Article | Open | Published:

# Large manipulative experiments revealed variations of insect abundance and trophic levels in response to the cumulative effects of sheep grazing

## Abstract

Livestock grazing can affect insects by altering habitat quality; however, the effects of grazing years and intensities on insect abundance and trophic level during manipulative sheep grazing are not well understood. Therefore, we investigated these effects in a large manipulative experiment from 2014 to 2016 in the eastern Eurasian steppe, China. Insect abundance decreased as sheep grazing intensities increased, with a significant cumulative effect occurring during grazing years. The largest families, Acrididae and Cicadellidae, were susceptible to sheep grazing, but Formicidae was tolerant. Trophic primary and secondary consumer insects were negatively impacted by increased grazing intensities, while secondary consumers were limited by the decreased primary consumers. Poor vegetation conditions caused by heavy sheep grazing were detrimental to the existence of Acrididae, Cicadellidae, primary and secondary consumer insects, but were beneficial to Formicidae. This study revealed variations in insect abundance and trophic level in response to continuous sheep grazing in steppe grasslands. Overall, our results indicate that continuous years of heavy- and over- sheep grazing should be eliminated. Moreover, our findings highlight the importance of more flexible sheep grazing management and will be useful for developing guidelines to optimize livestock production while maintaining species diversity and ecosystem health.

## Introduction

Excessive grazing on grassland ecosystems by livestock poses a serious threat to the grasslands by lowering productivity, biodiversity and stability1, leading to ecological deterioration2, especially for insect diversity3. Insects are a major, but often under-appreciated component of terrestrial ecosystems4,5,6, and the effects of grazing on insect diversity have been thoroughly investigated in previous studies. Some studies have shown that grazing may increase insect diversity7,8,9, while others showed decreased insect diversity10, 11 or no change in insect diversity12, 13 in response to grazing. These inconsistent results may be due to factors such as variation in vegetation6, intensity of grazing14, 15, and herbivore size6. However, growing evidence shows that insects are experiencing local/regional species loss or even global extinction16, and that the diversity of insects apparently declines more rapidly than that of vertebrates and plants6, 17. Therefore, understanding critical factors that determine their diversity and species composition has become an urgent task facing ecologists and conservation biologists.

The effects of large herbivores on insect abundance are grazer species-specific and pre-grazing plant diversity-dependent9, 18. Large herbivore species can alter vegetation features due to diet selection and body size, potentially influencing the insect community13, 19. Low grazing intensities result in taller swards, providing more forage and shelter for herbivorous insects20. Grazing profoundly changes insect taxonomic composition6. Although extensive research investigating the effects of livestock grazing on grassland insect abundance has provided valuable insights21, 22, several important gaps in our knowledge remain. Previous studies investigating the effects of grazing intensities on insect diversity and abundance often only compared two or three stocking rates13, 23, 24, and continuous grazing years have rarely been considered. Indeed, some studies only provided one year of data, or conducted short term investigation of grazing at few levels of grazing intensity25, 26 that lacked continuous treatments6, 19. Moreover, other studies only focused on limited groups, such as pollinators24. The use of taxonomic hierarchies such as order and species levels could be advantageous to biodiversity assessments27. Thus, there is a pressing need for studies that examine how the taxonomic compositions of insect communities respond to livestock grazing. Furthermore, the effects of continuous manipulative grazing on other insect groups, such as Formicidae, Acrididae, Cicadellidae, and on various trophic levels, are not well understood. Such knowledge will facilitate reasonable grazing management and maintenance of species diversity and ecosystem health.

Grasslands comprise the largest terrestrial ecosystem in China, and play a critical role in maintaining the structure, function and stabilization of surrounding natural ecosystems28, 29. Sheep grazing is a key management tool in the steppe grassland of Northern China and the most important economic income source of herdsman30. However, rapid steppe degradation has led to reduced biodiversity, decreased productivity and, in some cases, desertification owing to livestock over-grazing31, 32. The effects of continuous sheep grazing on insect abundance and trophic level in such areas are unknown; therefore, we conducted a 3-year large manipulative experiment with five grazing levels (0, 4, 8, 12 and 16 sheep per 1.33 ha) in Inner Mongolia, China. The specific goals of this study were to determine (i) how dominant insect groups respond to various intensities and years of sheep grazing; (ii) how insect trophic levels change with sheep grazing; and (iii) potential reasons for insect abundance variation for sheep grazing. Suggestions regarding how to manage sheep grazing to maintain insect diversity are also discussed.

## Results

### Insect Shannon-Wiener index, species richness, and abundance

Grazing intensities and grazing years were found to have significant effects on the insect Shannon-Wiener index, species richness, and abundance (all P < 0.05, Table 1). When compared with the control, moderate grazing (MG), heavy grazing (HG) and over grazing (OG) led to a significant (P < 0.05) reduction in the insect Shannon-Wiener index in 2016 (Fig. 1a). Moreover, insect species richness decreased significantly (P < 0.05) in 2015 and 2016 in response to grazing treatments MG, HG and OG (Fig. 1b). Furthermore, insect abundance was significantly lower (P < 0.05) in response to grazing treatments HG and OG in all three experimental years (Fig. 1c). These results indicate that grazing intensity can significantly influence insect variation.

