Introduction

The Cetoniidae (Insecta: Coleoptera), known as flower and fruit beetles, are insects that have diurnal feeding habits. They feed on nectar, pollen, plant exudates, and ripe fruits (Ritcher 1958; Krikken 1984a), and are highly attracted to fermented fruits such as banana and pineapple (Pacheco et al. 2006; Jákl 2009; Orozco 2012; Evangelista Neto et al. 2017). Some species are recorded as agricultural, horticultural, and ornamental pests, while others are considered pollinators in various parts of the world (Molina 2001; Vuts et al. 2008, 2009; Aydin 2011; Subchev et al. 2011).

They form a diverse group of approximately 4,000 species, 300 of which occur on the American continent (Krikken 1984a; Orozco 2012). In Brazil, 75 species and 22 genera have been reported (Vaz-de-Mello and Puker 2021): three species are recorded from the Cerrado-Pantanal transition (Garcia et al. 2013), 10 in the Atlantic Forest (Gonçalves and Louzada 2005; Puker et al. 2014; Correa et al. 2021a), 16 in the Amazon (Valois and Silva 2015; Alves-Oliveira et al. 2016; Puker et al. 2020a, b), and 21 in the Cerrado (Rodrigues et al. 2013; Evangelista Neto et al. 2017; Correa et al. 2018). The Cerrado biome has the highest richness, accounting for 28% of the species and 36.4% of the genera of Cetoniidae known in the country (Garcia et al. 2013; Rodrigues et al. 2013; Evangelista Neto et al. 2017). This reinforces the importance of preserving the biome for the conservation of this important group of organisms.

In diverse groups of organisms, the spatio-temporal availability of food resources shapes the diversity (species richness and abundance) and behavior of communities of individuals, generating morphological adaptations and preferences for certain habitats (Arruda Almeida et al. 2018; Ribeiro et al. 2022). Representatives of Cetoniidae have a pronounced posthumeral marginalization that enables fast, long (in search of food and other resources), and sonorous flights with closed elytra (Le Gall 2010; Cherman and Morón 2014). For beetles, the elytra protect against dehydration, allowing adaptation to various habitats and as an aid to flight (Le et al. 2010; Johansson et al. 2012). Unlike other Coleoptera, Cetoniidae have better flight control, can hover, and have a quick response to predator threats (Krikken 1984b). The closed elytra flight may be related to diurnal life and sparse food resources, creating the need for long flights (Šípek et al. 2016). The diversity of this group seems to be influenced by microhabitats; for example, areas of cerrado sensu stricto show high species diversity (Evangelista Neto et al. 2017) in contrast to pasture areas and eucalyptus plantations, where fewer species are found (Correa et al. 2021b). Comparing urban parks and the Cerrado nature reserve area, abundance was 1.85 times higher in the nature reserve area even though the species richness was similar (nine species) (Correa et al. 2018).

The Cerrado presents high heterogeneity and strong seasonal variation, with respect to biodiversity and phytophysiognomies with two well-defined seasons: the dry (April to September) and the rainy season (October to March) (Silva et al. 2008). Three main types of vegetation formations are registered in this biome: forest, savanna, and grassland, which have different levels of tree density, humidity, and availability of resources. Among the forest formations, the gallery forest generates a dense cover and maintains high humidity even in the dry season because it accompanies streams and has tall trees (20–30 m tall) (Ribeiro and Walter 2008). In the savannah formations, the cerrado sensu stricto is the characteristic physiognomy of the biome representing 70% of the area (Felfili and Silva Júnior 2005) and presents low, twisted plants, with thick barked trunks and rigid leaves that can be 1.5–5 m tall, demonstrating adaptations to the dry season. In the grassland formations, campo sujo (common in Central Brazil) constitutes an open area with sparse shrubs and subshrubs (Sano et al. 2010).

