Insects have important role in the ecosystems as herbivores, pollinators, nutrient cyclers, regulating populations of other organisms and feeding on or serving as food for other species (Losey & Vaugan 2006; Losey & Vaugan 2008; Lemos et al. 2003). The Lepidoptera are among the largest groups of insects with great diversity and different ecological services (Summerville & Crist 2008; Pereira et al. 2008). However, climate change can reduce the number and diversity of Lepidoptera (Gimesi et al. 2012).

The insect species richness in tropical rainforest ecosystems is high and its evenness is low, with a few abundant ones and mostly rare species (Hill & Hamer 1998; Nummelin 1998). Tropical rainforests have more species than temperate habitats, but their abundance and distribution in these communities need to be better studied (DeVries et al. 2009). Local and regional species richness may improve understanding of the relative influence of regional and local processes on community structure (Ricklefs 1987).

More homogeneous environments such as forest plantations may have continuous modifications affecting the composition and structure of animal communities due to interactions between biotic and abiotic factors (Ganho & Marinoni 2005). Furthermore, species diversity may be related to the effective-functioning of the ecosystem (Loreau et al. 2001). Therefore, studies on the abundance of species in natural communities and their variation in time and space are important.

The increase of eucalyptus plantations in Brazil facilitates a larger number of native Lepidoptera to adapt to these exotic plant species, such as Eupseudosoma aberrans Schaus and Eupseudosoma involuta Sepp (Arctiidae), Automeris sp. Walker and Eacles imperialis Walker (Saturniidae), Oxydia vesulia Cramer, Sabulodes caberata Guenée and Thyrinteina arnobia (Stoll) (Geometridae) (Guedes et al. 2000; Zanuncio et al. 2006). Lepidoptera pests like T. arnobia are important to forestry because of population outbreaks and consequent damage in eucalyptus plantations in Brazil (Zanuncio et al. 2006).

The occurrence of Lepidoptera pests and species with the potential to acquire pest status has been defined in monitoring programs by forestry companies (Pereira et al. 2001; Zanuncio et al. 2001; Zanuncio et al. 2006) in different regions using light traps (Kitching et al. 2000; Osaki et al. 2011).

Lepidoptera can be dependent on plants in a community and most of them have some degree of host specificity (Dyer et al. 2007; Harrison & Berenbaum 2013). The spatial occurrence of these species may be limited to environments suitable for their development, which may have implications in monitoring and managing their populations.

The objective of this study was to evaluate and to compare the spatial distribution of species diversity of Lepidoptera across eucalyptus plantations in the Amazon region of Brazil.

Materials and Methods

Alpha Diversity

Richness is a measure of the number of species in an area, but may present systematic flaws (Ricklefs 1996). This index has been calculated with rarefaction methods to obtain the «expected» number of species according to the number of sampled individuals (Ramalho & Silva 2002). The first order Jackknife procedure is one of the more accurate nonparametric methods to express the richness of a community and to correcting its shortcomings (Palmer 1991). Comparisons between sampling efforts were conducted by correction using randomized cumulative species curves based on samples, after 100 randomizations calculated by the software Estimates version 8.2 (Colwell 2009), using the Statistica 9.0 (Statsoft 2009) analysis to compare the results.

Table 1.

Characterization of sampling sites of lepidopteran defoliators on eucalyptus in Caracuru, Pacanari and Ponte Maria, Pará State and Felipe, Amapá State, Brazil.

A species accumulation curve for each sampled site with the Jackkinfe estimator and values estimated at a confidence interval of 95% (Colwell & Coddington 1994) was made. The estimate by Jackknife is:

where: S_obs = is the total number of observed species in all samples; m = total number of samples.

Beta Diversity

The beta diversity indices with presence and absence data are still controversial (Wilson & Shmida 1984; Harrison et al. 1992) and with a wide variety of indexes. The approach of this study was based on dissimilarity measures estimated by the Jaccard index. The index is defined as:

where: Dij = is the measure of dissimilarity between the squares; i and j = the number of common species or genera; S. and S. = species richness of species or genera in each area.

The mean dissimilarity per point compared to the others was used as a measure of beta diversity. High values of this measure mean a place with very particular species composition and preservation.

Lepidopterans collected were divided in groups I, II and III, which consisted of primary pests, secondary pests and those with undefined importance to eucalyptus, respectively (Zanuncio et al. 2006).

Results

For the eucalyptus plantations in Caracuru, Felipe, Pacanari and Ponte Maria, the respective numbers of lepidopteran individuals collected were 8,381, 1,010, 3,459 and 1,015 for the primary pests; 666, 324, 249 and 351 for the secondary ones and 3,195, 3,558, 4,034, and 5,615 for those species with undefined importance for the eucalyptus. The estimated value of the richness (Fig. 1), calculated by the Jackknife index, ranged from 10 to 11 species of group I; 9 to 10 species of group 11 and 378 to 409 species of group III, between the sampling sites, without differences per group between areas.

The species accumulation curves showed stability for the primary pests group between the 15th and 35th sampling and for the secondary ones in the 70th, 45th, 35th and 20th samplings in Caracuru, Felipe, Pacanari and Ponte Maria, respectively. The accumulation curves for the species with undefined importance to eucalyptus differed from those of other groups for the 4 sampling areas (Fig. 2). At the beginning of the sampling, the estimated species richness of this group ranged from 50 to 100 species and increased with a steep slope curve until approximately the 60th sample, when it slowly increased and tended to stabilize in the 4 areas.

The primary pest species were the same in all areas, and therefore, the beta diversity of this group was zero. Species diversity of group II was low (Table 2, Fig. 1) and similar in Caracuru, Pacanari and Ponte Maria, differing only in Felipe. Group III showed the highest beta diversity, possibly due to major differences in species richness, and the region of Felipe again showed the most difference between the areas (Table 2).

