Neotrop Entomol DOI 10.1007/s13744-015-0280-y
ECOLOGY, BEHAVIOR AND BIONOMICS
Leaf Beetle (Chrysomelidae: Coleoptera) Assemblages in a Mosaic of Natural and Altered Areas in the Brazilian Cerrado M PIMENTA1, P De MARCO Jr2 1
Lab de Ecologia, Embrapa Recursos Genéticos e Biotecnologia, Brasília, DF, Brasil Lab de Ecologia Teórica e Síntese - LETS, Univ Federal de Goiás, Goiânia, GO, Brasil
2
Keywords Bioindicators, beta diversity, community ecology, habitat heterogeneity, herbivorous insects Correspondence P De Marco, Lab de Ecologia Teórica e Síntese - LETS, Univ Federal de Goiás, Goiânia, GO, Brasil; [email protected] Edited by Wesley AC Godoy – ESALQ/USP Received 8 April 2014 and accepted 1 February 2015 * Sociedade Entomológica do Brasil 2015
Abstract In landscape mosaics, species may use different vegetation types or be restricted to a single vegetation type or land-use feature highlighting the importance of the interaction of species requirements and environmental heterogeneity. In these systems, the determination of the overall pattern of β-diversity can indicate the importance of the environmental heterogeneity on diversity patterns. Here, we evaluate leaf beetles (Coleoptera: Chrysomelidae) as habitat quality bioindicators in a system with varying intensities of human impacts and different phyto-physiognomies (from open field to forests). We collected 1117 leaf beetles belonging to 245 species, of which 12 species and 5 genus were considered possible bioindicators based on IndVal measures. Higher species richness was observed in forests and regenerating fields, and habitats with lower species richness included pastures, mines, and veredas. Natural fields, regenerating fields, natural cerrado, and forest had higher values of βdiversity. Bioindicator systems that include not only species richness and abundance but also assemblage composition are needed to allow for a better understanding of Chrysomelidae response to environmental disturbance.
Introduction Landscape mosaics may be defined by the existence of different vegetation types and/or land uses interconnected in a given landscape portion. In these systems, it is expected that the dynamic interaction of different patches, mainly through dispersal and habitat selection, strongly affects the structure of ecological communities (Halffter 1998, Law & Dickman 1998, Martin et al 2006). Considering the possible trade-off between dispersal and habitat selection in a given set of taxonomic-related species, we may expect to find in landscape mosaics both some “landscape species” (Sanderson et al 2002, Coppolillo et al 2004), which uses different vegetation types in the area, and habitat specialists, which may be restricted to a vegetation type or a land-use feature. Taken together, they will determine the overall pattern of beta diversity among the vegetation types and may indicate the importance of the environmental heterogeneity on diversity patterns. For instance, the Cerrado biome (tropical savanna)
in Brazil is an excellent example of complex mosaic landscapes with different, but integrated, vegetation units (Silva et al 2006). It ranges from open areas of “campo limpo” (grasslands) to the more dense cover of the “Cerrado sensu stricto,” “Cerradão” (forest savanna), and other vegetation types, as riparian forests and wetlands dominated by Buriti palms (“veredas”). Rapid changes in the natural communities of the Brazilian Cerrado, resulting from human-induced land-use changes, have greatly affected the biological diversity of this system (Klink & Machado 2005, Borges & Marini 2010). The majority of these changes is related to habitat fragmentation associated with expansion of agriculture frontiers, rangeland, and urban development (Sano et al 2001, Silva et al 2006). Other important environmental impacts reported for this system include alien species invasions, pollution of aquifers, ecosystem degradation, fire, carbon cycle disequilibria, and other climatic effects related to global change (Klink & Machado 2005). Those anthropogenic impacts can generate a
Marco et al