Climatic seasonality in the Cerrado is reflected in seasonality in insect diversity by increasing or decreasing food resources (Oliveira and Frizzas 2008; Silva et al. 2011; Oliveira et al. 2021) and influencing vegetation (Becerra et al. 2009). Generally, insect diversity tends to be lower in dry months in regions with a well-defined dry season (Wolda 1978; Pinheiro et al. 2002). Studies conducted in the Cerrado have shown seasonality in Coleoptera diversity, with population dynamics being directly influenced by abiotic factors like precipitation and temperature (Evangelista Neto et al. 2017; Frizzas et al. 2020; Evangelista et al. 2021; Oliveira et al. 2021).

The high diversity of plant species recorded in the Cerrado, with 12,356 cataloged species (Mendonça et al. 2008), suggests that a wide range of food resources may contribute to the high diversity of Cetoniidae. However, to date, no study has evaluated the effect of physiognomies, which present variation in food resources and microclimate, on the diversity of these organisms. Thus, the objectives of this study were (i) to evaluate the diversity of Cetoniidae among the physiognomies: gallery forest, cerrado sensu stricto, and campo sujo in the Cerrado biome in Central Brazil, and (ii) to assess the seasonality of this group in these physiognomies.

Since the Cerrado physiognomic forms have different densities of trees, leading to variations in food resources and microclimate, we hypothesized that there will be significant differences in the diversity and seasonality of Cetoniidae among these environments. We expected that the campo sujo will present the highest diversity because it is an open area, which facilitates dispersal, foraging, and the perception of fermented fruit odors characterized by fast and long flights. We also hypothesized that this group of insects would present a seasonal distribution in the phytophysiognomies of cerrado sensu stricto and campo sujo because of the lower humidity in the dry season in this region. We expected that the highest diversity would be synchronized with the rainy season when the availability of food resources is greater. In the gallery forest, we expected a dispersed and more homogeneous distribution of species throughout the year because these areas retain moisture even in the dry season and may have greater availability of food resources throughout the year.

Methods

Study areas

The study was conducted in three conservation units located in Brasilia, Distrito Federal, Brazil (Fig. 1), which have characteristic vegetation of the Cerrado biome. The list of sites is as follows: (i) Parque Nacional de Brasília (PNB) has an area of 42,389 ha and represents the typical ecosystem of the Cerrado of the Central Plateau, including several types of vegetation and abundant diverse fauna. Moreover, it also facilitates the connection with fragments of native Cerrado that are essential for the formation of ecological corridors and the genetic exchange of fauna and flora species (ICMBio 2018). (ii) Fazenda Água Limpa (FAL), belonging to the University of Brasília (UnB), has an area of 4,500 ha including the area of ecological interest (ARIE Capetinga/Taquara) and falls under the Environmental Protection Area (APA of the Basins of Gama and Cabeça do Veado). Environmental heterogeneity provides diverse phytophysiognomies in the Cerrado biome. Both, floristic and faunistic diversities are very high and include species representing endemic plants and animals (PELD 2018). (iii) Reserva Ecológica do IBGE (RECOR) is part of the Environmental Protection Area (APA Gama-Cabeça de Veado District) with an area of 10,000 ha of continuous protected area. It is one of the core areas of the Cerrado Biosphere Reserve, featuring several types of Cerrado biome phytophysiognomies, springs of important watercourses that form the Paranoá Basin, and diverse wildlife (RECOR 2016; PELD 2018).

Fig. 1
figure 1

(Source: Google Earth Pro, 2021)

Conservation units selected for the collection of Cetoniidae in Brasília, Distrito Federal, Brazil (October/2016 to September/2018). PNB - Parque Nacional de Brasília, FAL - Fazenda Água Limpa (FAL), and RECOR - Reserva Ecológica do IBGE, location of the phytophysiognomies in each conservation unit and trap used for the collection of Cetoniidae. Green dots: gallery forest; blue dots: cerrado sensu stricto and yellow dots: campo sujo.

Full size image

Three phytophysiognomies were selected in each conservation unit: gallery forest, cerrado sensu stricto, and campo sujo, representing forest, savanna, and grassland vegetation formations, respectively. These physiognomies were chosen to select areas with a gradient in tree density, from physiognomy with the lowest tree density (campo sujo) to the physiognomy with the highest tree density (gallery forest) (campo sujo > cerrado sensu stricto > gallery forest).