There were significant inter-annual changes in three insect variables (Fig. 1). When compared with 2014, continuous sheep grazing in 2015 and 2016 led to a significant (P < 0.05) reduction in the insect Shannon-Wiener index, species richness, and abundance, especially for over grazing (OG). These results showed that sheep grazing had a significant negative cumulative effect on insects.

### Formicidae, Acrididae, and Cicadellidae abundance

Grazing intensities and years had significant effects on the largest families, Formicidae, Acrididae, and Cicadellidae (all P < 0.05, Table 1), as well as a significant interactive effect on Formicidae (P < 0.05). Acrididae abundance exhibited a significant negative linear relationship to grazing intensities in 2016 (y = −0.642x + 14.867, R2 = 0.33, P = 0.01425). Cicadellidae abundance also showed significant negative linear relationships to grazing intensities in all three experimental years (2014: y = −0.642x + 14.867, R2 = 0.33, P = 0.01425; 2015: y = −2.9x + 78, R2 = 0.312, P = 0.0178; 2016: y = −2.25x + 43.8, R2 = 0.33, P = 0.0004). In contrast, there were no significant changes between Formicidae abundance and grazing intensities in any of the experimental years. These results suggest that the dominant insect groups responded differently to altered sheep grazing intensities.

There were also significant inter-annual changes in these three dominant insect groups (Fig. 2). When compared with 2014, continuous heavy grazing (HG) and over grazing (OG) by sheep in 2015 and 2016 led to a significant (P < 0.05) reduction in the abundance of Formicidae, Acrididae, and Cicadellidae, which suggested a significant negative cumulative effect on grazing years.

### Primary and secondary consumers

Grazing intensity and years all significantly influenced primary consumers (phytophagous insects) and secondary consumers (parasitoids and carnivorous insects) (all P < 0.05, Table 1), and these factors exerted a significant interaction effect on secondary consumers (P < 0.05). Additionally, both primary and secondary consumers showed a significant negative linear relationship with grazing intensity (primary consumer, 2014: y = 246.933 − 6.858x, R2 = 0.262, P = 0.030, 2015: y = 147.877 − 6.767x, R2 = 0.309, P = 0.018, 2016: y = 91.8 − 4.667x, R2 = 0.600, P = 0.0004; secondary consumer, 2015: y = 22.067 − 1.05x, R2 = 0.415, P = 0.0057, 2016: y = 15.333 − 0.667x, R2 = 0.268, P = 0.028). These results indicate that both primary and secondary consumers were also susceptible to altered sheep grazing intensities.

There were also significant inter-annual changes (Fig. 3). Compared with 2014, the continuous sheep grazing in 2015 and 2016 led to a significant (P < 0.05) reduction on these two insect trophic levels, especially for over grazing (OG). These results showed that there also had a significant negative cumulative effect of sheep grazing years on primary consumer and secondary consumer.

The relationship between primary and secondary consumer abundance were examined. The result showed that the secondary consumer abundance significantly (P < 0.05) increased with increasing primary consumer abundance in all grazing treatments (Fig. 4). These findings suggest that the secondary consumer insects were limited by decreased primary consumer insects following increased sheep grazing intensities.

### Grazing-vegetation-insect relationship

The Importance Value showed that the plant Leymus chinensis (Trin.) Tzvel. is the most widely distributed across each grazing plot, followed by Stipa grandis P. Smirn., Cleistogenes squarrosa (Trin.) Keng., Chenopodium glaucum L. and Carex korshinskyi Kom. (Table S1). There were no significant differences in plant diversity and species richness among grazing intensities in 2015 (Fig. S1). The relationship between vegetation structure heterogeneity and grazing intensities was an approximate U-curve (Fig. S1) with a tendency to first ascend, then descend.

The grazing-vegetation-insect relationships in 2015 were next examined by redundancy analysis (RDA) (Fig. 5). Axis 1 explained most of the variation in insect abundance (37.3%). The results showed that the grasslands subjected to heavy grazing (HG) and over grazing (OG) were characterized by decreased plant coverage, biomass, density and height (Fig. 5, Table S2). In contrast, grasslands subjected to no grazing (CK), light grazing (LG), and moderate grazing (MG) showed high plant coverage, biomass, density and height (Fig. 5, Table S2). These four vegetation variables showed positive relationships with Acrididae, Cicadellidae, primary consumer and secondary consumer insect abundance (Fig. 5, Table S3). Moreover, the results showed that the poorer vegetation attributes of the HG and OG groups could be detrimental to the existence of Acrididae, Cicadellidae, primary consumer and secondary consumer insects, but would be beneficial to Formicidae. These results suggest that the changes in vegetation in response to different grazing intensities resulted in variations in the insect community and explained these changes well.

