had similar topography and long-term history (>20 years) of grassland management and were used for hay production. At least 10–20 years have passed since management of the abandoned grasslands was discontinued. Extensive and intensive grasslands ranged in size from 550 to 2150 m², whereas those that were abandoned ranged in size from 300 to 2300 m². Grasses and legumes (i.e., Bromus erectus, Lotus corniculatus) dominated plant communities of the extensive grasslands, whereas few species of legumes and grasses (i.e., Lolium perenne, Trifolium repens) were dominant in intensive grasslands, and abandoned grasslands were dominated by Ajuga reptans and Cruciata laevipes. The soils of the study areas were all loamy humus Cambisols “http://www.unserboden.at/ (accessed on 3 February 2023).

2.2. Syrphids (Syrphidae) Sampling

Syrphids (Diptera: Syrphidae) sampling was achieved using two different methods: observation plot survey and sweep netting [24]. In observation plot surveys, we established five 4 m² plots each at a distance of 10 m from a predetermined starting point. Each plot was observed for 10 min, and fast moving syrphid individuals were caught with an entomological hand net. Syrphids were also sampled with standardized sweep-netting. A total of 90 sweeps (three transects of 30 sweeps each, separated by 20 m) were performed from May to August 2021 and 2022 in each study site. All study sites were sampled in a randomized sequence to reduce sampling bias and considered most syrphid activity times between 9 am and 6 pm. Since the weather is known to have a significant impact on syrphid activity, syrphid sampling was only conducted while it was sunny weather [25]. No attempts were made to observe syrphid larvae. Syrphids within observation surveys and sweep netting were either identified on the study site or caught and preserved in 70% ethanol for later identification in the laboratory [26,27].

2.3. True Bugs (Hemiptera) Sampling

Similar to syrphid, we applied a sweep netting method with a 40 cm diameter sweep net to sample true bugs. In the middle of each study site, we applied three parallel 30 m long sweep net samplings each separated by 10 m distance. The sweep net was emptied after a few sweeps to avoid damage to the specimens. The sampling was repeated four times (May to August) in both 2021 and 2022 and completed under favorable weather i.e., sunny and dry days with temperatures ≥25 °C and little wind. The individuals captured were preserved in 70% ethanol and identified to species in the laboratory (adult true bugs were identified to species level while larvae were indeterminable and thus only counted). We identified true bugs using a variety of entomological handbooks of Wagner [28,29,30] and Strauss [31].

2.4. Vegetation Parameters

We estimated overall flower cover and plant height as a predictor of studied insect diversity. We recorded plant height (height from the ground to the top of the plant) and flower cover (abundance of insect-pollinated flowers) in five randomly selected plots per study site. The distance between each plot was 10 m. We placed a wooden frame of 1 × 1 m size, divided into 25 squares, on the ground to estimate overall flower cover [32]. The number of squares containing at least one nectar-producing flower was counted. Similar to the flower cover, plant height was measured five times per study site using a measuring tape, vertically inserted up to the soil level.

2.5. Statistical Analysis

We applied generalized linear mixed models (GLMMs) from the R package lme4 to evaluate abundance and species richness of true bugs and syrphids in the studied grassland management regimes [33]. We considered only true bugs adults for species richness because nymphs of many species cannot be identified up to species level, while adults and nymphs both were used in abundance analysis. We pooled all plots (five observation plots) and sweep nettings (three sweep nettings) per study site and added replications as a group level variable. This tells the lmer to fit GLMM models with a varying-intercept group effect using the variable replication. We used the Poisson family in GLMMs models since the response variables in our dataset (abundance and species richness) were count data. In all models we applied the dispersiontest function from the package AER [34] and the dispersion_glmer from the blmeco package [35] to counter overdispersion. To correct for overdispersion in a model, an extra observation-level random effect was added [36]. If a significant difference between studied grassland management regimes was identified, the glht function from package multcomp was used to perform Tukey’s post-hoc pairwise comparisons [37]. The multicollinearity in regression analysis among predictor variables (flower cover and plant height) was checked by using variance inflation factor (VIF) from package car that measures the correlation and strength of correlation between the predictor variables [38].

