Diversity of the families of Coleoptera captured with pitfall traps in five different environments in Santa Maria , RS , Brazil

The conservationist community is giving special attention to epigean insects due to their importance in the preservation of terrestrial habitats. This work analysed the diversity, richness, abundance and similarity at the soil surface of Coleoptera composition among five environments: native forest, native grassland, Pinus elliottii plantation, Eucalyptus saligna plantation and degraded area by soil use in southern Brazil, from October 2004 to October 2005. A total of 1,812 individuals were collected, attributed to 45 morph-species and 14 families. The higher richness and abundance were observed in native forest (31 species and 782 individuals) and the lower richness and abundance in degraded area (14 species, 86 individuals). Scarabaeidae was the richest family captured, with nine morph-species, and the most frequent family was Nitidulidae (1,113 individuals). According to the Shannon-Wiener index, the degraded area had smaller diversity in relation to the native forest, E. saligna and Pinus elliottii plantations. No difference in diversity between evaluated areas was found for the Simpson diversity index. The most dominant site was the degraded. The correlation between the total number of morph-species captured was not different to the degraded area and the P. elliottii monoculture (r = 0.47) and the correlation between the total individuals number was not significant between native forest and native grassland (r = 0.46) and between degraded areas and the other sites. According to the Jaccard Similarity Index, the greatest similarity for the organism composition occurred between P. elliottii plantation and E. saligna plantation, presenting 74% of overlap.

The following were analysed: a) the richness and abundance variability of Coleoptera families between environments; b) the existence of patterns on the presence of families in each environment; c) if these patterns may be related to the of degree human actions in ecosystems; d) the structure of communities in relation to the trophic level occupied by the Coleoptera families; e) the relationship between the time period and the increase in the number of Coleoptera species; f) richness and diversity estimates and uniformity values of the sample areas; and g) the similarity between the investigated areas based on species abundance and richness of Coleoptera families.

Characteristics of the study locations
The study was developed in the "Centro de Pesquisas de Florestas e Conservação do Solo da Fundação Estadual de Pesquisa Agropecuária (FEPAGRO)", Santa Maria, Rio Grande do Sul (29° 43' 12" S and 53° 43' 01" W) (Figure 1).According to Maluf (2000) the climate is classified as Temperate Superhumid (Te Su) with an annual mean temperature of 19.2 °C.Rainfall is more or less evenly distributed throughout the year, with an annual mean of 1,708 mm (Maluf, 2000).
Five quarterly collections were made between October 2004 and October 2005 in five different environments.Native grassland: characterised by the homogeneity of the vegetation, with strong predominance of grasses.Degraded area (reject): where the soil does not present its surface layer, being completely naked, compacted, and having suffered the strong action of erosion processes.A 15-year-old Pinus elliottii plantation: showing almost absentee subforest vegetation, and litter with a thickness around ten centimetres.A 30-year-old Eucalyptus saligna plantation: showing considerable vegetation in the sub-forest and litter layer with an approximate thickness of five centimetres.In the native forest studied, the families Anacardiaceae, Annonaceae, Bignoniaceae, Caesalpinaceae, Fabaceae, Mimosaceae, Moraceae and Myrtaceae are found.This area is distinguished by the largest heterogeneity and density of the vegetation.The litter layer is around two centimetres thick.The size of each area evaluated and the minimum distance between them was calculated with the program Image J 1:41 (Rasband, 2009), from satellite images, obtained by Google Earth, with a point of view of two kilometres of altitude and with a mean total area of 140 ha (Table 1).