## Discussion

Understanding the relationship between grazing and biodiversity and how it affects insect populations may provide insights that improve monitoring strategies, early warning alerts and management strategies for species conservation22, 33. In this study, the cumulative effects of sheep grazing intensities and years on insect abundance and trophic level were analyzed by a large manipulative experiment over three years. Several broad conclusions can be drawn based on the results. First, sheep grazing intensities significantly influenced the insect community, with increased grazing having detrimental effects on insect abundance, diversity and species richness. These findings confirm those of previous studies linking changes in insect population dynamics to variations in grazing intensity3, 9, 25, 34. Second, sheep grazing years also negatively affected the insect community with a significant cumulative effect. This finding agrees with those of previous studies showing that continuous grazing years were detrimental to insect abundance35, especially under heavy grazing and over grazing. Third, we found the largest families, Acrididae and Cicadellidae, were susceptible to sheep grazing intensity, but Formicidae was not. This result also supports the general opinion that insect groups respond to grazing differently6, and that not all species are sensitive to grazing27. Fourth, grazing can affect insect trophic level. Interestingly, we found that primary and secondary consumers were both negatively impacted by increasing grazing intensity, with the secondary consumer being limited by the primary consumer. These results were contrary to those of previous studies that indicated extensive grazing enhanced phytodiversity36, 37, and in turn promoted species richness of higher trophic levels38. This indicates that there is a cascade reaction to insects in response to grazing. Alterations in one or some insect species usually change other the specific interspecies relationships of other insect species39. Finally, we found that poorer vegetation attributes associated with increased sheep grazing intensity were detrimental to insect existence, and that plant attributes for grazing were closely associated with variations in insect abundance3, 6, 9.

The question of which selective factors for grazing have driven the insect change is of great interest, and has been investigated in many previous studies1, 6, 40, 41. Livestock grazing can affect insect diversity and abundance directly and indirectly. The direct effects include unintentional ingestion or trampling42, while the indirect effects mainly include changes in microclimates, vegetation, and interspecies relationships in response to grazing14, 40. As human disturbances of native grasslands increase, understanding the grazing-plant-insect relationship becomes increasingly important6, 18. Grazing and trampling can modify vegetation, thereby altering the insect community structure43, 44. Large herbivores may also lower the quality of plants by reducing their nitrogen levels22 and change vegetation structure by exposing bare soil; thus, increasing the risk of predators45. Foraging by livestock reduces plant density and coverage, reducing food availability to insects. Herbivorous insects normally need adequate food resources to support their development and reproduction46. When food resources decrease and become scarce, their population dynamics are also be negatively affected by food shortages47. Therefore, there is generally a positive relationship between herbivorous insects and plant biomass, especially during grazing of degraded grasslands of limited resources48, 49. Since insects primarily feed on plants for survival19, 20, plant diversity has enhances insect diversity and abundance21,22,23. Accordingly, decreased plant coverage, height, biomass and plant litter can be presumed to lead to decreased quality of habitat through exhausted food resources, shifts in the nutrient status of host plants and development of inhospitable microclimates17. For example, fruit flies inhabit grasslands50 when plant coverage is high, but areas of exposed bare soil do not attract colonizing adults51. Zhu et al.19, showed that large herbivores strongly affected insect species richness by modifying plant structural heterogeneity, which reversed the positive relationship between plant and insect diversity. Zhong et al.9, found that the positive interactions between large herbivores and grasshoppers were driven by differential herbivore foraging preferences for plant resources that break down the associational plant defense between grasses and forbs. In this study, we found that vegetation structure heterogeneity tended to ascending, then descending in succession in response to increased sheep grazing intensity, which supported the intermediate disturbance hypothesis52. Plant coverage, biomass, density and height were all tightly correlated with insect abundance. Specifically, decreased vegetation variables in response to increased grazing intensity deteriorated insect food resources and refuge, which resulted in decreased insect abundance, especially for phytophagous insects (the primary consumer), such as the dominant groups Acrididae and Cicadellidae. As a result, secondary consumers suffered from the decreased food availability owing to the shortage of primary consumers. Interestingly, we found that different grazing intensities did not influence plant biodiversity and species richness, while insect diversity and abundance decreased with increased grazing intensity in 2015. This finding strongly supports a recent report that showed arthropod diversity is often more negatively affected by grazing than plant diversity40. Furthermore, the microclimate variables in different grazing habitats, such as humidity, light, and temperature, can also influence insect growth performance and population dynamics53. These variables may be more unfavorable to insects following increased grazing intensity. Thus, future research should be focus on differences in micro-climate variables and how they influence insect abundance for sheep grazing.