We generated Venn diagrams to understand true bugs and syrphids species overlapping attributes between studied grassland management regimes. Further, the species assemblage patterns were assessed by using Non-metric Multidimensional Scaling (NMDS) [39] after Hellinger transformation [40]. We used adonis from package vegan to calculate PERMANOVA (Bray-Curtis dissimilarities, 999 permutations) as performed in Hussain [32]. PERMANOVA assumption “homogeneity of variances” was fulfilled by using the function betadisper. Afterward, a multilevel pairwise comparison based on adonis, i.e., pairwise.adonis, was used to generate test results of species assemblages. The R software version 3.5.1 [41] was used for analysis, and model results were reported using criteria following Zuur and Ieno [42].

3. Results

In total, we found 6641 individuals (5119 adults and 1522 nymphs) and 98 species of true bugs, and 2315 individuals and 95 species of syrphids. Trigonotylus caelestialium (Kirkaldy, 1902), Leptopterna dolabrata (Linnaeus, 1758), Stenotus binotatus (Fabricius, 1794) and Megaloceroea recticornis (Geoffroy, 1785) comprised 51.7% of true bugs abundance, and Melanostoma mellinum (Linnaeus, 1758), Sphaerophoria scripta (Linnaeus, 1758) and Myathropa florea (Linnaeus, 1758) comprised 43% of syrphids abundance. Total abundance and species richness of true bugs and syrphids for each grassland management regime are shown in Appendix Table A1.

Extensive grasslands exhibited significantly higher true bug and syrphid abundance compared to abandoned grasslands. However, true bugs and syrphid species richness had a varied response among grassland management regimes, i.e., extensive grasslands had higher true bugs species richness compared to intensive grassland but was similar to abandoned grassland, and syrphids species richness was similar among the grassland management regimes studied (Table 1, Figure 1). True bugs and syrphids abundance increased with plant height, while flower cover caused a significant increase in syrphid abundance and species richness only. True bugs species richness was unaffected by plant height and flower cover. A VIF value of about 1 indicates that there is no correlation between a studied predictor variable, and that the coefficient estimates and p-values in the regression output are likely reliable (Appendix Table A2).

Table 1. Generalized linear mixed effects models (GLMM) and Tukey’s post-hoc pairwise tests for differences in true bugs and syrphids abundance and richness amongst grassland management regimes (abandoned as intercept, extensive, and intensive) are summarized. The estimate of the corresponding variable in the final model (estimate), z-value (Z), standard errors (Std. error), and significant effects at p < 0.05 are marked in bold. CI = 95% confidence intervals; R²m = marginal R²; R²c = conditional R²; dispersion: 0.75 to 1.4.

True Bugs			Estimate	Std. Error	Z	p	95% CI
abundance	GLMM	Abandoned (Intercept)	1.5759	0.3565	4.421	<0.001	2.40–9.72
		extensive	0.4121	0.1198	3.439	0.001	1.19–1.91
		intensive	−0.0736	0.1245	−0.591	0.554	0.73–1.19
		Dispersion	0.9635	R2m	0.027	R2c	0.301
	Tukey’s post-hoc pairwise comparisons	extensive x abandoned	0.4121	0.1198	3.439	<0.001
		intensive x abandoned	−0.0736	0.1245	−0.591	0.82474
		intensive x extensive	−0.4857	0.1211	−4.01	<0.001
richness
	GLMM	Abandoned (Intercept)	0.8846	0.1804	4.902	<0.001	1.70–3.45
		extensive	0.1380	0.0630	2.191	0.280	1.01–1.30
		intensive	−0.0102	0.0667	−0.153	0.878	0.87–1.13
		Dispersion	0.9959	R2m	0.01	R2c	0.273
	Tukey’s post-hoc pairwise comparisons	extensive x abandoned	0.1380	0.0630	2.191	0.072
		intensive x abandoned	−0.0102	0.0667	−0.153	0.987
		intensive x extensive	−0.1482	0.0635	−2.332	0.05
Syrphids
abundance	GLMM	Abandoned (Intercept)	0.7567	0.0515	14.679	<0.001	1.93–2.36
		extensive	0.1501	0.0645	2.327	0.020	1.02–1.32
		intensive	0.1101	0.0653	1.685	0.092	0.98–1.27
		Dispersion	0.9038	R2m	0.011	R2c	0.027
	Tukey’s post-hoc pairwise comparisons	extensive x abandoned	0.1501	0.0645	2.327	0.05
		intensive x abandoned	0.1101	0.0653	1.685	0.2103
		intensive x extensive	−0.0400	0.0580	−0.69	0.7689
richness
	GLMM	Abandoned (Intercept)	0.5641	0.0515	10.953	<0.001	1.59–1.94
		extensive	0.1069	0.0622	1.719	0.0857	0.99–1.26
		intensive	0.1073	0.0628	1.709	0.0874	0.98–1.26
		Dispersion	0.7762	R2m	0.005	R2c	0.007
	Tukey’s post-hoc pairwise comparisons	extensive x abandoned	0.1069	0.0622	1.719	0.198
		intensive x abandoned	0.1073	0.0628	1.709	0.201
		intensive x extensive	0.0004	0.0555	0.008	1

