Abstract
Caryocar brasiliense (Caryocaraceae) trees have a wide distribution in the Brazilian savanna. This plant is protected by federal laws and is untouched in deforested areas of the Brazilian savanna. This situation increases the damage to leaves, flowers, and fruits from chewing insects. We studied populations of herbivorous Lepidoptera, Coleoptera, and Hymenoptera and their natural enemies on C. brasiliense trees for three successive years during each season in the Brazilian savanna. Phytophagous insects were most abundant at the beginning of winter and with more species and diversity in the summer. Natural enemies were most abundant in the spring and in the winter and with highest species and diversity at end of the winter. Fruits bored by Carmenta sp. (Lepidoptera: Sesiidae) and Naupactus sp.3 (Coleoptera: Curculionidae) on the leaves were highest in the summer. Fruits scraped by Naupactus sp.1 and sp.2 were more numerous in the spring and summer, percentage of defoliation in the autumn. Apoptus sp. (Coleoptera: Elateridae) on the leaves in the spring, and leaf miners (Lepidoptera) in the winter. In the case of natural enemies, Crematogaster sp. (Hymenoptera: Formicidae) were most abundant on the leaves and in the flowers in the winter and spring, and on the fruits in the spring when new leaves and flowers are formed. The number of Zelus armillatus (Hemiptera: Reduviidae), Holopothrips sp. (Thysanoptera: Phlaeothripidae), and the complex of spiders were greater on the leaves in the winter; and Trybonia sp. (Thysanoptera: Phlaeothripidae) was most abundant on the leaves in the autumn.
Keywords
Introduction
About 23% of Brazil is covered with savanna or cerrado (Da Silva and Bates 2002). This region is characterized by high diversity of plants and insects (Bridgewater et al. 2004). This area is used for grain and cattle production (Aguiar and Camargo 2004), as well as reforestation (Zanuncio et al. 2002). Caryocar brasiliense Camb., 1828 (Caryocaraceae) trees have a wide distribution in the Brazilian savanna (Brandão and Gavilanes 1992; Bridgewater et al. 2004; Leite et al. 2006) and can reach over 10 meters in height and six meters in width of canopy (Leite et al. 2006, 2011a, b). The leaves of C. brasiliense are alternate, trifoliate and have high tricome density; its flowers are hermaphrodite but mostly cross are pollinated (Araújo 1995). The fruit is a drupe with 1-4 seeds, weighing 160g (fresh weigh) and with a volume of 315 cm³ (Araújo 1995; Leite et al. 2006). Fruit production is annual, and C. brasiliense blooms between July and September, or dry period, with fructification from October to January which correspond to the rainy season (Leite et al. 2006). The internal mesocarp of the fruit is rich in oil, vitamins, and proteins and contains many compounds of medicinal importance. Moreover, it is also used by humans for food, production of cosmetics, lubricants, and in the pharmaceutical industry (Araújo 1995; Segall et al. 2005; Ferreira and Junqueira 2007; Garcia et al. 2007; Khouri et al. 2007). This plant represents the main source of income of many communities of the region (Leite et al. 2006).
Coleoptera, Lepidoptera, and Hymenoptera are the most abundant orders of chewing insects in the Brazilian savanna (Pinheiro et al. 2002; Zanetti et al. 2003; Zanuncio et al. 2003; Leite et al. 2007, 2009, 2011b) and as Caryocar brasiliense trees are protected by federal laws are left in deforested areas which increases the damage to leaves, flowers, and fruits from chewing insects (according to communication reported by the collectors of C. brasiliense fruits). Insects that damage this plant are poorly known, in general, only at family level (Araújo 1995) and only a a single pest has been formally described (Freitas and Oliveira 1996; Oliveira 1997; Lopes et al. 2003; Boiça et al. 2004).
Our objective was to research the seasonality of herbivorous insects in the orders: Lepidoptera, Coleoptera, and Hymenoptera, their natural enemies, and the phenophases of C. brasiliense that influences them on C. brasiliense trees in savanna strictu sensu of Montes Claros in the state of Minas Gerais, Brazil.
Materials and Methods
We conducted this study in the Municipality of Montes Claros (43°55′7.3″W 16°44′55.6"S and 943 masl), in the state of Minas Gerais, Brazil, from June 2001 to June 2004 in a region with dry winter and rainy summer, which is classified as climate Aw: tropical savanna according to Köppen (Vianello and Alves 2000). The design was completely randomized with 25 replicates (25 trees) in savanna vegetation strictu sensu with dystrophic yellow red latosoil (sandy texture), and density of 13 C. brasiliense trees/ha (Leite et al. 2006). The savanna sensu strictu, a species-rich dense scrub of shrubs and trees, 8-10 m height, with a dense understory, is the more common Brazilian savanna than the grassland open forms (Ribeiro and Walter 1998; Durigan et al. 2002).