We also found that yearly fluctuations of grazing would impact on insect abundance more than grazing treatment. One reason is the cumulative effects of continuous sheep grazing, that have been demonstrated by Minckley et al.35 and Marriott et al.36 and in our present study. Another reason may be climatic variation of different grazing years, such as the temperature and rainfall. Stige et al 54. have demonstrated that climatic changes of yearly fluctuation could significantly impact insect population dynamics. In addition, insect migration and short distance dispersal in response to habitat change within years may also impact on insect abundance variation45. Those related researches can well explain grazing years’ effect on insect abundance.

Insects are important indicators of the ecological environment54. Monitoring insect communities to measure the benefits of changes in grazing intensities to improve biodiversity should consider the continuous effects of grazing disturbance. Although Cicadellidae is one of the largest families of phytophagous insects and among the dominant groups of prairie herbivores55, its abundance significantly declined during three years of consecutive grazing, indicating that the insects may not be able to tolerate the habitat conditions created by this grazing regimen. Additionally, grazing has been shown to significantly reduce Acrididae, which are important grassland pests14. Acrididae density in 2016 differed markedly from the first year of grazing, suggesting that continuous grazing may also negatively influence their population dynamics and spatial distribution. The Acrididae and Cicadellidae were susceptible and fragile to sheep grazing. Although most insect groups decreased with decreased habitat quality, the Formicidae group did not change. The mechanism regulating this behavior is yet to be determined; however, we presume that this may have resulted from the availability of suitable food resources such as herbivore feces, plant litter and microclimate, such as adequate sunshine. These differences in insect responses indicate a difference in sensitivity for grazing based on a likely complex mechanism, which will be addressed in future studies. Nevertheless, our results showed that not all insect species are susceptible to grazing by large herbivores6, implying that susceptible insect species may decline or disappear and should therefore be conserved in grazed grasslands. The results of the present study also suggest that insect groups and species respond to herbivore grazing in a complex manner, and it is essential to quantify this effect in further studies.

Livestock grazing is a key management tool in grasslands, and its widespread prevalence has generated great interest in understanding its ecological effects, especially for insects40, 56. Against the background of decreased biodiversity57, 58, appropriate management of remaining grassland sites is required to maintain biodiversity. Grazing appears to have a high potential for combining these targets with the growing social demands for animal welfare59. However, the main function of pastures for farmers is to meet agronomic and financial interests. Therefore, identification of a threshold grazing intensity that fulfils both environmental and livestock production objectives is essential. The insights into the relationship between sheep grazing and insect abundance provide an opportunity to devise strategies to improve livestock and insect management that reduce the potential conflict of livestock production and insect conservation. It is well known that high stocking rates combined with intensive grassland management contribute to the deterioration of insect diversity37. However, it is still not clear what level of grazing intensity is appropriate to conserve insects and by which mechanisms sheep grazing intensity affects insect diversity in Chinese Steppe grasslands. We found that the continuous years of heavy and over grazing should be eliminated to prevent a large decrease in insect abundance. This can be accomplished through interventions such as improved livestock grazing management and implementation of fallow periods to allow habitat recovery. Moreover, grazing intensity and years should be considered to ensure good livestock management. Finally, although we studied the vegetation and insect responses to sheep grazing for three years, the observed responses to sheep grazing may only have just been beginning. Overall, a better understanding of sheep grazing ecosystems requires long-term monitoring, particularly in relation to improving grassland habitats to ensure sustainability, plant and insect biodiversity and the livelihood of farmers.

## Materials and Methods

### Study site

The study site is located in the eastern region of the Eurasian Steppe Zone, Inner Mongolia, China (116°32′E, 44°15′N) and managed by the institute of Grassland Research, Chinese Academy of Agricultural Sciences. The site has a semi-arid continental climate with a mean annual temperature of −0.1 °C, with the coldest temperatures occurring in January (average −22.0 °C, extreme minimum −41 °C) and the highest in July (average 18.3 °C, extreme maximum 38.5 °C). The annual accumulated temperature ranges from 2100 °C to 2400 °C, and the annual precipitation was 350 to 450 mm. The major soil type in the area is calcic chestnut soil and the main vegetation type is typical steppe dominated by two perennial grasses, Leymus chinensis (Trin.) Tzvel. and Stipa grandis P. Smirn. Other common grass species include Cleistogenes squarrosa (Trin.) Keng., Artemisia frigida Wild., Salsola collina Pall., and Chenopodium glaucum L. The period of maximum vegetation biomass and highest insect occurrence is from early July to early August. The experimental site was not grazed, but was mowed from 2007 to 2013.

### Experimental design and animal management

A relatively flat area within the study site with homogenous conditions of soil and plants was blocked in 2014. We constructed 15 sets of 1.33-ha enclosures with a 100 m × 125 m dimension and a 1.5 m-high iron netting above the ground to confine sheep within the enclosures. Fifteen enclosures were randomly assigned into five treatments with three replicates per treatment: control (CK), light grazing (LG), moderate grazing (MG), heavy grazing (HG) and over grazing (OG). Based on the local standard set60 for sheep grazing intensity, 0 (grazing pressure: 0 SSU·d·(hm2)−1·y−1), 4 (170 SSU·d·(hm2)−1·y−1), 8 (340 SSU·d·(hm2)−1·y−1), 12 (510SSU·d·(hm2)−1·y−1) and 16 (680 SSU·d·(hm2)−1·y−1) sheep per 1.33 ha were used for the CK, LG, MG, HG and OG groups, respectively. The grazing season lasted from June 10 to September 10 each year.