significant reduction in biodiversity not only due to habitat loss but also from the loss of habitat heterogeneity in a given landscape (Fischer & Lindenmayer 2007). The conversion of these natural habitats will maintain a diverse mosaic of these different vegetation types, but its effects on the maintenance of overall biodiversity are still poorly understood. Our current understanding suggests that pasturedominated areas of the Cerrado in central Brazil has lower level of fragmentation and are more capable of maintaining populations of endangered mammals than crop-dominated landscapes (Carvalho et al 2009). Bioindicator species can be used to evaluate the effect of human activities on the environment, determine regional distribution patterns of biodiversity, evaluate changes in the structure and function of communities and estimate the value for conservation of target areas (Staines & Staines 1998). Despite the existence of many criticisms about the use of bioindicators (Simberloff 1998, Rolstad et al 2002), several authors have claimed that indicator taxa should be effectively defined and applied to environmental management (Hilty & Merelender 2000, McGeoch 2007). In terrestrial environments, many insect groups have received considerable attention as potential bioindicators because they are characterized by: (1) high levels of diversity, (2) functional importance in ecosystems, (3) sensitivity to environmental changes and (4) relatively high levels of ease for sampling and identification (Pearson 1994, Silva et al 2010). In this study, we focus on the leaf beetle assemblage (Coleoptera: Chrysomelidae) which exhibits these general properties as bioindicator. In addition, leaf beetles have with a high ecological fidelity (Brown Jr. 1991, Staines & Staines 1998, Iannuzzi et al 2003). Chrysomelidae is one of the largest families of Coleoptera, both in terms of the number of species as well in abundance, with nearly 37,000 described species, possibly up to 50,000 species, arranged in 19 subfamilies and more than 2000 genera (Jolivet & Verma 2002). Larvae and adults of this family are herbivorous, feeding on different plant parts (leaves, fruits, seeds, pollen, and roots) (Jolivet & Verma 2002) mainly from the Asteraceae, Solanaceae, Convolvulaceae, Fabaceae, Malvaceae, Salicaceae, and Verbenaceae (Solorio & Rosales 2004). Some species have a great economic importance due to their role as herbivores of crops and their potential as control agents of invasive plants (Konstantinov & Vandenberg 1996, Jolivet & Verma 2002, DeLoach et al 2003). Leaf beetles exhibit narrow relationships with their host plants, and are probably among the most selective phytophagous insects (Jolivet 1992, Staines & Staines 1998), depending on a small set of plants to persist in a given system. For this reason, it is expected that they should respond to environmental alterations, especially anthropogenic interventions in natural habitats. Monitoring indicators and their complex relationships and responses to environmental impacts may allow us to use them as tools to for evaluating
environmental conditions (Bonvicino et al 1996, Staines & Staines 1998, Marinoni & Ganho 2006). The proportion of Chrysomelidae to total Coleoptera present in an area is considered an indicator of environmental disturbance (Hutcheson 1990, Linzmeier et al 2006, Marinoni & Ganho 2006). Herbivorous insects are expected to be more common in disturbed habitats, because of the resulting increase in young and more palatable plants. In contrast, predatory and detritivorous insects are expected to increase their representativeness in natural areas (Hutcheson 1990). The complex mosaic nature of Cerrado landscapes may represent a large variation in habitat suitability for leaf beetles and, thus, pose and interesting opportunity to understand their patterns of habitat association and assemblage organization. Following this reasoning, we aimed here to evaluate leaf beetles as habitat quality bioindicators in a system characterized by varying intensities of human impacts and different phyto-physiognomies. We test two general hypotheses about the effects of environmental alterations on beetle assemblages across habitat mosaic. First, we investigate whether species richness and β-diversity are determined by the degree of disturbance of a given habitat. We also sought to evaluate the main differences in species composition between habitats, which may help explain observed patterns. Second, we tested the hypothesis that spatial proximity provides a better explanation for similarity in species composition than habitat type.