The distribution of Lepidoptera and Coleoptera defoliators and their natural enemies, the percentage of defoliation, the number of flowers damaged by Hymenoptera, the number of fruits scraped and bored by insects, were recorded in four fully expanded leaves, four bunches of flower, and four fruits of 25 C. brasiliense trees with 4.07 ± 0.18 m (average ± standard error) in height and crown width of 2.87 ± 0.13 m (Leite et al. 2006). This sampling was conducted in the morning by direct visual observation every month (Horowitz 1993). Defoliation was determined visually by estimating the percentage of leaf area loss on a scale from 0-100% with increments of 5% of the total areas removed (Sastawa et al. 2004; Mizumachi et al. 2006). Insects were collected with tweezers, brushes, or aspirators and preserved in vials with 70% alcohol for identification by taxonomists.
Ecological indices such as abundance of individuals, species richness, and diversity were calculated for the species identified in the samplings per tree in each season. The formula of Hill (Hill 1973) was used to calculate the diversity, and Simpson indices were used to calculate the abundance and richness of species (Townsend et al. 2006).
Given the lack of normality of the data Spearman's correlation was applied to data for phytophagous insects, their damage, natural enemies, flowers and fruits. Seasonal differences were tested with ANOVA with subsequent Tukey's test, using transformed data (√x + 0.5). In all tests, the significant level was 5%.
Results
Phytophagous insects were most abundant in winter and least abundant in spring and autumn (df = 72; F = 4.153; P = 0.009). There were more species and greatest diversity in the summer than in autumn (df = 72; F = 5.769; P = 0.001; df = 72; F = 6.422; P = 0.0006, respectively) (Table 1). Natural enemies were most abundant in spring and winter (df = 72; F = 5.761; P = 0.001). The fewest species were present in the autumn and the most in winter (df = 72; F = 15.127; P <<0.001). Diversity in this group was lowest in the autumn and highest in the winter (df = 72; F = 8.848; P = 0.00004) (Table 1).
Hill's diversity index, number of individuals and of species of natural enemies and herbivorous insects per tree of Caryocar brasiliense. Montes Claros, Minas Gerais State, Brazil. Autumn 2001 to Autumn 2004.
Means followed by the same letter (± standard error) in each row are not different by the test of Tukey (P < 0.05).
The number and percentage of bored fruits by Carmenta sp. (Lepidoptera: Sesiidae) (df = 72; F = 15.487; P << 0.001; df = 72; F = 34.982; P = 0.00003, respectively) and the number of Naupactus sp.3 (Coleoptera: Curculionidae) on the leaves (df = 72; F = 5.376; P = 0.002) were greatest in the summer (Table 2). Number and percentage of scraped fruits by Naupactus sp.1 and sp. 2 (df=72; F = 13.838; P<<0.001; df = 72; F = 33.892; P << 0.001, respectively) were greatest in the spring and summer, percentage of defoliation (df = 72; F = 19.489; P << 0.001) was highest in the autumn, number of Apoptus sp. (Coleoptera: Elateridae) on the leaves was highest in the spring (df = 72; F = 4.336; P = 0.00725), and number of leaf miners (Lepidoptera) (df = 72; F = 8.335; P << 0.001) was highest in the winter (Table 2).
Number of herbivorous insects and their damage per tree of Caryocar brasiliense. Montes Claros, Minas Gerais State, Brazil. Autumn 2001 to Autumn 2004.
Means followed by the same letter (± standard error) in each row are not different by the test of Tukey (P < 0.05).
In the case of natural enemies, Crematogaster sp. (Hymenoptera: Formicidae) had the highest abundance on the leaves (df = 72; F = 5.097; P = 0.003) and in the flowers (df = 72; F = 7.594; P = 0.0001) in the winter and spring, while it was on the fruits (df = 72; F = 5.697; P = 0.001) in the spring. The number of Zelus armillatus (Lep. and Servi, 1825) (Hemiptera: Reduviidae) (df = 72; F = 6.008; P = 0.001); Holopothrips sp. (Thysanoptera: Phlaeothripidae) (df = 72; F = 3.308; P = 0.02), and spiders such as Cheiracanthium inclusum (Hentz, 1847) (Miturgidae), Peucetia rubrolineata (Keyserling, 1877) (Oxyopidae), Anelosimus sp., Achaearanea hirta (Taczanowski, 1873) (Theridiidae), Gastromicans albopilosa (Simon, 1903), Chira bicirculigera (Soares and Camargo, 1948), Rudra humilis (Mello-Leitão, 1945), Thiodina melanogaster (Mello-Leitão, 1917) and Lyssomanes pauper (Galiano, 1945) (Salticidae), Dictyna sp. and sp.1 (Dictynidae); Tmarus sp. and sp.1 (Thomisidae), Argiope argentata (F, 1775), Gasteracantha cancriformes (L., 1758), Argiope sp., Parawixia sp. and sp.1 (Araneidae), and Anyphaenidae (df = 72; F = 6.460; P = 0.0006) were highest on the leaves in the winter; and Trybonia intermedius (Bagnall, 1910) and Trybonia mendesi (Moulton, 1933) (Thysanoptera: Phlaeothripidae) (df = 72; F = 2.853; P = 0.043) on the leaves in the autumn (Table 3).