### Insect survey

Since the general records for the study area indicate that the highest insect diversity occurs in July, insect samples were collected on 2014 (July 10), 2015 (July 10) and 2016 (July 11), respectively. Insects were collected using a suction sampler as previously described61. To estimate the abundance of each insect species per square meter, insects were sucked from a steel framed 1 m × 1 m × 1 m quadrat randomly placed in five locations within each plot. Each surface of the quadrat frame, excluding the undersurface, was covered with fine 1 mm2 cloth mesh to ensure that no insects escaped. The covered upper mesh could be opened when using suction sampler. To avoid man-made disturbances and collect as many insects as possible, the quadrat frame was thrown into the air and allowed to freely fall to the ground. We then used the suction sampler to collect all insects within the quadrat, which took 2 minutes for each suction sample. These five sample locations for each plot had a relatively flat terrain with uniform vegetation and were at least 10 m away from the plot boundary to minimize edge effects. Insect specimens were only collected under favorable conditions (sunny days with minimal cloud cover and calm or no wind) from 09:00 to 15:00 h. All grazed plots were sampled in a random order. All contents of the suction sampler were preserved in ziplock bags marked with respective plot numbers and taken to the laboratory. We then killed all insects by freezing them (−20 °C for 20 hours).

The insects were examined and sorted into 10 orders; Hemiptera, Orthoptera, Hymenoptera, Coleoptera, Diptera, Lepidoptera, Neuroptera, Collembola, Thysanoptera and Mantodea. All individuals were identified to species, but immature insects omitted from analysis. Insects that could not be identified as a species were separated into recognizable taxonomic units based on morphological characteristics.

### Vegetation survey

Vegetation was sampled in parallel with insect surveys in 2015. The following vegetation attributes were evaluated in five randomly selected quadrats (1 m × 1 m) within each plot using the same methods by Zhu et al.24: plant coverage, height, density, biomass. Plant samples were not taken from about 10 m from the plot boundary to avoid any edge effects. The tufted plant (S. grandis) were counted by the number of tufts. The Importance Value (I.V.) of each plant species/treatment was calculated by the formula62: I.V. = relative cover + relative density + relative frequency (Table S1).

### Statistical analyses

Insect and plant Shannon-Weiner index63 was calculated as

$$H=-{\sum }_{1}^{{\rm{S}}}(Pi\times \,\mathrm{ln}\,Pi)$$

where Pi is the proportion of individuals represented by species i, S is the number of species and H is the Shannon-Weiner diversity index. Insect species richness and abundance within each plot from 2014 to 2016 were also analyzed. Vegetation structural heterogeneity was estimated as the coefficient of variation (CV) of plant height in each quadrat19.

Normality of these measured variables of plants and insects was assessed using SAS64, and insect species and abundance were log-transformed prior to analysis. We used repeated-measures ANOVA to evaluate the effects of grazing intensity and grazing years on the insect Shannon-Wiener index, insect species and insect abundance, with grazing intensity treatment as a between-subject factor (main effect) and years as a within-subject factor (repeated), considering plots as experimental units. Furthermore, one-way ANOVA with LSD post hoc tests was used to compare insect variables within each year or each intensity using SAS version 8.0. Relationships between grazing intensity and abundance of the dominant insect groups (Acrididae, Cicadellidae and Formicidae) and both trophic primary and secondary consumer insect abundance were analyzed using the General Linear Model (GLM), respectively.

Multivariate ordination redundancy analysis (RDA) allows simultaneous representation of observations, Y variables, and X variables in two or three dimensions, which is optimal for the covariance criterion65, 66. We evaluated the relationships between grazing intensity-vegetation attribute-insect abundance by RDA using treatment plots as observations, vegetation parameters (biomass, coverage, density, height) as environmental variables and abundance of insect groups as species variables.

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## References

1. 1.

Gossner, M. M., Weisser, W. W. & Meyer, S. T. Invertebrate herbivory decreases along a gradient of increasing land-use intensity in German grasslands. Basic. Ap. Ec. 15, 347–352 (2014).

2. 2.

Zhao, T. et al. Ecosystem services and their valuation of china grassland. Acta. Ecol. Sin. 24, 1101–1110 (2004).

3. 3.

Moran, M. D. Bison grazing increases arthropod abundance and diversity in a tallgrass prairie. Env. Entomol. 43, 1174–1184 (2014).

4. 4.