Figure 1. Effects of studied grassland management regimes on the abundance and species richness of true bugs and syrphids. Box–Whisker plots display the medians (—), mean (●), notches, 25% and 75% percentiles, and outlier values (•). Box–Whisker plots with distinct letters differ significantly from one another (p < 0.05) using Tukey’s post-hoc pairwise tests.

The variation in species between the studied grassland management regimes was characterized using Venn diagrams (Figure 2). Overall, 30 true bugs and 28 syrphids were shared between grassland management regimes. Further, 8 true bugs and 11 syrphids species occurred exclusively in intensive grassland, 11 true bugs and 16 syrphid species occurred exclusively in extensive grassland, and 25 true bugs and 20 syrphid species occurred exclusively in abandoned grassland. Large numbers of species (25.5% true bugs and 21.5% syrphids) were only observed in abandoned grassland (Figure 2). True bugs assemblages in abandoned grassland differed significantly from those in intensive and extensive grassland (Table 2; Figure 3). Syrphids assemblages appeared to be similar in all three studied grassland management regimes. As syrphids assemblage data did not fulfil homogeneity of variance assumption, we did not calculate PERMANOVA.

Figure 2. Venn diagrams showing the main overlaps and differences in true bugs and syrphids species among grassland management regimes. Numbers and percentages are calculated using the total data set.

Table 2. PERMANOVA (Pairwise-adonis) for true bugs in grassland management regimes i.e., abandoned, extensive, and intensive grassland. Significant p-values (<0.05) are shown in bold.

True Bugs		Df	Sum Sq	R2	F	p
Adonis	Management regimes	2	1.4242	0.20617	3.5061	<0.001
	Residual	27	5.4839	0.79383
	Total	29	6.9082	1
Pairwise Adonis		F. Model	R2	p	Adjusted p
	abandoned × extensive	3.1035	0.1470	0.001	0.003
	abandoned × intensive	5.4626	0.2328	0.001	0.003
	extensive × intensive	1.8765	0.0944	0.059	0.177

Figure 3. Non-metric Multidimensional Scaling (NMDS) depending on species assemblages of true bugs and syrphids in grassland management regimes.

4. Discussion

It is well established that extensive grassland makes a considerable contribution to the insect biodiversity in agro-ecosystems [43]. There is a relatively large agreement that intense grassland management practices like frequent cutting or heavy grazing have a negative impact on insect biodiversity [13]. Of the three grassland management regimes, extensive grassland had a significant effect on true bug and syrphid abundance when compared to abandoned grassland. Surprisingly, abandoned and extensive grassland had a similar true bug and syrphid species richness. It is known that farming methods have a significant impact on true bugs [44]. As previously stated by Moir [45], it appears that the abundance of true bugs is a more sensitive indicator of the stability of grassy ecosystems than species richness. Woody vegetation, like in the abandoned grassland, is also known to influence insect assemblages [46]. Even in the alpine region, where semi-natural grasslands still support a rich invertebrate fauna [47], abandoned grassland probably serves as a secondary habitat for true bugs. Syrphids are reported to require pollen and nectar as a food source [48], and a variety of floral resources in extensive grassland might have a favorable impact on the abundance of syrphids.