Introduction
Insects are the largest group among animals, occupying several ecological niches, being considered very important in the dynamics of natural ecosystems (Borror et al., 1992;Kim, 1993;Gullan and Cranston, 1996;Thomazini and Thomazini, 2000).These organisms exhibit great diversity in terms of species and colonisation of habitats, are easy to sample (Rosenberg et al., 1986) and present high sensitivity to environmental interference.Therefore, the composition of this group reflects the performance of the ecosystem (Correia and Pinheiro, 1999) and is fundamental in the analysis of the landscape structure.
The quality and quantity of litter are related to the richness and composition of epigean insects, since these characteristics may create different conditions on the soil surface (Loranger et al., 1998).The abundance and the structure of the edaphic insect communities depend on the alimentary supply and on the environmental properties conditioned by the local vegetation (Dunxiao et al., 1999).More complex environments support larger diversity of niches, resulting in a greater number of fauna nesting sites and feeding areas (Savolainen and Vepsäläinen., 1988).Thus, the structural diversity of the vegetation is highly correlated to the diversity of arthropods, and it is used in the understanding of disturbances caused by the simplifications of the natural ecosystems (Hutcheson, 1990;Loranger et al., 1998).Generally, the changes in abundance, diversity and composition of indicative species that depend on particular resources of the system may evaluate and explain the results of disturbance (Lewinsohn, 2001;Thomazini and Thomazini, 2002).According to Buchs (2003), biological diversity, in particular, is one of the most important tools in monitoring of health and functioning of ecosystems.
Coleoptera is the largest order of insects, with about 400 thousand species worldwide, representing 30% of animals and about 40% of all insects (Lawrence and Britton, 1991;Lawrence and Newton, 1995;Costa, 2000).The success of the group is due, mainly, to the presence of hardened fore wings, the ability to consume a wide variety of materials and holometabolia (Daly et al., 1998;Costa, 2000), characteristics that have enabled the conquest of different environments during their evolution.
Insects of this order are important in decomposition, nutrient cycling, pollination, seed dispersal and control of other animal populations (Costa, 2000;Davis et al., 2001).Furthermore, beetles are indicators of soil properties (Dunxiao et al., 1999), temperature and humidity variation of the environments (Stork and Eggleton, 1992), forest disruptions (Halffter and Favila, 1993;Davis et al., 2001) and environmental disturbances in the landscape structure (Buchs, 2003).According to Brown Junior (1997), this order satisfies all requirements to be used for environmental bioindication (Marinoni and Dutra, 1997).
This study aimed to evaluate the differences in the composition and structure of epigean Coleoptera communities in five different environments in Santa Maria, RS, Brazil.
The diversity analysis among the ecosystems were evaluated by the following diversity indexes: Shannon-Wiener (H'), which consider equal weight to the rare and abundant species and Simpson's index (1-D), which is characterized by being sensitive to changes in the most abundant species composition (Peet, 1974) were employed.The differences between these indexes were calculated using the Student test (t-test).
The uniformity among the coleopterans caught in the five sampling areas was calculated with the Berger and Parker index.Non-parametric tests, Jackknife one and two were used to estimate the number of species in the studied environments, based on abundance data.The relationship between the "singletons" and "doubletons" was utilised to estimate the richness and a Bootstrap procedure of 1.000 random samples and confidence interval of 95% was employed.
10% and they were kept in the sample areas for two days.After this period, they were transported to the laboratory to identify the material.

Identification and data analysis
After being mounted on entomological pins, the insects were separated and quantified at the family level (Borror et al., 1992;Lawrence et al., 1999), and morph-species level.Although not conventional, the analyses were performed through the abundance of Coleoptera at the family level (Marinoni and Dutra, 1997).Pielou (1975) and Magurran (1988) consider it valid.The Kruskal-Wallis test was employed to evaluate the differences in richness and abundance among the Coleoptera families in the investigated environments.To compare the addition of species during the study period, a rarefaction curve was utilised, because the sample size was different among the studied environments (Moreno, 2001).Table 1.Size of the five areas assessed in Santa Maria, Rio Grande do Sul, Brazil, from October/2004 to October/2005 and the distance between them.

Eucalyptus saligna
Native Analysing each study area separately, Nitidulidae was the richest family among the ecosystems.However, in the native forest, Scarabaeidae and Carabidae had the same number of morph-species as the supracited family (6).Exceptionally, in native grassland, Scarabaeidae was the richer family (6).The Nitidulidae family was also the most abundant family in all ecosystems, except in native grassland where the Staphylinidae represented the most of the captured individuals.Mordelidade and Ostomidae families were registered only in the native forest, while the Cucujidae family occurred only in the P. elliottii plantation and native grassland.The Cicindela genus (Carabidae -Cicindelinae) occurred only in native forest and in the E. saligna area.Twenty-six individuals were captured of the same morph-specie of the Carabidae in the degraded area.Among the families that did not occur in the degraded area are: Coccinelidae, Cucujidae, Histeridae, Mordelidae, Ostomidae and Scolytidae.The last three families also did not appear in native grassland.