Belovsky, G. E. & Slade, J. B. Insect herbivory accelerates nutrient cycling and increases plant production. Proc. Natl. Acad. Sci. 97, 14412–14417 (2001).

5. 5.

Whiles, M. R. & Charlton, R. E. The ecological significance of tallgrass prairie arthropods. Ann. R. Entom. 51, 387–412 (2006).

6. 6.

Zhu, H., Wang, D. L., Guo, Q. F., Liu, J. & Wang, L. Interactive effects of large herbivores and plant diversity on insect abundance in a meadow steppe in China. Agr. Eco. Env. 212, 245–252 (2015).

7. 7.

Bakker, E. S., Ritchie, M. E., Olff, H., Milchunas, D. G. & Knops, J. M. H. Herbivore impact on grassland plant diversity depends on habitat productivity and herbivore size. Ecology Letters. 9, 780–788 (2006).

8. 8.

Freymann, B. P., Buitenwerf, R., Desouza, O. & Olff, H. The importance of termites (Isoptera) for the recycling of herbivore dung in tropical ecosystems: a review. European Journal of Entomology. 105, 165–173 (2008).

9. 9.

Zhong, Z. et al. Positive interactions between large herbivores and grasshoppers, and their consequences for grassland plant diversity. Ecology. 95, 1055–1064 (2014).

10. 10.

Wardle, D. A., Barker, G. M., Yeates, G. W., Bonner, K. I. & Ghani, A. Introduced browsing mammals in New Zealand natural forests: aboveground and belowground consequences. Ecol. Monogr. 71, 587–614 (2001).

11. 11.

Howe, H. F., Brown, J. S. & Zorn-Arnold, B. A rodent plague on prairie diversity. Ecology Letters. 5, 30–36 (2002).

12. 12.

Adler, P. B., Milchunas, D. G., Sala, O. E., Burke, I. C. & Lauenroth, W. K. Plant Traits and Ecosystem Grazing Effects: Comparison of US Sagebrush Steppe and Patagonian Steppe. Ecol. Appl. 15, 774–792 (2005).

13. 13.

Kruess, A. & Tscharntke, T. Grazing intensity and the diversity of grasshoppers, butterflies, and trap-nesting bees and wasps. Conser. Biol. 16, 1570–1580 (2002).

14. 14.

O’Neill, K. M. et al. Effects of livestock grazing on rangeland grasshopper (Orthoptera: Acrididae) abundance. Agr. Eco. Env. 97, 51–64 (2003).

15. 15.

O’Neill, K. M. et al. Effects of Livestock Grazing on Grasshopper Abundance on a Native Rangeland in Montana. Agr. Eco. Env. 39, 775–786 (2010).

16. 16.

Collinge, S. K. Effects of grassland fragmentation on insect species loss, colonization, and movement patterns. Ecology. 81, 2211–2226 (2000).

17. 17.

Thomas, J. A. et al. Comparative losses of British butterflies, birds, and plants and the global extinction crisis. Science. 303, 1879–1881 (2004).

18. 18.

Liu, J. et al. Impacts of grazing by different large herbivores in grassland depend on plant species diversity. J. App. Ecol. 52, 1053–1062 (2015).

19. 19.

Zhu, H. et al. The effects of large herbivore grazing on meadow steppe plant and insect diversity. J. Appl. Ecol. 49, 1075–1083 (2012).

20. 20.

Pétillon, J. et al. Influence of abiotic factors on spider and ground beetle communities in different salt-marsh systems. Basic. Ap. Ec. 9, 743–751 (2008).

21. 21.

Littlewood, N. A. Grazing impacts on moth diversity and abundance on a Scottish upland estate. Insect Conservation & Diversity. 1, 151–160 (2008).

22. 22.

Cease, A. J. et al. Heavy Livestock Grazing Promotes Locust Outbreaks by Lowering Plant Nitrogen Content. Science. 335, 467–469 (2012).

23. 23.

Wallisdevries, M. F., Van Swaay, C. A. M. & Plate, C. L. Changes in nectar supply: A possible cause of widespread butterfly decline. Current Zoology. 58, 384–391 (2012).

24. 24.

Sjödin, N. E., Bengtsson, J. & Ekbom, B. The influence of grazing intensity and landscape composition on the diversity and abundance of flower-visiting insects. J. Appl. Ecol. 45, 763–772 (2008).

25. 25.

Debano, S. J. Effects of livestock grazing on aboveground insect communities in semi-arid grasslands of southeastern Arizona. Biodivers. C. 15, 2547–2564 (2006).

26. 26.

Verdú, J. R. et al. Grazing promotes dung beetle diversity in the xeric landscape of a Mexican Biosphere Reserve. Biol. Conser. 140, 308–317 (2007).

27. 27.

Williams, P. H. & Gaston, K. J. Measuring more of biodiversity: Can higher-taxon richness predict wholesale species richness? Biol. Conser. 67, 211–217 (1994).

28. 28.