Insect collection

We collected adults monthly for 24 months (October 2016 to September 2018) using baited traps. The trap consisted of a bait-holder pot (500 g) coupled with a larger plastic pot (1 kg). The larger pot had three 8 × 8 cm side windows and the bottom was removed. The lid of the bait-holder was glued to the bottom of the larger pot (with a hole in the center for the insects to pass through), allowing the bait-holder to be screwed on (Fig. 1). This trap is an adaptation of the pet bottle trap usually used for such collections (Orozco and Pardo-Locarno 2004; Orozco 2012; Rodrigues et al. 2013; Puker et al. 2014). Adaptations were made to facilitate the transport, installation, and removal of traps in the field. The bait-holders could be filled with the bait in the laboratory. In the field, the bait-holders with the collected specimens can be unscrewed and removed to replace with new bait-holders. The bait-holder contained 200 mL of bait, consisting of approximately 150 mL of Nanica banana (Dwarf Cavendish) and 50 mL of sugarcane juice that was fermented for 48 h (Rodrigues et al. 2013; Puker et al. 2014). All traps were installed at approximately 1.50 m from the ground level and remained in the field for seven days. In the gallery forest and cerrado sensu stricto, the traps were tied directly to the trees. In campo sujo, they were set on iron bars because the trees were sparse and low in this phytophysiognomy.

In each phytophysiognomy, two 500 m transects were demarcated, where 10 traps were set, spaced 50 m apart, and 50 m between the transect. Thus, we had 20 traps for each phytophysiognomy. The conservation units were considered repetitions with a total of 180 traps (3 conservation units × 3 phytophysiognomies × 20 traps). A 20 m margin was established from the entry point to the installation of the first trap.

Taxonomic identification

The Cetoniidae were pinned, labeled, separated by species, and counted. Taxonomic identification was performed by comparison with specimens from the Entomological Collection of the Zoology Department of the University of Brasília (DZUB, Brasília, DF, Brazil), and with the assistance of Dr. Fernando Zagury Vaz-de-Mello (UFMT). Vouchers of the collected material were deposited in the DZUB and in the collection of the Universidade Federal do Mato Grosso (CEMT, Cuiabá, MT, Brazil). Climatic data on temperature, relative humidity, and precipitation were obtained from the National Institute of Meteorology (INMET) throughout the collection period.

Data analysis

All data analyses were performed with R Program (R Core Team 2019). The data obtained (abundance and species richness) did not meet normality assumptions relating to the residues and the homogeneity of the variance. A generalized linear model (GLM) was used to verify if there were significant differences in the abundance and species richness (response variables) of Cetoniidae for the three phytophysiognomies (campo sujo, cerrado sensu stricto, and gallery forest). For these comparisons, a GLM was used in a logistic regression model using a negative binomial distribution. GLM selection was previously performed to choose the best model to fit count data using the “hnp” package (Moral et al. 2017). Mean values were statistically separated by Tukey’s HSD test at p < 0.05 using the function “glht” in “multcomp” package (Hothorn et al. 2016).

Similarly, we used GLM in a logistic regression model using a negative binomial distribution to verify the influence of climate variables (monthly average air temperature, monthly average relative air humidity, and monthly cumulative total precipitation) on Cetoniidae abundance and species richness. To identify indicator species among phytophysiognomies types, an indicator value (IndVal) approach was employed to characterize the habitats (Dufrêne and Legendre 1997). The ‘indicspecies’ package was used for the IndVal analysis in R (Caceres et al. 2016). The IndVal indicator attains the maximum value (1.00) when all of the individuals of a species are found in a single treatment (high specificity) and when the species occurs in all replicates of treatment (high fidelity).

The number of the individuals-based protocol was employed, and the accumulation curve was constructed in the R environment, using the “iNEXT” package (Hsieh et al. 2016). In the iNEXT analysis, individual-based abundance data and diversity order based on species richness (q = 0) were used. iNEXT can interpolate and extrapolate species richness by taking into account a measure of sample coverage (Chao and Jost 2012; Chao et al. 2014). Shannon diversity (q = 1) and Pielou evenness indices were calculated using the “BiodiversityR” package (Kindt 2018).