Material and Methods Study area and sampling procedures The study was conducted in Barro Alto and Niquelândia municipalities in the state of Goiás, central Brazil, a region with nickel mining operations. The region belongs to the Cerrado biome, which is characterized by dry winters and rainy summers (“Aw” according to Köppen’s classification). Mean annual temperatures vary between 18 and 28°C (Ratter et al 1997); mean annual precipitation is 1500 mm and varies from 750 to 2000 mm (Ribeiro & Walter 1998). The area is also recognized as a priority region for biodiversity conservation for the Cerrado (MMA/SBF 2002) due to a high degree of endemism and also a high level of anthropogenic pressure. According to a recent evaluation of fragmentation patterns in the Brazilian Cerrado (Carvalho et al 2009), this region has a relatively high proportion of Cerrado remnants, mostly because of its geomorphology, characterized by numerous hills that prevent certain agricultural practices or other land uses that require mechanization. We sampled areas with different management practices that were representative of the different Cerrado physiognomies (Fig 1, Table 1). Native areas were Cerrado sensu stricto,
Leaf beetle in a Mosaic of Brazilian Cerrado
Fig 1 Location of sampled areas in Niquelândia and Barro Alto, Goiás, Brazil.
forests, vereda, and natural open fields; these vegetation types are described by Ribeiro & Walter (1998). Cerrado sensu stricto was characterized by short trees (very few trees higher than 10 m), a relatively open canopy, a low cover of herbaceous plants, and a high plant biodiversity. In the Cerrado landscapes, forests are usually found in narrow strips of riparian vegetation along rivers and creeks. The vereda was also associated with hydromorphic soils along narrow water courses and was vegetated predominantly by buriti palms (Mauritia flexuosa or Mauritia vinifera) associated with dense groups of shrubs and herbaceous vegetation. Grasslands (campo limpo) were represented by natural open fields without shrubs or trees. Managed areas included Eucalyptus urophylla (Myrtaceae) plantations, pastures, and mining areas. Pastures included areas that were previously devoted to cattle but were in the initial stages of succession. Mining areas were places where mining operations had ceased but its effects, including a higher presence of nickel and physical disturbances, were still affecting the system. The decreasing order of structural complexity of the nine vegetation types was forest, natural cerrado, grasslands, fields under succession, cerrado under succession, vereda, pasture, eucalyptus plantation, and mine (Table 1). This order also reflects the
degree of disturbance of the system with the first three areas representing systems with low or absent direct anthropogenic disturbance, the next two with intermediate status (successional areas), and the final three areas suffering from direct human use. The vereda areas studied here were considered disturbed since they are surrounded by very disturbed pastures. We collected leaf beetles in field samples in Niquelândia in November 2007, February and April 2008, and March, September, and October 2009, and in Barro Alto in May, November, and December 2009. All samples were collected in the rainy season to coincide with the peak abundance of Coleoptera (Pinheiro et al 2002, Santos et al 2003). We used Malaise traps (Townes 1972) due to their efficiency in sampling flying insects. Each Malaise trap was maintained in the field for seven consecutive days, after which the collecting vials were removed. After sorting, individual beetles were stored in 70% ethanol. Species identification was conducted to the lowest taxonomic level possible using standard taxonomic keys with the assistance of Mr. Ayr de Moura Bello and comparisons with specimens deposited in the collection at the Universidade Federal de Goiás and at the entomological collection Pe. Jesus Santiago Moure (DZUP). When species determination was
Marco et al Table 1
Description of sample sites in Niquelândia (NI) and Barro Alto (BA), Goiás, Brazil.