Trophic groups
Of the 14 sampled families, four have an exclusively carnivorous habit, three are herbivores, two are herbivores or detritivore, one is fungivore or detritivore, one is Spearman correlation was applied to group areas of larger similarity based on coleopterans species abundance.This analysis was processed with the Past program version 1.32 (Hammer et al., 2004).In the grouping of similar environments based on the coleopterans families found in each one, the Jaccard similarity index was used, which considers the presence and absence of species.The UPGMA method was used to group the areas.The evaluation of the trophic groups of the coleopteran families registered in the five environments was made according to Marinoni et al. (2001).The normality of all data was tested by the Shapiro-Wilk method.All tests considered a significance of 0.05.

Richness and abundance
In the five survey areas, 1,812 beetles were caught, distributed in 45 morph-species from 14 different families (Table 2).The "singletons" frequency in the areas ranged from 9.7% in native forest to 27.8% in P. elliottii plantation.The degraded area showed the lowest number of "singletons".The native forest had the highest richness, presenting 31 morph-species (68.9%), while the lowest richness occurred in the degraded area.At this site, only 14 morphspecies (31.1%) were found.The native forest also showed the greatest number of individuals (n = 782), including 43.2% of the total captured beetles, followed by E. saligna monoculture, with 473 individuals (26.1%).The lowest abundance of beetles was found in the degraded area, where only 86 specimens (4.7%) were captured.
Among the families recorded, Scarabaeidae was the family that exhibited the highest number of morph-species (9),

Richness estimates
Jackknife two was the species richness estimate which resulted in the largest values for all the areas the whole sampling period (Table 4) was analysed.The bootstrap procedure resulted in the smallest values.The native fungivore or herbivore and three possess varied habits.The percentage of each group is given in Table 3.

Species accumulation curve
The graph of the species collected shows a decrease in the number of the new species of beetles captured in each sample period (Figure 2).In the first sample, the total number of individuals captured exceeded 50% of The correlation between the total number of morphspecies found was not different to the degraded area and P. elliottii monoculture (r = 0.47) and the correlation between the total number of individuals was not significant between native forest and native grassland (r = 0.46) and between the degraded area and others sites (Table 7).
Cluster analysis from the Jaccard similarity index revealed that P. elliottii and E. saligna plantations had the most similar composition of organisms, showing an overlap of 74% (Figure 3).

Discussion
The greater coleopterans richness and coleopterans abundance in forest is related to the benefits obtained by these organisms from the soil moisture and microclimate stability offered by forests (Butterfield et al., 1995).Panzer and Schwartz (1998) observed that the plant species richness explained more than 49% of the variance of the insect species richness among the studied areas.The abundance of beetles found in E. saligna plantation may be associated to the ecological plasticity of some species forest and native grassland presented the highest richness estimates and the degraded area the smallest ones for both jackkinfe one and jackknife two.

Diversity indexes and uniformity
Diversity indexes show some differences among the five sites (Table 4).For the Shannon-Wiener index, native forest (2.26), P. elliottii (2.17) and E. saligna monocultures (2.17) are the ecosystems with the largest diversity.Simpson's index shows the same pattern.The values obtained for these sites are not different for both diversity indexes (Table 5).The degraded area exhibited a significantly smaller diversity in relation to the other investigated areas (except to the native grassland) for the Shannon-Wiener index.However, for the Simpson index, the degraded area did not differ in relation to the diversity from other sites.The most dominant site was the degraded area (BP = 0.29) (Table 4) and native forest had the lowest dominance (BP = 0.19), coinciding with the high diversity found in this site.

Structure of the communities
There was no significant difference in abundance among Coleoptera families collected in the environments assessed (Table 6).However, a significant difference   (Rice, 1974;Silva, 1978).Furthermore, this site presented a discontinuous canopy and high solar radiation (Ferreira and Marques, 1998).