The diversity was analyzed by β-diversity, which expresses species turnover in terms of spatial and/or temporal scales. The β-diversity index expresses the faunistic change (turnover) between sampling (months) for each phytophysiognomy (intra-site heterogeneity) and the species turnover between adjacent phytophysiognomies (inter-site heterogeneity). This index was estimated using the following two formulas. The temporal β-diversity was first measured and calculated using Whittaker’s index: βw = (S/α) – 1, where α is the average of the α-diversity for all sampling periods (months) and S is the total number of species recorded in all sampling periods (Whittaker 1960). The βw value varied between 0 and n-1 (n = 12). The second formula measured the inter-site (spatial) β-diversity and was calculated using Cody’s index: βc = 100 * (Sg + Sl)/2α, where Sg is the increase in the number of species between two phytophysiognomies, Sl is the number of species lost, and α is the average of the α-diversity for the two phytophysiognomies (Cody 1975).

Circular analysis was used to examine the abundance of species in the different months of the year and among all three phytophysiognomies (Agostinelli and Lund 2013). The Rayleigh uniformity test was used to analyze temporal variation data, which analyzes the functional relationship of variables under the concept of dimensional homogeneity (Mendoza 1994).

Results

A total of 1,574 specimens, 8 genera, and 17 species of Cetoniidae were collected in 24 months (Supplementary Material S1). The species with the highest abundance were Gymnetis hebraica, comprising 22.7% of the total collection followed by Hoplopyga brasiliensis (19.8%), and Allorrhina menetriesii (16.3%). For the first time, we recorded the species Gymnetis hieroglyphica, Gymnetis pantherina, Hoplopyga miliaris, and Macrocranius similis.

We collected 106 individuals and 9 species from the gallery forest, 423 individuals and 15 species from the cerrado sensu stricto, and 1,045 individuals and 15 species from campo sujo (Fig. 2). The abundance and species richness of Cetoniidae were significantly lower (p < 0.01) for the gallery forest and did not differ between the phytophysiognomies cerrado sensu stricto and campo sujo (Fig. 2). Eight species were common to the three types of physiognomy while G. hieroglyphica and M. similis were collected only from campo sujo (Fig. 3b). No exclusive species were found in cerrado sensu stricto and gallery forest areas (Fig. 3c and d).

Fig. 2
figure 2

Boxplot graph of abundance and species richness of Cetoniidae species collected in traps baited with fermented fruit in campo sujo, cerrado sensu stricto, and gallery forest, in three conservation units (Fazenda Água Limpa, Parque Nacional de Brasília, and Reserva Ecológica do IBGE) in Brasília/DF, Brazil (October 2016 to September 2018). Values followed by the same letter in the columns do not differ significantly (p < 0.01) by Tukey’s HSD test

Full size image
Fig. 3
figure 3

Cetoniidae species collected in traps baited with fermented fruit in campo sujo, cerrado sensu stricto, and gallery forest, in three conservation units (Fazenda Água Limpa, Parque Nacional de Brasília, and IBGE Ecological Reserve) in Brasília/DF, Brazil (October 2016 to September 2018). (A) Total, (B) campo sujo, (C) cerrado sensu stricto and (D) gallery forest

Full size image

The seven species considered as indicators of campo sujo physiognomy were Hoplopyga albiventris (IndVal = 0.749; p < 0.001), E. lurida (IndVal = 0.726; p < 0.001), A. menetriesii (IndVal = 0.599; p < 0. 001), Hologymnetis undulata (IndVal = 0.581; p < 0.001), Hoplopyga singularis (IndVal = 0.580; p < 0.001), G. hieroglyphica (IndVal = 0.535; p < 0.001), and Hoplopyga miliaris (IndVal = 0.498; p < 0.001). For the gallery forest, the two indicator species were Inca bonplandi (IndVal = 0.665; p < 0.001) and Hoplopyga liturata (IndVal = 0.602; p < 0.001). For campo sujo and cerrado sensu stricto, the four indicator species were: G. hebraica (IndVal = 0.833, p < 0.001), H. brasiliensis (IndVal = 0.691; p < 0.001), Gymnetis rufilatris (Indval = 0.667, p < 0.001), and Allorrhina cincta (IndVal = 0.386; p < 0.001). No species were indicated as exclusive indicators of cerrado sensu stricto physiognomy.