Locals
City
Latitude
Longitude
Area description
Fruta do Lobo Morro seco Mica Verde Pedra Verde Horto Aranha
NI NI NI NI NI
14°07′43.5″S 14°12′59.4″S 14°11′58.6″S 14°11′20.1″S 14°23′27.4″S
48°21′00″W 48°22′ 39.1″W 48°22′ 27.5″W 48°22′ 14.1″W 48°41′ 24.9″W
Natural open fields: a field, grassland without shrubs or trees
Morro Dois Corregos Morro Dois Corregos Morro Dois Corregos Casa de Pedra Fruta do Lobo Mica Verde
BA BA BA BA NI NI
15°4′56.60″S 15°4′44.50″S 15°3′45.00″S 15°6′32.80″S 14°11′20.1″S 14°12′27.4″S
49°0′ 38.60″W 49°0′ 6.70″W 48°59′ 59.70″W 49°1′ 45.30″W 48°22′ 14.1″W 48°21′ 58.1″W
Pedra Verde Morro seco Fruta do Lobo Horto Aranha
NI NI NI NI
14°11′23.6″S 14°12′58.3″S 14°07′27.6″S 14°16′33.1″S
48°22′ 01.8″W 48°22′ 30.5″W 48°21′ 46.6″W 48°43′ 42.1″W
Morro seco Mica vede Horto Aranha Fazenda Pedro Ferreira Fazenda Pedro Ferreira Casa de Pedra Área 1A/Anglo Horto Aranha Horto Aranha Horto Aranha Fruta do Lobo Horto Aranha/Mata de Galeria Mata da Barragem Área 1B/Anglo Fazenda Pedro Ferreira Morro Dois Corregos Fazenda Dirani Fazenda Pedro Ferreira Fazenda Pedro Ferreira Fazenda Nossa Senhora de Lourdes Área 2/Anglo Área 3/Anglo Área 1C/Anglo
NI NI NI BA BA BA BA NI NI NI NI NI NI BA BA BA BA BA BA BA BA BA BA
14°12′39.8″S 14°11′58.76″S 14°22′55.1″S 15°3′42.60″S 15°4′3.30″S 15°5′38.70″S 15°6′26.50″S 14°22′23.0″S 14°22′23.3″S 14°21′47.1″S 14°07′57.0″S 14°16′20.2″S 14°08′57.6″S 15°5′36.70″S 15°3′46.70″S 15°4′23.30″S 15°5′40.60″S 15°3′58.90″S 15°3′24.70″S 15°2′36.10″S 15°4′54.60″S 15°3′32.30″S 15°5′36.10″S
48°22′46.7″W 48°22′31.42″W 48°41′00″W 48°58′28.90″W 48°58′0.90″W 49°1′47.80″W 49°1′11.60″W 48°40′59.3″W 48°41′18.4″W 48°41′17.9″W 48°20′54.3″W 48°43′39.4″W 48°20′13.9″W 49°0′38.20″W 48°59′31.50″W 48°59′51.60″W 48°59′8.60″W 48°59′1.60″W 48°59′2.00″W 48°57′15.80″W 48°58′34.80″W 48°57′30.20″W 48°59′40.10″W
Área adjacente a Rodovia Go 437 Casa de Pedra Fazenda do Sr. Antônio
BA BA BA
15°4′38.50″S 15°5′46.70″S 15°5′22.00″S
48°54′44.30″W 49°2′18.80″W 48°58′48.50″W
impossible, individuals were assigned to morph species based on similar general characters from the taxonomic keys and other morphological traits. The specimens from this study were deposited in the zoology collection at Universidade Federal de Goiás.
Regeneration grasslands: fields areas where the vegetation was removed and they are regeneration of vegetal cover Regeneration cerrado: Cerrado sensu stricto where the vegetation was removed and they are regeneration of vegetal cover Cerrado sensu stricto: characterized by the presence of a high plant biodiversity with short trees (very few trees higher than 10 m) with a relative open canopy and a low cover the herbaceous plants
Eucalyptus urophylla ST Blake (Myrtaceae) plantation
Forest: patches of forest along river and streams
Pasture
Mine: mining were areas where mining operation ceased but its effects, including higher presence of nickel and physical disturbance are still affecting the system The vereda is a habitat associated to hydromorphic soils along narrow water courses with predominance of the buriti palms (Mauritia flexuosa or Mauritia vinifera) that emerging in the midst of more or less dense groups of shrubs and herbaceous vegetation
Data analysis We identified indicator species that corresponded to levels of habitat quality using the IndVal method (Dufrene & Legendre 1997), which combines species abundance (specificity) and
Leaf beetle in a Mosaic of Brazilian Cerrado
frequency (fidelity) in the samples. Species that have high fidelity to a given habitat type and high abundance are the best indicators. This method identifies indicators without any a priori or a posteriori classification of sites. Another important advantage of the method is that the IndVal metric was shown to be independent across species (McGeoch & Chown 1998). For the indicator species analysis, only species with more than two sampled individuals were analyzed, reducing the number of evaluated species from 245 to 67. The significance of the indicator value was tested using a Monte Carlo 10,000 randomization procedure. Additionally, statistically significant (p
ECOLOGY, BEHAVIOR AND BIONOMICS
Leaf Beetle (Chrysomelidae: Coleoptera) Assemblages in a Mosaic of Natural and Altered Areas in the Brazilian Cerrado M PIMENTA1, P De MARCO Jr2 1