Kang, L., Han, X., Zhang, Z. & Sun, J. O. Grassland ecosystems in China: review of current knowledge and research advancement. Phi. T. Roy. B. 362, 997–1008 (2007).

29. 29.

Sun, H. L. The chinese ecosystem. (Science Press, 2005).

30. 30.

Ren, H. et al. Do sheep grazing patterns affect ecosystem functioning in steppe grassland ecosystems in Inner Mongolia? Agr. Eco. Env. 213, 1–10 (2015).

31. 31.

Chen, Z. Z. & Wang, S. P. T ypical Steppe Ecosystems of China. (Science Press, 2000).

32. 32.

Han, J. G. et al. Rangeland degradation and restoration management in China. Rangeland. J. 30, 233–239 (2008).

33. 33.

Huang, X. et al. Quantitative Analysis of Plant Consumption and Preference by Oedaleus asiaticus (Acrididae: Oedipodinae) in Changed Plant Communities Consisting of Three Grass Species. Env. Entomol. 45, 163–170 (2016).

34. 34.

Rambo, J. L. & Faeth, S. H. Effect of Vertebrate Grazing on Plant and Insect Community Structure. Conser. Biol. 13, 1047–1054 (2001).

35. 35.

Minckley, R. L. Maintenance of richness despite reduced abundance of desert bees (Hymenoptera: Apiformes) to persistent grazing. Insect Conservation & Diversity. 7, 263–273 (2013).

36. 36.

Marriott, C. A., Hood, K., Fisher, J. M. & Pakeman, R. J. Long-term impacts of extensive grazing and abandonment on the species composition, richness, diversity and productivity of agricultural grassland. Agr. Eco. Env. 134, 190–200 (2009).

37. 37.

Jerrentrup, J. S., Wrag‐Mönnig, N., Röver, K., Isselstein, J. & McKenzie, A. Grazing intensity affects insect diversity via sward structure and heterogeneity in a long‐term experiment. J. Appl. Ecol. 51, 968–977 (2014).

38. 38.

Siemann, E. Experimental Tests of Effects of Plant Productivity and Diversity on Grassland Arthropod Diversity. Ecology. 79, 2057–2070 (1998).

39. 39.

Scherber, C. et al. Bottom-up effects of plant diversity on multitrophic interactions in a biodiversity experiment. Nature. 468, 553–6 (2010).

40. 40.

van Klink, R. et al. Effects of large herbivores on grassland arthropod diversity. Biol. Rev. 90, 347–366 (2015).

41. 41.

Hao, S., Wang, S., Cease, A. & Kang, L. Landscape level patterns of grasshopper communities in Inner Mongolia: interactive effects of livestock grazing and a precipitation gradient. Landsc. Ecol. 30, 1657–1668 (2015).

42. 42.

Bonal, R. & Muñoz, A. Multi-trophic effects of ungulate intraguild predation on acorn weevils. Oecologia. 152, 533–540 (2007).

43. 43.

Mcnaughton, S. J., Oesterheld, M., Frank, D. A. & Williams, K. J. Ecosystem-level patterns of primary productivity and herbivory in terrestrial habitats. Nature. 341, 142–4 (1989).

44. 44.

Burns, C. E., Collins, S. L. & Smith, M. D. Plant community response to loss of large herbivores: comparing consequences in a South African and a North American grassland. Biodivers. C. 18, 2327–2342 (2009).

45. 45.

Roy, D. B. & Thomas, J. A. Seasonal variation in the niche, habitat availability and population fluctuations of a bivoltine thermophilous insect near its range margin. Oecologia. 134, 439–444 (2003).

46. 46.

Powell, G., Tosh, C. R. & Hardie, J. Host selection by aphids: Behavioural, evolutionary, and applied perspectives. Annu. Rev. Entomol. 51, 309–330 (2006).

47. 47.

Scriber, J. M. Evolution of insect-plant relationships: chemical constraints, coadaptation, and concordance of insect/plant traits. Proceedings of the 11th International Symposium on Insect-Plant Relationships. Springer Netherlands. 104, 217–235 (2002).

48. 48.

Fielding, D. J. & Zhang, M. Populations of the northern grasshopper, Melanoplus borealis (Orthoptera: Acrididae), in Alaska are rarely food limited. Env. Entomol. 40, 541–548 (2011).

49. 49.

Guihe, L. et al. The diet composition and trophic niche of main herbivores in the Inner Mongolia Desert steppe. Acta. Ecol. Sin. 33, 856–866 (2013).

50. 50.

Southwood, T. & Jepson, W. F. The frit fly-a denizen of grassland and a pest of oats. Ann. Ap. Biol. 49, 556–557 (1961).

51. 51.

Adesiyun, A. A. Effects of seeding density and spatial distribution of oat plants on colonization and development of Oscinella frit (Diptera: Chloropidae). J. Appl. Ecol. 15, 797–808 (1978).

52. 52.

Huston, M. Biological Diversity: The Coexistence of Species on Changing Landscapes (Cambridge, UK: Cambridge Univ. Press, 1994).