The species accumulation curve indicated that sampling effort was satisfactory for campo sujo and cerrado sensu stricto, whereas, for the gallery forest, a greater sampling effort is necessary to reach the asymptote (Fig. 4). The campo sujo physiognomy showed the highest Shannon diversity index (H’ = 2.12) and highest Pielou’s equitability (J = 0.78), followed by the gallery forest (H’ = 1.61 and J = 0.73) and the cerrado sensu stricto (H’= 1.37 and J = 0.50).

Fig. 4
figure 4

Individuals-based accumulation curve of Cetoniidae species [based on species richness (q = 0)] collected in traps baited with fermented fruit in campo sujo, cerrado sensu stricto, and gallery forest in three conservation units (Fazenda Água Limpa, Parque Nacional de Brasília, and Reserva Ecológica do IBGE) in Brasília/DF, Brazil (October 2016 to September 2018)

Full size image

The β-diversity index expressing the faunistic change (turnover) between sampling months was higher in the gallery forest (βw = 3.50), followed by cerrado sensu stricto (βw = 2.40) and campo sujo (βw = 1.90). In terms of species turnover between adjacent phytophysiognomies (inter-site heterogeneity-β-spatial diversity), the largest difference occurred between campo sujo and gallery forest with βc = 41.7%, followed by cerrado sensu stricto and gallery forest with βc = 29.2%, and the smallest difference was found between campo sujo and cerrado sensu stricto (βc = 13.3%).

Cetoniidae were collected in all months of the year, with the highest abundance (n = 811) and richness (n = 15) observed in November, while the lowest abundance (n = 6) was observed in September and the lowest species richness (n = 1) in July and August (Fig. 5). In the rainy season (October to March), 1,424 individuals (90.5% of the total collected) and 17 species were collected. In the dry season (April to September), 150 individuals (9.5% of the total) and 7 species (42% of the species) were collected (Figs. 3a and 5). Except for G. hebraica, which was found throughout the year, most species were collected during the rainy season. Although April falls in the dry season, it represents a transition period, with 37 individuals and 7 species collected during this period. Gymnetis flava was collected only during the transition period and in the dry season. Precipitation showed a significant effect on abundance (z = 3.346 and p = 0.00082) and species richness (z = 3.059 and p = 0.00222) of Cetoniidae.

Fig. 5
figure 5

Species richness and abundance of Cetoniidae collected in traps baited with fermented fruit in campo sujo, cerrado sensu stricto, and gallery forest in three conservation units (Fazenda Água Limpa, Parque Nacional de Brasília, and Reserva Ecológica do IBGE) in Brasília/DF, Brazil (October 2016 to September 2018) during dry and rainy seasons in the Cerrado

Full size image

The circular analysis showed that the abundance of Cetoniidae presents a well-defined peak in November for all physiognomies (Fig. 6); however, the distribution throughout the rainy season varied. The aggregate distribution of abundance was concentrated in campo sujo during November and in the gallery forest between November and January. Cerrado sensu stricto showed a dispersed distribution with a wider distribution throughout the rainy season (October to January) (Fig. 6), with small peaks during the dry season due to the presence of G. hebraica (Table 1). Rayleigh’s uniformity test showed that there was significant temporal variation in Cetoniidae abundance (p < 0.0001) in the three phytophysiognomies (Table 1).