Lab de Ecologia, Embrapa Recursos Genéticos e Biotecnologia, Brasília, DF, Brasil Lab de Ecologia Teórica e Síntese - LETS, Univ Federal de Goiás, Goiânia, GO, Brasil
2
Keywords Bioindicators, beta diversity, community ecology, habitat heterogeneity, herbivorous insects Correspondence P De Marco, Lab de Ecologia Teórica e Síntese - LETS, Univ Federal de Goiás, Goiânia, GO, Brasil; [email protected] Edited by Wesley AC Godoy – ESALQ/USP Received 8 April 2014 and accepted 1 February 2015 * Sociedade Entomológica do Brasil 2015
Abstract In landscape mosaics, species may use different vegetation types or be restricted to a single vegetation type or land-use feature highlighting the importance of the interaction of species requirements and environmental heterogeneity. In these systems, the determination of the overall pattern of β-diversity can indicate the importance of the environmental heterogeneity on diversity patterns. Here, we evaluate leaf beetles (Coleoptera: Chrysomelidae) as habitat quality bioindicators in a system with varying intensities of human impacts and different phyto-physiognomies (from open field to forests). We collected 1117 leaf beetles belonging to 245 species, of which 12 species and 5 genus were considered possible bioindicators based on IndVal measures. Higher species richness was observed in forests and regenerating fields, and habitats with lower species richness included pastures, mines, and veredas. Natural fields, regenerating fields, natural cerrado, and forest had higher values of βdiversity. Bioindicator systems that include not only species richness and abundance but also assemblage composition are needed to allow for a better understanding of Chrysomelidae response to environmental disturbance.
Introduction Landscape mosaics may be defined by the existence of different vegetation types and/or land uses interconnected in a given landscape portion. In these systems, it is expected that the dynamic interaction of different patches, mainly through dispersal and habitat selection, strongly affects the structure of ecological communities (Halffter 1998, Law & Dickman 1998, Martin et al 2006). Considering the possible trade-off between dispersal and habitat selection in a given set of taxonomic-related species, we may expect to find in landscape mosaics both some “landscape species” (Sanderson et al 2002, Coppolillo et al 2004), which uses different vegetation types in the area, and habitat specialists, which may be restricted to a vegetation type or a land-use feature. Taken together, they will determine the overall pattern of beta diversity among the vegetation types and may indicate the importance of the environmental heterogeneity on diversity patterns. For instance, the Cerrado biome (tropical savanna)
in Brazil is an excellent example of complex mosaic landscapes with different, but integrated, vegetation units (Silva et al 2006). It ranges from open areas of “campo limpo” (grasslands) to the more dense cover of the “Cerrado sensu stricto,” “Cerradão” (forest savanna), and other vegetation types, as riparian forests and wetlands dominated by Buriti palms (“veredas”). Rapid changes in the natural communities of the Brazilian Cerrado, resulting from human-induced land-use changes, have greatly affected the biological diversity of this system (Klink & Machado 2005, Borges & Marini 2010). The majority of these changes is related to habitat fragmentation associated with expansion of agriculture frontiers, rangeland, and urban development (Sano et al 2001, Silva et al 2006). Other important environmental impacts reported for this system include alien species invasions, pollution of aquifers, ecosystem degradation, fire, carbon cycle disequilibria, and other climatic effects related to global change (Klink & Machado 2005). Those anthropogenic impacts can generate a
Marco et al
