How does the transformation of Cerrado vegetation into exotic pastures affect insect communities? A meta-analysis

Cerrado is the largest and most biodiverse savannah worldwide. However, only a reduced amount of this ecosystem is conserved, and in the last decades, the vast agricultural expansion resulted in the substitution of native vegetation for exotic pasturelands.

Insects are key components for both native and exotic ecosystems in Cerrado, performing critical functions for ecosystem maintenance. Our goal was to evaluate how the transformation of natural Cerrado into cattle pastures (exotic pastures) affects the insect communities in this biome.

We conducted a meta-analysis using data on species richness and abundance data of insect assemblages from conserved Cerrado and exotic pastures. The analysis was based on a set of 25 scientific articles. Fourteen studies focused on Coleoptera and nine on Hymenoptera.

Conserved Cerrado encompassed a higher species richness of insects than exotic pastures, but their abundance did not differ between these two habitats. However, dung beetle abundance increases significantly in disturbed Cerrado over time. Synanthropic species that thrive in altered environments could be masking the actual decline in abundance of species sensitive to environmental disturbances provoked by exotic pastures.

Our results highlight the consequences of the structural homogenization of the Cerrado into cattle pastures. That is, despite having a similar open-canopy vegetation structure, this land-use change leads to a significant loss of insect species, resulting in a drastic simplification of insect communities. This study reinforces the importance of maintaining native Cerrado patches in order to maintain diverse and functional ecosystems in this region.

Methods
Literature search

We conducted a systematic search of peer-reviewed literature using the Web of Science (WoS) and Scopus databases to investigate the effects of Cerrado conversion into exotic pastures on insect communities. The search was performed during the second half of 2023 and included articles published from 1984 to 2023. The following search terms were used: (“pasture*” OR “introduced pasture*” OR “livestock*” OR “land use*” OR “land-use*”) AND (“land-use change*” OR “land-use modif*” OR “Anthro*" “disturb*” OR “Disturb*” OR “Anthropiza*”) AND (“Brazilian Cerrado*” OR “Brazilian Savanna*” OR “Cerrado*” OR “native grassland*” OR “natural grassland*”) AND (“insect*” OR “bioindicator*” OR “ant*” OR “beetle*” OR “flies*” OR “fly*” OR “invertebrate*” OR “termite*” OR “butterfl*” OR “moth*” OR “Lepidotper*” OR “Hymenopt*” OR “bee*” OR “Coleopt*” OR “dragonfl*” OR “Odonat*” OR “Isopter*”). A complementary search was conducted in SciELO-Brazil (https://www.scielo.br/) to include articles in Portuguese. The same keywords used in the main search were translated into Portuguese (“pasto*” OR “pasto introduzido*” OR “pecuária*” OR “uso do solo*”) AND (“mudança do uso do solo*” OR “Antro*" “disturb*” OR “Disturb*” OR “Antropiza*”) AND (“Cerrado brasileiro*” OR “Savana brasileira*” OR “Cerrado*” OR “campos nativos*” OR “campos naturais*”) AND (“inset*” OR “bioindicador*” OR “formiga*” OR “besouro*” OR “mosca*” OR “invertebrados*” OR “cupins*” OR “borboleta*” OR “mariposa*” OR “Lepidotper*” OR “Hymenopt*” OR “abelha*” OR “Coleopt*” OR “libelula*” OR “Odonat*” OR “Isopter*”).

Inclusion criteria and screening protocol

After removing duplicates, the search yielded 594 publications (Fig. A.1). A study was included in our database if it met the following criteria: the research focused on the diversity or abundance of terrestrial insect communities within the Cerrado ecosystem, the study featured both a control group, comprising conserved Cerrado, and a treatment group, consisting of exotic pastures. We chose 39 papers for comprehensive assessment in accordance with the PRISMA declaration (Page et al., 2021; Fig. A.1). The complementary search yielded a total of 101 results. Titles and abstracts were screened to assess thematic relevance, and full texts of potentially eligible studies were examined to confirm eligibility. After this process, only one article in Portuguese was selected for a full review.

Data extraction

We extracted the following information from the selected literature: a) taxonomic identity and abundance of the insects reported in each study; b) sample size (n), represented by the number of replicates of each sampled habitat type (i.e., Cerrado and exotic pasture); c) measures of variability for diversity or abundance data, including standard error, standard deviation, or any other form of variance; d) geospatial data, encompassing specific location of each study; and e) the year the data was collected, as reported in each paper. When the means and/or variance data was not explicitly expressed in the text of a given study, then we extracted this data from the figures. Hozo et al.’s (2005) approach was used to obtain the means and standard deviations when a paper provided variance data such as medians and interquartile range. We omitted only two papers because they were single-site studies lacking any measure of variance.