Fig. 6
figure 6

Circular analysis for total abundance of Cetoniidae (A) collected in traps baited with fermented fruit in campo sujo (B), cerrado sensu stricto (C) and gallery forest (D) phytophysiognomies in three conservation units (Fazenda Água Limpa, Parque Nacional de Brasília, and Reserva Ecológica do IBGE) in Brasília/DF, Brazil, (October 2016 to September 2018)

Full size image
Table 1 Measurements of the concentration (r), Rayleigh Test, mean group, distribution, and season of the year with the greatest total abundance of Cetoniidae species collected in three phytophysiognomy of Cerrado from October 2016 to September 2018
Full size table

Discussion

Our results confirm our hypothesis that more open Cerrado physiognomies (campo sujo and cerrado sensu stricto) support the highest diversity of Cetoniidae when compared to forested areas (gallery forest). We also confirmed that although the Cetoniidae occur throughout the year in the Cerrado, in general, abundance and species richness were concentrated in the rainy season. However, the seasonality of the group varies among the three phytophysiognomies, being aggregated in campo sujo and gallery forest and dispersed in cerrado sensu stricto. Thus, the hypothesis that the community in the gallery forest would not be affected by the seasonality of the climate was not confirmed, possibly because the species are linked to the seasonality of the Cerrado biome and not to the specific microclimate of the physiognomy. Further studies are required to address the biology of the species, interactions with plants, natural history, and ecosystem services.

The gallery forest showed the lowest diversity compared to the grasslands and savannah vegetation formations (Fig. 2). Higher diversity observed in campo sujo and cerrado sensu stricto can be explained by the drier microclimate in these areas, which may favor a faster loss of moisture from the diet and the intensification of volatiles from fermentation. As open areas are more, the volatiles can spread easily and quickly via air. The lower presence of physical barriers (lower density of woody plants) compared to forest formations may also facilitate the displacement of specimens (Fuentes et al. 2016). Cetoniidae are diurnal beetles with long flights (Le Gall 2010; Cherman and Morón 2014) and probably take advantage of more open Cerrado areas. This is because despite sparser food resources, locating is faster and less energetically expensive; the time spent foraging is reduced thereby reducing the probability of predation (Goodell 2003). Another important fact is that grassland (8,848 plant species) and savanna (7,618 plant species) formations have a higher vegetation diversity than forest formations (6,998 plant species) (Mendonça et al. 2008), which may also have contributed to the higher diversity of Cetoniidae in these types of vegetation formation by representing a greater source of food resources.

Insects have flight properties such as height, speed, and maneuvering ability. These may vary among species and are directly linked to ecological factors such as feeding habits, habitat, and predation (Šípek et al. 2016; Kojima and Kato 2017; Le Roy et al. 2019; Farisenkov et al. 2020). Most beetles open the elytra during flight; however, the families Cetoniidae, Scarabaeidae, Cerambycidae, and Buprestidae have species that are known to fly with the elytra closed or partially raised (Krikken 1984a; Šípek et al. 2016; Kojima and Kato 2017). Rapid flight has the benefit of dodging predators, and the closed elytra may reduce water loss during diurnal flight (Chown et al. 2011; Šípek et al. 2016). This type of flight is associated with diurnal life, allowing insects to utilize dispersed and unpredictably distributed food resources represented by uneven flowering of fruit trees (Šípek et al. 2016), as occurs in the Cerrado.

It is also worth noting that the larvae and adults of Cetoniidae have distinct nutritional biology. Larvae are usually saprophytes or saproxylophages, and can often be found in soil, decaying vegetables, rotting wood, tree cavities, feces (Luederwaldt 1911; Morón and Arce 2002; Puker et al. 2014), and may also live in association with social insects (Krikken 1984a; Micó et al. 2000; Peter and Johnson 2009; Puker et al. 2012). Adults typically seek out flowers or inflorescences that are large and open, yellow, white, purple, or pink, as well as fermented fruits (Peter and Johnson 2009; Puker et al. 2014, 2015), foods with high sugar concentration, and fermenting yeasts. After emergence, males search for trunks, branches, flowers/inflorescences, or fruits to find mating partners (Arce-Pérez and Morón 1999). The search for mates and food, therefore, has an important relationship with the dispersal ability of these insects, especially in males. After mating, males disperse widely in search of food in flowers. Despite the distinct biology of immatures and adults, most developmental requirements of the Cetoniidae are met in the Cerrado explaining the high richness of the group in this biome. The Brazilian Cerrado contains 12,356 plant species (Mendonça et al. 2008) and has approximately 44% of endemic flora making it the most diverse savannah on the planet (Klink and Machado 2005). It contains several endemic fruiting species that are rich in nutrients and diverse flowers.

