The abundance and richness of Lepidoptera in the PNMRB varied seasonally and were higher in the dry season. The species composition may indicate how habitat fragmentation impacts the structure of habitats and the response of butterfly communities to changes, such as resource availability, food guilds, patterns of occurrence, and species composition (Pignataro et al., 2020). In addition to reducing the availability of host plants for larval stages, fragmentation of the restinga can reduce the availability of native fruit trees, affecting the butterflies throughout the year (Ramírez-Restrepo & MacGregor-Fors, 2017).
This temporal pattern with greater richness and abundance in the dry season can also be an escape strategy from natural enemies mainly because, during the rainy season, the availability of plant resources increases and, consequently, higher rates of parasitism and predation on the larvae as well (Morais et al., 1999). Despite the importance of natural enemies as biotic pressure, the sandbank has severe physical characteristics, due to open vegetation and sandy soil, which increases the incidence of solar radiation. During the rainy season, which corresponds to the summer in the tropics, the average temperature is higher, and there is more intense insolation, factors that increase the risk of desiccating eggs and interfering with the physiology of insects (Monteiro et al., 2007).
The higher number of species restricted to the dry season (n = 15) suggests they may be sensitive to increased temperature. Patterns of the temporal distribution of Lepidoptera were found in some areas of restinga, with higher peaks of abundance between March and April (end of the rainy season) and in the dry season, between July and September (Flinte et al., 2006; Monteiro et al., 2007), a period that precedes the growth of new leaves of host plants. However, most studies indicate that the butterfly species richness and abundance are higher in the rainy season (Brown Jr. & Freitas, 2000; Silva et al., 2007; Iserhard et al., 2013; Carneiro et al., 2014; Kishimoto-Yamada & Itioka, 2015), mainly due to a greater supply of food resources and oviposition sites and host plants with new and soft leaves, and greater nutritional value (Wolda, 1989; Grotan et al., 2012; Iserhard et al., 2013). The rainy season increases the availability of resources, but climate change has raised average temperatures in recent decades. Climate change can alter the interaction between butterfly species and their host plants and affect ecosystem functions (Ferro et al., 2014; Sousa et al., 2019). Changes in the temporal distribution of species include a shift in synchronization between larvae, host plants, butterflies, flowers, and natural enemies between dry and rainy seasons (Logan et al., 2003; Carneiro et al., 2014; Lourenço et al., 2019; Sousa et al., 2019).
Most of the collected species showed low population densities, which does not necessarily characterize them as rare species. Low population densities may be an effect of fragment isolation caused by the urbanization process, which affects the availability of food resources, host plants, or alters the microclimate of habitats in restinga (Brown Jr. & Freitas, 2002, De Souza & Guillhermo-Ferreira, 2015). Sousa et al. (2019) suggest that a high incidence of singletons and doubletons species may help explain the high species turnover along urban gradients.
Butterflies are mainly sampled in the hours of greatest activity, with peaks of abundance in the morning and the early hours of the afternoon (Basset et al., 2011; Ritter et al., 2011; Freire & Diniz, 2014; Pignataro et al., 2020). Thermoregulation allows these insects to control body temperature through behavioral or physiological mechanisms (Iserhard et al., 2018), such as selection of microhabitats, adaptations to absorb sunlight, and changes in activity or color patterns (Rossato et al., 2018; Pereira et al., 2020).
Studies point out that the definition of protected areas for biodiversity conservation needs to consider the effects of climate change on species distribution (Hannah, 2010; Mawdsley, 2011; Lemes & Loyola, 2013). Several species have shown migrations to colder regions, both in temperate and tropics regions, as a response to the increase in global temperature (Parmesan et al., 1999; Colwell et al., 2008; Ferro et al., 2014). Some studies have applied modern techniques to predict the distribution of invertebrates (Diniz-Filho et al., 2010; Ferro et al., 2014) and revealed that climate change is modifying species’ phenology patterns, distribution and abundance, including Lepidoptera species (Kocsis & Hufnagel, 2011). For the Arctiinae subfamily, a long-term study conducted from 1968 to 1999 in Great Britain showed an approximately 30% decrease in abundance and proportion of occupied places after 1984, coinciding with an increase in average temperature. Ferro et al. (2014) suggest that climate change should cause contractions of up to 100% in the distribution areas of several Arctiinae species that inhabit the Atlantic Forest. Local extinctions can cause cascades of extinction and consequent disruption of ecosystem services, with consequences for ecological interactions and economic losses due to insect infestation and pestilence (Griffith et al., 2009; Araújo et al., 2011; Ferro et al., 2014).