Areas
The lack of vegetation in the degraded area may have contributed to reduce the richness and abundance of the beetle species due to the low availability of food resources (Thomazini and Thomazini, 2002) and to the decrease of soil moisture at this site (Stork and Eggleton, 1992).Perner and Malt (2003) reported that beetles have clear reactions to microclimate and soil moisture changes.Among the identified coleopteran families, Scarabaeidae presented the highest number of morph-species.These beetles are important in organic matter recycling (Kim, 1993) and in the seed dispersals (Davis and Sutton, 1998).Meanwhile, the Nitidulidae family was the most abundant, followed by Staphylinidae.The abundance of Nitidulidae in the study areas is due mainly to the different habits of its species.They can distribute themselves in the most diverse ecosystems, fuelling up on different materials (Audino et al., 2007).The abundance of Staphylinidae, however, is associated with its large diversity.This group is among the most common and important organisms (Butterfield et al., 1995) and to the amount of litter in this area, originated from the great sub-forest presence.This characteristic stimulates the beetles that are responsible for nutrient cycling and organic matter decomposition.However, the richness in E. saligna plantation is different from native area and native grassland.This result may are responsible for the input energy.The activity of these organisms, therefore, influences the whole ecosystem (Crossley et al., 1992).Detritivores require environments with relatively dense vegetation and soils with thick layers of leaf litter (Iannuzzi et al., 2003).The phytophages are particularly dependent on the physical structure and floristic composition of their habitats (Brown Junior, 1997).Thus, the units of landscape of this study should have influenced, especially, Cerambycidae, Curculionidae and Mordelidae diversity.This hypothesis may be verified with the use of a methodology to capture species that occur in another stratum of vegetation.
The greater richness and greater diversity of insects found in native forest is related to its complex structure, which includes many plants species, vertical stratification and interconnected canopies (Elton, 1973).This environment provides the beetles with a greater diversity of food, more stable microclimate, higher humidity and greater quantity of refuges against predators (Vallejo et al., 1987).There was no difference in Coleoptera diversity found in native forest and E. saligna plantation.This result may reflect the ability of E. saligna plantation to support distinct species originated from the great sub-forest presence (Butterfield et al., 1995), since this environment presented an inferior amount of beetles in relation to the native forest.Monoculture areas that have a discontinuous canopy, like the P. elliotti site studied, allows a large sunshine input, high evaporation and strong impact of rain on the soil, causing the reduction of the soil fauna (Ferreira and Marques, 1998).The variation of the species diversity is influenced also by other factors such as phylogenetic diversity (May, 1990) and endemism (Jetz et al., 2004).
As the pattern observed by Odum (1985), the high dominance of beetle species in the degraded area reflected a significant difference between the richness estimate in this environment and in the native grassland and native forest.The degraded area provides a lower number of insect species due to its low diversity of plants and the high compression of the first centimetres of soil, damaging the animals that move among its particles (Dorst, 1973).
The large presence of plant material in native forest soil and E. saligna monoculture should have allowed the coexistence of a larger number of morph-species beetles in these sites.In contrast, there was no correlation in the total number of Coleoptera species captured between the degraded area and the P. elliottii monoculture, and there was no correlation in the total number of individuals collected between native forest and native grassland and between degraded area and the other sites.Thus, the abundance and richness of beetles in the studied environments seems to depend not only on the composition and on plant structure, but on the quantity of plant product in each ecosystem.
The most similar areas in the composition of Coleoptera species were E. saligna and P. elliottii plantations.The similarity in their vegetable physiognomies characterised by structural homogeneity may explain this result.Areas with Eucalyptus plantation have many opportunistic ant species and low occurrence of rare species (Ramos et al., of soil fauna and they are found in almost all types of ecosystems (Bohac, 1999).
Considering all the environments, Nitidulidae was the family that had the highest abundance and Scarabaeidae the highest richness.On the other hand, they did not obtain the same morph-species composition and dominance among the sites investigated.The vegetation density and structure of each ecosystem might have exerted an important influence on these guidelines.In P. elliottii and E. saligna monocultures, there possibly prevailed the species favoured by dry environments (Bohac, 1999).When the sites were analysed separately, there were some changes in this pattern.In native grassland, Staphylinidae was the most abundant group and in E. saligna and P. elliottii plantation and degraded area, Nitidulidae exhibited the highest richness.These differences in coleopteran composition and number reflect the different environmental requirements of the group based in the ecosystem structure and plant composition.
Some groups had higher richness or appeared only in poor soils, such as the Cucujidae family, which was observed only in P. elliottii plantation and in native grassland.Kevan (1999) associated the Carabidae diversity with the vegetation richness, but one morph-specie of Carabidae was found only in the degraded area.Steinborn and Meyer (1994) and Eyre et al. (2003) reported that soil type and soil moisture are factors of great influence in the communities of this family.Thus, Cucujidae and Cicindelinae might have obtained benefits from the lack of competition in those sites (Stork and Eggleton, 1992).Although Carabidae beetles are characterised by their great variety of species in the ecosystems (Lövei and Suderland, 1996), in the present study, they were not abundant.The low abundance of this group might have occurred due to its gregarious distribution in response to the pattern of resources availability (Paarmann et al., 2001).
Based on the lack of an asymptote in the rarefaction curve, it was observed that were not collected all beetles species in the study areas.The acquisition of such data requires a greater sampling effort (Soberón and Llorente, 1993).The use of other methodologies certainly would contribute with an asymptote.
The composition of beetles in each environment differs due to the needs, trophic level and behaviour of each group (Nouhuys, 2005).Most collected species belonged to families exclusively carnivorous as observed by Marinoni et al. (2001).These kinds of beetles depend on the diversity of animal products of an ecosystem, which in turn depend on other organisms and so on.Thus, Nouhuys (2005) suggested that in a community it is expected that organisms from higher trophic levels are more sensitive to environmental changes.Our study supports the author's idea, because only one exclusively carnivorous family (Carabidae -Cicindelinae) was recorded in the degraded area.
The herbivore and detritivore arthropods have received considerable attention, because they are organisms of high impact on ecosystems.Detritivores are responsible for increasing the flow of nutrients in the soil, while herbivores 2001).Moreover, the smaller similarity between native grassland and native forest should reflect differences in floristic composition and vertical structure of these ecosystems.Furthermore, Coleoptera composition in the degraded area differs from others, possibly because of the naked soil and the little food resources available to the beetles.
Epigean beetles show changes in the landscape and respond to environmental impacts through their close relationship with the processes that occur in soils and their susceptibility to interference in ecosystems.This study reinforces that the Coleoptera order may reflect the condition of each environment.Thus, it is necessary to identify taxa for each stress factor of the ecosystem and for different intensities of these factors.With this such knowledge, we will be able to carry out the appropriate management of environments with only a minor degree of disturbance.