significant reduction in biodiversity not only due to habitat loss but also from the loss of habitat heterogeneity in a given landscape (Fischer & Lindenmayer 2007). The conversion of these natural habitats will maintain a diverse mosaic of these different vegetation types, but its effects on the maintenance of overall biodiversity are still poorly understood. Our current understanding suggests that pasturedominated areas of the Cerrado in central Brazil has lower level of fragmentation and are more capable of maintaining populations of endangered mammals than crop-dominated landscapes (Carvalho et al 2009). Bioindicator species can be used to evaluate the effect of human activities on the environment, determine regional distribution patterns of biodiversity, evaluate changes in the structure and function of communities and estimate the value for conservation of target areas (Staines & Staines 1998). Despite the existence of many criticisms about the use of bioindicators (Simberloff 1998, Rolstad et al 2002), several authors have claimed that indicator taxa should be effectively defined and applied to environmental management (Hilty & Merelender 2000, McGeoch 2007). In terrestrial environments, many insect groups have received considerable attention as potential bioindicators because they are characterized by: (1) high levels of diversity, (2) functional importance in ecosystems, (3) sensitivity to environmental changes and (4) relatively high levels of ease for sampling and identification (Pearson 1994, Silva et al 2010). In this study, we focus on the leaf beetle assemblage (Coleoptera: Chrysomelidae) which exhibits these general properties as bioindicator. In addition, leaf beetles have with a high ecological fidelity (Brown Jr. 1991, Staines & Staines 1998, Iannuzzi et al 2003). Chrysomelidae is one of the largest families of Coleoptera, both in terms of the number of species as well in abundance, with nearly 37,000 described species, possibly up to 50,000 species, arranged in 19 subfamilies and more than 2000 genera (Jolivet & Verma 2002). Larvae and adults of this family are herbivorous, feeding on different plant parts (leaves, fruits, seeds, pollen, and roots) (Jolivet & Verma 2002) mainly from the Asteraceae, Solanaceae, Convolvulaceae, Fabaceae, Malvaceae, Salicaceae, and Verbenaceae (Solorio & Rosales 2004). Some species have a great economic importance due to their role as herbivores of crops and their potential as control agents of invasive plants (Konstantinov & Vandenberg 1996, Jolivet & Verma 2002, DeLoach et al 2003). Leaf beetles exhibit narrow relationships with their host plants, and are probably among the most selective phytophagous insects (Jolivet 1992, Staines & Staines 1998), depending on a small set of plants to persist in a given system. For this reason, it is expected that they should respond to environmental alterations, especially anthropogenic interventions in natural habitats. Monitoring indicators and their complex relationships and responses to environmental impacts may allow us to use them as tools to for evaluating
environmental conditions (Bonvicino et al 1996, Staines & Staines 1998, Marinoni & Ganho 2006). The proportion of Chrysomelidae to total Coleoptera present in an area is considered an indicator of environmental disturbance (Hutcheson 1990, Linzmeier et al 2006, Marinoni & Ganho 2006). Herbivorous insects are expected to be more common in disturbed habitats, because of the resulting increase in young and more palatable plants. In contrast, predatory and detritivorous insects are expected to increase their representativeness in natural areas (Hutcheson 1990). The complex mosaic nature of Cerrado landscapes may represent a large variation in habitat suitability for leaf beetles and, thus, pose and interesting opportunity to understand their patterns of habitat association and assemblage organization. Following this reasoning, we aimed here to evaluate leaf beetles as habitat quality bioindicators in a system characterized by varying intensities of human impacts and different phyto-physiognomies. We test two general hypotheses about the effects of environmental alterations on beetle assemblages across habitat mosaic. First, we investigate whether species richness and β-diversity are determined by the degree of disturbance of a given habitat. We also sought to evaluate the main differences in species composition between habitats, which may help explain observed patterns. Second, we tested the hypothesis that spatial proximity provides a better explanation for similarity in species composition than habitat type.