Urbanization is seen as a strong selective pressure that homogenizes habitats and restricts the geographic distribution and sharing of niches by rare and common species (McDonnell & Hahs, 2015; Iserhard et al., 2018). Urban centers tend to form heat islands, increasing local temperature, affecting the microclimate and the structure of habitats. However, butterfly species respond differently to disturbances, with their abundances dependent on specific distribution patterns and life histories (Ramírez-Restrepo & MacGregor-Fors, 2017; Santos et al., 2019).
The richness and abundance of species in the PNMRB varied between seasons and between Lepidoptera families, as observed in other studies (Ribeiro et al., 2010; Freire and Diniz, 2014; Carreira, 2015; Santos et al. 2017). Nymphalidae was the most frequent family, a pattern commonly reported in Neotropical fauna inventories (Martins et al., 2017; Pérez et al., 2019; Sousa et al., 2019). In Brazil, about 800 species of Nymphalidae have been registered. This family has wide morphological and behavioral variety and rapidly responds to environmental variations, making this group essential for environmental monitoring in the tropics (Bonfantti et al., 2009; Ribeiro et al., 2012; Henriques et al., 2019).
In Brazil, patterns with a high richness of Lepidoptera species are registered mainly in large and more conserved fragments of the Atlantic Forest. Approximately 241 species were registered (Monteiro et al., 2004) in the Restinga of Jurubatiba National Park (PNRJ), and 98 species in the Rio Doce State Park (PERD; Lourenço et al., 2019). However, a study carried out on eight urban forest fragments, where fragment size ranged between 1.4 to 103.4 hectares, in Curitiba city, south Brazil, were found 298 species of butterflies but the richness was not directly proportional to the fragment size (Pérez et al., 2019). This work allows us to infer those urban fragments, regardless of size, can serve as a refuge for many species of butterflies, making these areas fundamental for biodiversity conservation (Pérez et al., 2019). However, comparisons between studies need to be considered due to the size of the areas and sampling efforts, collection methods, degree of disturbance, and heterogeneity of the Atlantic Forest (Martins et al., 2017).
The diversity of Lepidoptera depends on the distribution pattern of the host plants (Bonebrake et al., 2010; Sousa et al., 2019). Restingas ecosystems have abiotic specificities such as high insolation, salinity, and endemism of plants that influence the colonization of herbivorous insects (Ferro et al., 2014). Bellaver et al. (2012) registered 64 lepidopteran species in an area of 1,000 hectares of restinga in southern Brazil, connected to forest fragments of Atlantic Forest. In the largest preserved area of restinga in Brazil, the PNRJ, with 14,860 hectares (148,600.000m²), 64 out of the 241 lepidopteran species sampled, were actively collected with a net (Monteiro et al., 2004). The present study was carried out in a small urban fragment of 32 hectares, separated from the PNRJ by a large urbanized area (Fig. 1). In this small and isolated restinga fragment, 40 species of butterflies were actively sampled. Such lower species richness may be related to the smaller area, availability of host plants, and the effects of urbanization that can directly affect the structure of the butterfly community, as has been shown in several studies (Clark et al., 2007; Konvicka & Kadlec, 2011; Lizée et al., 2012; Melliger et al., 2017; Ramírez-Restrepo & MacGregor-Fors, 2017; Merckx & Van Dyck, 2019; Tzortzakaki et al., 2019 Kuussaari et al., 2020). Pérez et al. (2019) highlighted the lack of connectivity and the isolation of the fragments as mechanisms affecting the structure of butterfly communities in urban ecosystems, in addition to demonstrating that small fragments can harbor a high diversity of butterflies.