Certain management practices, like mowing, directly affect insects by harming or eliminating individuals or by altering the habitat [49]. This is substantiated in our results due to the presence of more unique true bugs and syrphids species in abandoned grassland, which proves that the abandonment of organic grassland is not a threat to alpine insect diversity. According to published research, grassland management changes the vegetation’s structure and plant species composition [50], and it could have an impact on the microclimate and other elements of the microhabitat [51]. Due to the high sensitivity of many insects to microclimatic conditions, such indirect effects may change true bugs species compositions for two main reasons. Firstly, true bugs are known to have close relationships with plant species and achieve their highest richness in extensive grasslands [15]. Rapid changes in the vegetation structure of the plant community due to management in intensive and extensive grassland seems to have direct effects on true bugs at a species level and might force true bugs towards more stable abandoned grassland. Secondly, true bugs have both larval and adult stages that coexist in the same habitat and are sensitive to environmental changes [52]. There are several behaviors similar for observed syrphid species (i.e., some species from genus Pipizella or Cheilosia) that were only seen in abandoned grasslands, i.e., they prefer dense vegetation and fly close to the ground cover and bushes [53]. These species may require the period of transitional abandonment, which serves as a suitable habitat that offers resources for survival, reproduction, and mobility for true bugs and syrphids at species level.

The results presented here demonstrated that management effects are important for the species assemblages of true bugs and syrphids. Non-metric Multidimensional Scaling (NMDS) indicates that abandoned grassland has different true bugs assemblages compared to intensive and extensive grassland. It is probable that true bugs used stable and undisturbed abandoned grassland as a refuge because intensive and extensive grasslands are generally more disturbed by agricultural activities [54]. Contrary to true bugs, syrphids are mobile and spread across various grassland management regimes.

Interestingly, true bugs and syrphids increased with plant height. It is probable that many abundant true bugs were zoophagous or polyphagous herbivores and thus more responsive to the greater structural complexity of vegetation [15]. This may be because undisturbed and stable vegetation may provide more resources for overwintering, feeding, oviposition, and resting, thereby providing a larger potential area for colonization [55]. However, syrphids are known to depend on pollen and nectar for nutrition [48], and a greater abundance of floral resources within extensive grassland was associated with a higher syrphid abundance [56]. Future studies should also consider other predictor variables, i.e., duration of the abandonment and the structure of the surrounding landscape, for a more elaborative explanation of insect response to grassland abandonment.

Our results demonstrate the relevance of grassland management regimes for insect biodiversity in alpine landscapes where a large proportion of conventionally maintained grassland is present. This appears to be important considering the current high land use pressure [1], which has resulted in a large proportion of agricultural landscapes being abandoned especially in the alpine region. In support of hypothesis 1, extensive grassland promotes a high abundance of true bugs and syrphids compared to abandoned grassland. Abandoned grassland seems to be vital for true bugs and syrphids biodiversity in providing habitat for more unique species compared to intensive and extensive grassland, thus confirming hypothesis 2. In addition, true bugs assemblages were found to be different in abandoned grassland, thus being partly in line with hypothesis 3, indicating that other factors such as habitat disturbance [15] might drive variation in assemblages apart from grassland management.

5. Conclusions

It is challenging to explain the divergent effects of a complex and spatially diverse process of grassland management regimes such as land abandonment. In the literature, grassland abandonment is frequently viewed as a barrier to biodiversity conservation, an impression generally widespread among plant ecologists [57], and extensive grassland considered as a biodiversity rich habitat [5,58]. We found that extensive grassland can manage to increase true bugs and syrphid abundance. However, extensive and intensive grasslands are more disturbed by agricultural activities, and therefore drive more true bugs and syrphids species towards natural, undisturbed ecosystems of abandoned grassland. Our findings indicate that there are no “one-size-fits-all” policy solutions for managing biodiversity conservation upon land abandonment in alpine grassland. A mosaic landscape consisting of abandoned grassland along with grassland having different, mainly extensive, management intensity could be an ideal arrangement for alpine biodiversity conservation.