Table 4 .
Number of individuals, abundance, diversity indexes, equability and richness estimates for the Coleoptera species collected in five different environments in Santa Maria, Rio Grande do Sul, Brazil, from October/2004 to October/2005.(N = number of individuals; S = number of morph-species; (H' = diversity index of Shannon-Wiener; 1-D = diversity index of Simpson; BP = index of dominance of Berger and Parker; J1 = Jackknife1; J2 = Jackknife2; and B = Bootstrap).

forest Native grassland
) and Nitidulidae (6).These three families, with Chrysomelidae and Cerambycidae were responsible for 73.3% (33) of the total richness.However, the most abundant family among the environments studied were Nitidulidae (n = 1.113), representing 61.4% of the total specimens sampled and Staphylinidae, representing 24.9% (n = 452) of the individuals sampled.Only these two families composed 86.4% of the total beetles collected.

Table 2 .
Number of individuals (N), morph-species (me), "singletons" (*) from each Coleoptera family collected in five different environments in Santa Maria, Rio Grande do Sul, Brazil, from October/2004 to October/2005.
Figure 2. Species accumulation curve for Coleoptera collected in five different environments in Santa Maria, Rio Grande do Sul, Brazil, from October/2004 to October/2005.wasidentified in Coleoptera richness captured between E. saligna monoculture and native forest, and between E. saligna monoculture and native grassland.

Table 5 .
Significance (t-test)to the Shannon-Wiener (H') and Simpson (1-D) diversity indexes calculated for the Coleoptera collected in five different environments in Santa Maria, Rio Grande do Sul, Brazil, from October/2004 to October/2005.*p < 0.05, ns = not significant.

Table 6 .
Significance for the abundance and richness of Coleoptera collected in five different environments in Santa Maria, Rio Grande do Sul, Brazil, from October/2004 to October/2005.*p < 0.05.

Table 7 .
The Spearman correlation based on the number of morph-species (matrix superior) and individuals (matrix inferior) of each Coleoptera family found in five different environments in Santa Maria, Rio Grande do Sul, Brazil from October/2004 to October/2005.*p < 0.05, ns = not significant.