We recorded 13 species that are considered indicators for the three physiognomies. We found seven species for campo sujo, two for gallery forest, and no exclusive indicator species for cerrado sensu stricto, indicating that this may be a transition area between closed and open formations. Among the indicator species of campo sujo, immatures of H. albiventris and H. singularis were found associated with termites and the adults feeding on the sap of Asteraceae Baccharis and Vernonia (Shaughney and Ratcliffe 2015). Adults of E. lurida can be considered generalists because they are reported pests of maize, apple, persimmon, grapes, peach, safflower (Bertels and Baucke 1966; Garcia and Corseuil 1999; Cunha et al. 2007; Androcioli et al. 2017). They also act as floral visitors of sunflower (Torretta et al. 2009) and cotton (Dutra et al. 2012) and as pollinators of orchids (Singer and Cocucci 1997). Adults of H. miliaris are described as floral visitors of the Apocynaceae (Koschnitzke 2015). Cetoniidae tend to be regarded as resource exploiters rather than pollinators (Di Iorio 2013; Iorio 2014) because they may pollinate flowers accidentally while searching for resources (Peter and Johnson 2009).

In the gallery forest, the indicator species I. bonplandi and H. liturata were more abundant in August (Fig. 3D), which is characteristic of the dry season in the Cerrado. Despite having smaller areas and being more isolated, gallery forests have a vegetational structure characterized by the dominance of tree species and canopy formation that are associated with watercourses (Ribeiro and Walter 2008). This probably maintains humidity in dry months and ensures the maintenance of resident species during this period. These species may also be favored by less competition from other species because the diversity of insects and the occurrence of natural enemies are lower in the Cerrado during the dry season (Ramos and Diniz 1993; Morais and Diniz 1999; Silva et al. 2011). Adults of I. bonplandi feed on plant exudates and flowers (Boos and Ratcliffe 1985) and immatures on decaying wood (Morón 1983; Costa et al. 1988) and the organic material derived from palm trees (Sousa et al. 2018). Hoplopyga liturata adults feed on exudates from plants of the families Boraginaceae, Fabaceae, Zingiberaceae, and Rhamnaceae, whereas, immatures are found in decomposing organic material of the families Myrtaceae, Burseraceae, Solanaceae, and Moraceae (Shaughney and Ratcliffe 2015).

The species H. brasiliensis, G. hebraica, G. rufilatris, and A. cincta were recorded as indicator species of open Cerrado vegetation formations (cerrado sensu stricto and campo sujo). Immatures of H. brasiliensis are found in association with the termite species Cornitermes cumulans, which is considered a key species because of its abundance and the consequent impact on the environment. Moreover, the nests of this species are capable of harboring several groups of arthropods (Redford 1984). The abundance of H. brasiliensis adults, mostly collected in campo sujo, may be linked to the presence of C. cumulans nests. Gymnetis hebraica is one of the most abundant species and is the only species collected year-round in the open Cerrado, mainly in cerrado sensu stricto. It is a generalist species well adapted to the Cerrado climate conditions and this behavior of G. hebraica was observed in another study in the Cerrado (Evangelista Neto et al. 2017). For most other collected species, no information on their biology or natural history was found.

Regarding the sampling effort, we observed that the species accumulation curves for campo sujo and cerrado sensu stricto reached the asymptote, but not for the gallery forest (Fig. 4). The gallery forest showed a large number of rare species (J = 0.73), which reinforces the need for greater sampling effort for this type of physiognomy in future studies.