Material and Methods Study area and sampling procedures The study was conducted in Barro Alto and Niquelândia municipalities in the state of Goiás, central Brazil, a region with nickel mining operations. The region belongs to the Cerrado biome, which is characterized by dry winters and rainy summers (“Aw” according to Köppen’s classification). Mean annual temperatures vary between 18 and 28°C (Ratter et al 1997); mean annual precipitation is 1500 mm and varies from 750 to 2000 mm (Ribeiro & Walter 1998). The area is also recognized as a priority region for biodiversity conservation for the Cerrado (MMA/SBF 2002) due to a high degree of endemism and also a high level of anthropogenic pressure. According to a recent evaluation of fragmentation patterns in the Brazilian Cerrado (Carvalho et al 2009), this region has a relatively high proportion of Cerrado remnants, mostly because of its geomorphology, characterized by numerous hills that prevent certain agricultural practices or other land uses that require mechanization. We sampled areas with different management practices that were representative of the different Cerrado physiognomies (Fig 1, Table 1). Native areas were Cerrado sensu stricto,
Leaf beetle in a Mosaic of Brazilian Cerrado
Fig 1 Location of sampled areas in Niquelândia and Barro Alto, Goiás, Brazil.
forests, vereda, and natural open fields; these vegetation types are described by Ribeiro & Walter (1998). Cerrado sensu stricto was characterized by short trees (very few trees higher than 10 m), a relatively open canopy, a low cover of herbaceous plants, and a high plant biodiversity. In the Cerrado landscapes, forests are usually found in narrow strips of riparian vegetation along rivers and creeks. The vereda was also associated with hydromorphic soils along narrow water courses and was vegetated predominantly by buriti palms (Mauritia flexuosa or Mauritia vinifera) associated with dense groups of shrubs and herbaceous vegetation. Grasslands (campo limpo) were represented by natural open fields without shrubs or trees. Managed areas included Eucalyptus urophylla (Myrtaceae) plantations, pastures, and mining areas. Pastures included areas that were previously devoted to cattle but were in the initial stages of succession. Mining areas were places where mining operations had ceased but its effects, including a higher presence of nickel and physical disturbances, were still affecting the system. The decreasing order of structural complexity of the nine vegetation types was forest, natural cerrado, grasslands, fields under succession, cerrado under succession, vereda, pasture, eucalyptus plantation, and mine (Table 1). This order also reflects the
degree of disturbance of the system with the first three areas representing systems with low or absent direct anthropogenic disturbance, the next two with intermediate status (successional areas), and the final three areas suffering from direct human use. The vereda areas studied here were considered disturbed since they are surrounded by very disturbed pastures. We collected leaf beetles in field samples in Niquelândia in November 2007, February and April 2008, and March, September, and October 2009, and in Barro Alto in May, November, and December 2009. All samples were collected in the rainy season to coincide with the peak abundance of Coleoptera (Pinheiro et al 2002, Santos et al 2003). We used Malaise traps (Townes 1972) due to their efficiency in sampling flying insects. Each Malaise trap was maintained in the field for seven consecutive days, after which the collecting vials were removed. After sorting, individual beetles were stored in 70% ethanol. Species identification was conducted to the lowest taxonomic level possible using standard taxonomic keys with the assistance of Mr. Ayr de Moura Bello and comparisons with specimens deposited in the collection at the Universidade Federal de Goiás and at the entomological collection Pe. Jesus Santiago Moure (DZUP). When species determination was
Marco et al Table 1
Description of sample sites in Niquelândia (NI) and Barro Alto (BA), Goiás, Brazil.