53. 53.

Stige, L. C., Chan, K. S., Zhang, Z., Frank, D. & Stenseth, N. C. Thousand-year-long Chinese time series reveals climatic forcing of decadal locust dynamics. Proceedings of the National Academy of Sciences of the United States of America 104, 16188–93 (2007).

54. 54.

Barton, P. S., Sato, C. F., Kay, G. M., Florance, D. & Lindenmayer, D. B. Effects of environmental variation and livestock grazing on ant community structure in temperate eucalypt woodlands. Insect Conservation & Diversity. 9, 124–134 (2016).

55. 55.

Whitcomb, R. F. & Hicks, A. L. Genus flexamia: new species, phylogeny, and ecology. Great Basin Naturalist Memoirs. 12, 224–323 (1988).

56. 56.

Joern, A. & Laws, A. N. Ecological mechanisms underlying arthropod species diversity in grasslands. Ann. R. Entom. 58, 19 (2013).

57. 57.

Vickery, J. A. et al. The management of lowland neutral grasslands in Britain: effects of agricultural practices on birds and their food resources. J. Appl. Ecol. 38, 647–664 (2001).

58. 58.

Stoate, C. et al. Ecological impacts of early 21st century agricultural change in Europe–a review. J. Envir. Mgm. 91, 22 (2009).

59. 59.

Polvan Dasselaar, A. V. D. et al. To graze or not to graze, that’s the question. Biodiversity and animal feed: future challenges for grassland production. Proceedings of the 22nd General Meeting of the European Grassland Federation, Uppsala, Sweden, 9–12 June 2008.

60. 60.

Schönbach, P. et al. Grassland responses to grazing: effects of grazing intensity and management system in an Inner Mongolian steppe ecosystem. Plant. Soil. 340, 103–115 (2011).

61. 61.

Wu, H. H. Study on the food adaptability of dominant grasshopper in Inner Mongolia steppe. Chinese Academy of Agricultural Sciences. (2012).

62. 62.

Odum, E. P. & Barrett, G. W. Fundamentals of ecology. Vol 3: Saunders Philadelphia (1971).

63. 63.

Shannon, C. E. & Weaver, W. The mathematical theory of communication (Urbana, IL: University of Illinois Press IL,1949).

64. 64.

Institute S. SAS/STAT user’s guide, Version 8. My Publications (1999).

65. 65.

Milauer, P. Multivariate analysis of ecological data using CANOCO. (Cambridge University Press, 2003).

66. 66.

ter Braak, C. & Smilauer P. CANOCO reference manual and CanoDraw for Windows user’s guide: software forcanonical community ordination (vers. 4.5), 500 p. Microcomputer Power, Ithaca, NY (2002).

## Acknowledgements

We are grateful to the anonymous reviewers for their constructive comments on earlier drafts of this manuscript. We also thank Prof. Mark McNeill (AgResearch, Canterbury Agriculture and Science Centre, Lincoln, New Zealand) for his invaluable suggestions on manuscript organization and linguistic revision. This study was supported by the National Natural Science Foundation of China, 31672485, the Earmarked Fund for China Agriculture Research System, CARS-34-07, and the Innovation Project of Chinese Academy of Agricultural Sciences.

## Author information

### Author notes

1. Jingchuan Ma and Xunbing Huang contributed equally to this work.

### Affiliations

1. #### State Key Laboratory for Biology of Plant Diseases and Insect Pests, Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing, 100193, P.R. China

• Jingchuan Ma
• , Xunbing Huang
• , Guangjun Wang
•  & Zehua Zhang
2. #### Scientific Observation and Experimental Station of Pests in Xilingol Rangeland, Ministry of Agriculture, Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Xilinhot, 02600, P.R. China

• Jingchuan Ma
• , Xunbing Huang
• , Guangjun Wang
•  & Zehua Zhang

• Xinghu Qin
4. #### Institute of Grassland Research, Chinese Academy of Agricultural Sciences, Hohhot, 010010, P.R. China

• Yong Ding
•  & Ning Wang
5. #### National Animal Husbandry Service, Ministry of Agriculture, Beijing, 100125, P.R. China

• Jun Hong
• , Guilin Du
•  & Xinyi Li
6. #### Grassland Workstation, Inner Mongolia, Hohhot, 010020, P.R. China

• Wenyuan Gao
•  & Zhuoran Zhang

### Contributions

Z.H.Z., N.W., J.C.M., X.B.H. designed the experiments. J.C.M., X.B.H., X.H.Q., G.J.W., W.Y.G., Z.R.Z. performed the experiments. J.C.M., X.B.H., J.H., Y.D., X.Y.L. analyzed the data. J.C.M., X.B.H. wrote the paper. All authors reviewed and considered the manuscript.

### Competing Interests

The authors declare that they have no competing interests.

### Corresponding authors

Correspondence to Ning Wang or Zehua Zhang.