The ability to fly allows organisms to colonize different environments, and the species that constitute these communities may share similarities or differences based on the specific requirements of the organisms and the differences and biotic and abiotic similarities of the environments. In the present study, temporal β-diversity was higher in the gallery forest, suggesting that a greater variation (qualitative and/or quantitative) in the availability of resources throughout the year possibly promotes changes in the species composition of Cetoniidae in this environment. Regarding the βc diversity index, which measures the changes in composition among the phytophysiognomies, it was observed that the communities of open areas were more similar to each other with a change in species composition of 13.3% throughout the study. This change in composition was maximum between campo sujo and gallery forest (41.7%), showing a gradient of similarity from the more open areas to the more closed areas. For the Cetoniidae, it was observed that the gallery forest has a specific community and that this community is the one that most varies over time, indicating that conservation efforts in this phytophysiognomy need more attention.

The seasonality in the abundance of insects is synchronized with the availability of food resources, and these in turn vary seasonally depending on the climate (Wolda 1978). The Cerrado has two well-defined seasons, and the majority of adult insect populations can be found in the rainy season from October to March (Silva et al. 2011; Oliveira et al. 2021). The seasonal variation in abundance and richness of Cetoniidae is directly related to the climatic characteristics of the Cerrado, with 90.5% of individuals and 100% of species occurring in the rainy season and 9.5% of individuals and 41.2% of species in the dry season. Furthermore, we recorded a direct relationship between precipitation and the diversity of Cetoniidae. The onset of rainfall is considered to trigger increased population density (Wolda 1978; Silva et al. 2011), and other studies conducted with Cetoniidae in the Cerrado (Evangelista Neto et al. 2017; Correa et al. 2018), Cerrado-Pantanal transition (Garcia et al. 2013), Atlantic Forest (Puker et al. 2014), and Amazon (Puker et al. 2020a) also recorded this behavior. The circular analysis confirmed this pattern, showing that the distribution of abundance was concentrated in the first quarter of the rainy season with November presenting the highest number of individuals. However, there was a variation in the distribution of abundance throughout the rainy season among the three phytophysiognomies. The overabundance of some species such as A. menetriesii in November in campo sujo or the presence of abundant species that are distributed throughout the year as G. hebraica in cerrado sensu stricto can generate differences in the patterns observed in different physiognomies.

The Cerrado offers favorable conditions for agriculture. Consequently, this region has experienced a constant reduction in natural vegetation. The Cerrado in the central-western region of Brazil has the highest degree of deforestation (61.6%) (Tisott and Schmidt 2021). The increase and strengthening of agribusiness in the biome trigger negative environmental effects, such as deforestation, biodiversity loss, groundwater pollution, and soil erosion (Brussaard et al. 2010). The fragmentation of natural habitats due to the advancement of agricultural frontiers modifies niches and the microclimate of these habitats, reducing the availability of food resources (quality and/or quantity), and shelter for numerous species, including the insects and plants that shelter these species (Turner 1997). The increase in agricultural areas in the Cerrado may also increase the incorporation of organic matter into the soil through no-till farming and generate a population growth of some species. For example, in Peru, an increase in the population of Gymnetis species was observed in apple plantations due to environmental factors (incorporation of organic matter into the soil and constant humidity) and the high number of adults searching for food resulted in the invasion of beehives by insects attracted by concentrated sugar substances (Tejada and Morón 2015).

Knowledge about the ecological requirements of species is crucial for their conservation, especially if they live in threatened habitat types such as old-growth forests. Studies on the factors that affect the abundance and distribution of insects and the ecological requirements that influence habitat selection are vital to ensure proper species and habitat management (Thomas et al. 1998, 2001; Krämer et al. 2012) because they elucidate the complex trade-off between various biotic and abiotic factors. The information gathered here is important for the assessment and preservation of habitats by assisting in public measures for the maintenance and creation of conservation units. The Cerrado presents 11 main types of physiognomies (Ribeiro and Walter 2008), and we evaluated only three of these physiognomies. Moreover, four species of Cetoniidae were recorded for the first time in the Federal District, suggesting the need for more studies in other physiognomies to uncover information about the real diversity of Cetoniidae in the Cerrado biome.