Locals
City
Latitude
Longitude
Area description
Fruta do Lobo Morro seco Mica Verde Pedra Verde Horto Aranha
NI NI NI NI NI
14°07′43.5″S 14°12′59.4″S 14°11′58.6″S 14°11′20.1″S 14°23′27.4″S
48°21′00″W 48°22′ 39.1″W 48°22′ 27.5″W 48°22′ 14.1″W 48°41′ 24.9″W
Natural open fields: a field, grassland without shrubs or trees
Morro Dois Corregos Morro Dois Corregos Morro Dois Corregos Casa de Pedra Fruta do Lobo Mica Verde
BA BA BA BA NI NI
15°4′56.60″S 15°4′44.50″S 15°3′45.00″S 15°6′32.80″S 14°11′20.1″S 14°12′27.4″S
49°0′ 38.60″W 49°0′ 6.70″W 48°59′ 59.70″W 49°1′ 45.30″W 48°22′ 14.1″W 48°21′ 58.1″W
Pedra Verde Morro seco Fruta do Lobo Horto Aranha
NI NI NI NI
14°11′23.6″S 14°12′58.3″S 14°07′27.6″S 14°16′33.1″S
48°22′ 01.8″W 48°22′ 30.5″W 48°21′ 46.6″W 48°43′ 42.1″W
Morro seco Mica vede Horto Aranha Fazenda Pedro Ferreira Fazenda Pedro Ferreira Casa de Pedra Área 1A/Anglo Horto Aranha Horto Aranha Horto Aranha Fruta do Lobo Horto Aranha/Mata de Galeria Mata da Barragem Área 1B/Anglo Fazenda Pedro Ferreira Morro Dois Corregos Fazenda Dirani Fazenda Pedro Ferreira Fazenda Pedro Ferreira Fazenda Nossa Senhora de Lourdes Área 2/Anglo Área 3/Anglo Área 1C/Anglo
NI NI NI BA BA BA BA NI NI NI NI NI NI BA BA BA BA BA BA BA BA BA BA
14°12′39.8″S 14°11′58.76″S 14°22′55.1″S 15°3′42.60″S 15°4′3.30″S 15°5′38.70″S 15°6′26.50″S 14°22′23.0″S 14°22′23.3″S 14°21′47.1″S 14°07′57.0″S 14°16′20.2″S 14°08′57.6″S 15°5′36.70″S 15°3′46.70″S 15°4′23.30″S 15°5′40.60″S 15°3′58.90″S 15°3′24.70″S 15°2′36.10″S 15°4′54.60″S 15°3′32.30″S 15°5′36.10″S
48°22′46.7″W 48°22′31.42″W 48°41′00″W 48°58′28.90″W 48°58′0.90″W 49°1′47.80″W 49°1′11.60″W 48°40′59.3″W 48°41′18.4″W 48°41′17.9″W 48°20′54.3″W 48°43′39.4″W 48°20′13.9″W 49°0′38.20″W 48°59′31.50″W 48°59′51.60″W 48°59′8.60″W 48°59′1.60″W 48°59′2.00″W 48°57′15.80″W 48°58′34.80″W 48°57′30.20″W 48°59′40.10″W
Área adjacente a Rodovia Go 437 Casa de Pedra Fazenda do Sr. Antônio
BA BA BA
15°4′38.50″S 15°5′46.70″S 15°5′22.00″S
48°54′44.30″W 49°2′18.80″W 48°58′48.50″W
impossible, individuals were assigned to morph species based on similar general characters from the taxonomic keys and other morphological traits. The specimens from this study were deposited in the zoology collection at Universidade Federal de Goiás.
Regeneration grasslands: fields areas where the vegetation was removed and they are regeneration of vegetal cover Regeneration cerrado: Cerrado sensu stricto where the vegetation was removed and they are regeneration of vegetal cover Cerrado sensu stricto: characterized by the presence of a high plant biodiversity with short trees (very few trees higher than 10 m) with a relative open canopy and a low cover the herbaceous plants
Eucalyptus urophylla ST Blake (Myrtaceae) plantation
Forest: patches of forest along river and streams
Pasture
Mine: mining were areas where mining operation ceased but its effects, including higher presence of nickel and physical disturbance are still affecting the system The vereda is a habitat associated to hydromorphic soils along narrow water courses with predominance of the buriti palms (Mauritia flexuosa or Mauritia vinifera) that emerging in the midst of more or less dense groups of shrubs and herbaceous vegetation
Data analysis We identified indicator species that corresponded to levels of habitat quality using the IndVal method (Dufrene & Legendre 1997), which combines species abundance (specificity) and
Leaf beetle in a Mosaic of Brazilian Cerrado
frequency (fidelity) in the samples. Species that have high fidelity to a given habitat type and high abundance are the best indicators. This method identifies indicators without any a priori or a posteriori classification of sites. Another important advantage of the method is that the IndVal metric was shown to be independent across species (McGeoch & Chown 1998). For the indicator species analysis, only species with more than two sampled individuals were analyzed, reducing the number of evaluated species from 245 to 67. The significance of the indicator value was tested using a Monte Carlo 10,000 randomization procedure. Additionally, statistically significant (p