The dissimilarity of secondary pests differed between the areas except in Felipe, which differed by 5.88% from the others (Table 3). The species with undefined importance to eucalyptus plantations showed lower dissimilarity between Caracuru and Ponte Maria (14.88%) and with greater values between Felipe and Pacanari (21.64%) (Table 3).

Discussion

The primary and secondary pest species collected were those commonly found in eucalyptus plantations in different regions of Brazil (Pereira et al. 2001; Zanuncio et al. 2003; Freitas et al. 2005; Zanuncio et al. 2006). This indicates that they are associated with alternative hosts, but have adapted to cultivated eucalyptus (Freitas et al. 2005; Pereira et al. 2009). Furthermore, forest plantations have lower diversity of plants and therefore can have lower numbers of natural enemies.

Fig. 1.

Estimated richness (Jackknife) for lepidopteran primary pests (A), secondary pests (B) and species with undefined importance to eucalyptus plantations (C) in Caracuru, Pacanari and Ponte Maria, Pará State and Felipe, Amapá State, Brazil. (The bars represent the 95% confidence interval).

The abundance of lepidopterans may be influenced by habitat type (Robinson et al. 2012); however, all of the areas studied were adjacent to E. urophylla plantations, which may have contributed to the herbivore and host specificity, as reported for other Brazilian regions (Pereira et al. 2001; Oliveira et al. 2005; Zanuncio et al. 2006).

Fig. 2.

Estimation of richness of lepidopteran species with undefined importance to Eucalyptus urophylla plan tations in Ponte Maria (A), Pacanari (B) and Caracuru (C), Pará State, and Felipe (D), Amapá State, Brazil.

The lower numbers of individuals of group I in Felipe and Ponte Maria and of species richness of the group III in Felipe are possibly related to the greater proximity of eucalyptus plantations to the native forest. This agrees with the lower number of lepidopteran pests found in E. cloeziana plantations with strips of native vegetation in Minas Gérais State, Brazil (Zanuncio et al. 1998b). Also, geographical factors such as area, distance and connectivity between sites affect the richness of lepidopteran species as reported in temperate environments (Bila et al. 2013). Furthermore, this can occur with natural enemy populations, suggesting that maintenance of areas or strips of native vegetation near eucalyptus plantations can be a reservoir of natural enemies such as hymenopteran parasitoids (Zanuncio et al. 2001). Protected and in post-successional stage forests may exhibit higher diversity and numbers of insects, because such areas have greater resources and varieties of microclimates (Merckx et al. 2012). This favors the migration of natural enemies to eucalyptus plantations where they can reduce pest populations (Zanuncio et al. 1998a; Teja & Roland 2004).

Table 2.

Beta diversity indices for lepidopteran primary pests (I), secondary pests (II) and species with undefined importance for eucalyptus (III) in the regions of caracuru, pacanari and ponte maria, Pará State And Felipe, Amapá State, Brazil.

The asymptotic curves of species accumulation showed that all primary and secondary pest species present in plantations were represented in all areas. The stabilization of these curves, over shorter periods, is attributed to the adaptation of these species to eucalyptus, as observed in other Brazilian regions (Freitas et al. 2005; Pereira et al. 2009). The cutting of native trees, whether managed or not, may favor lepidopteran pest species that feed on alternative hosts (Summerville et al. 2013) to migrate to the plantations, which justify the conservation of these areas.

Table 3.

Dissimilarities of alpha diversities calculated by the Jaccard Index for lepidopteran secondary pests (II) and species with undefined importance to eucalyptus (III) in the regions of caracuru, Pacanari and Ponte Maria, Pará State and Felipe, Amapá State, Brazil.

The lack of stabilization of species accumulation curves in short term for the species of group III may be caused by the low variability of plants and scarcity of alternative sources of food for lepidopterans in the plantation understory in the first months of planting. Moreover, it can be attributed to the smallest canopy volume of young plantations, because species richness may be positively correlated with tree cover (Houlihan et al. 2013).

The group III species may not feed on eucalyptus plants and, therefore, require other hosts, which would explain the greater effort, and the sampling period to stabilize the curve for this group. Rare lepidopteran species may be associated with plant species in areas with forest dominance or underbrush vegetation (Meehan et al. 2013). The high species diversity of this group may be related to vegetation diversity in native forests near the plantations, whose insect fauna can migrate to forest plantations (Zanuncio et al. 2001).

The higher richness and abundance of lepidopteran species of group III can be explained by the positive correlation with the size and distance of the native forest fragment from the eucalyptus plantation (Rosin et al. 2012). Areas with such plantations were surrounded by the Amazon forest with large trees, mainly in Caracuru and Pacanari, which may have increased the species diversity of group III at these sites.

The similarity between areas shows that a lower sampling effort was required to achieve the stabilization of species accumulation curves of the primary and secondary pests. Furthermore, the low dissimilarity between sampling sites indicates that the monitoring methodology was appropriate to represent the fauna of lepidopteran pests of this group of eucalyptus. This type of survey is important for the development of programs of integrated management aiming to control these insects (Maalik et al. 2013). The eucalyptus plantations, even at 50 km away from each other, have biotic and abiotic conditions without high variations, which contribute to the similar results between them.

The knowledge of the population dynamics and estimation of species richness are important strategies for monitoring defoliator lepidopterans of eucalyptus. The survey of primary and secondary pests of this plant can be concentrated in shorter periods to streamline the management and to reduce the costs of monitoring and control. This can be done because the primary and secondary pests are the same or with very low variations between the sampled sites.