Number of natural enemies on the leaves, flowers, and fruits per tree of Caryocar brasiliense. Montes Claros, Minas Gerais State, Brazil. Autumn 2001 to Autumn 2004.
Means followed by the same letter (± standard error) in each row are not different by the test of Tukey (P < 0.05).
Eunica bechina (Talbot, 1852) (Lepidoptera: Nymphalidae) (df = 72; r = 0.28; P = 0.04) and Naupactus sp.1+3 (r = 0.30; P = 0.03) showed a positive correlation with percentage of defoliation; Carmenta sp. showed a positive correlation with fruits bored (r = 1.00; P << 0.001) and Naupactus sp.2 with fruits scraped (r = 0.40; P = 0.007). Trigona spinipes (F., 1793) (Hymenoptera: Apidae) showed a positive correlation with damaged flowers (r = 0.80; P << 0.001).
The ants Crematogaster sp. and Pseudomyrmex termitarius (Smith, 1855) showed a positive correlation with flowers (r = 0.63; P = 0.0009), fruits bored by Carmenta sp. (r = 0.88; P = 0.0001) and those scraped by Naupactus sp.2 (r = 0.59; P = 0.02). The percentage of defoliation had a negative correlation with Crematogaster sp. (r = -0.40; P = 0.008); P. termitarius (r = -0.44; P = 0.004), Holopothrips sp. (r = -0.16; P = 0.047), Z. armillatus (r = -0.28; P = 0.024), spiders (r = -0.73; P = 0.003), and with the flower formation of C. brasiliense (r = -0.53; P = 0.0007). Lepidoptera leaf miners had a negative correlation with Z. armillatus (r = -0.27; P = 0.002), a barely significant relation with P. termitarius (r = -0.27; P=0.05) and a significant relation with spiders (r = -0.62; P = 0.01). Lepidoptera + Coleoptera had no correlation with P. termitarius (r = -25; P = 0.06).
Discussion
The higher abundance of phytophagous insects and their natural enemies in the winter is probably determined by the reduction in the number of C. brasiliense leaves available due to their gradual loss during the dry period and by the end of this season (Leite et al. 2006), which results in a concentration of insects per leaf. The largest species richness and diversity of the natural enemies in the winter probably indicate that their population depends on their prey and follows those of the phytophagous insects (Oberg et al. 2008; Venturino et al. 2008). In contrast, the higher species richness and diversity of phytophagous insects in the spring/summer indicated that these insects can be limited by the quality of available food i.e. autumn. For instance, the higher abundance of flowers and fruits on C. brasiliense during this season (Leite et al. 2006) can favor an increase in the diversity of chewing insects (Peeters 2002; Coley et al. 2006; Kursar et al. 2006) reducing the dominance of one or more species i.e. greater equitability.
Caryocar brasiliense loses its leaves in August/September with new ones in September (Leite et al. 2006). The ants Crematogaster sp. and P. termitarius were more abundant during the formation of new leaves and flowers at end of the winter, probably due to the nectaries of leaves and flowers (Oliveira 1997; Orivel and Dejean 2002; Oliveira and Freitas 2004). In addition, these ants visited fruits bored by Carmenta sp. and scraped by Naupactus sp.2, perhaps due to the presence of sugary secretions of damaged C. brasiliense fruits. Crematogaster ants may hinder Lepidoptera and Coleoptera colonizing alive trees, except for leaf miners, as observed for larvae of Hallonympha paucipuncta (Spitz, 1930) (Lepidoptera: Riodinidae), an endemic butterfly of the Brazilian savanna. Spatial distribution of larvae and tending ants were strongly aggregated, suggesting an influence of ants on oviposition or larval survival (Kaminski 2008; Sendoya et al. 2009).
Herbivorous Lepidoptera and Coleoptera appear to be negatively affected by the presence of the ants P. termitarius, and other predators such as Holopothrips sp., Z. armillatus as well as spiders, reducing defoliation. Mobile predators could respond to a local increase in vegetation complexity and the presence of an alternative prey and effectively suppress herbivores (Auslander et al. 2003). Ants can reduce E. bechina infestations as well as Edessa rufomarginata (De Geer, 1773) (Hemiptera: Pentatomidae), Prodiplosis floricola (Felt, 1908) (Diptera: Cecidomyiidae) and petiole gall insects on C. brasiliense (Hymenoptera: Chalcidoidea) (Freitas and Oliveira 1996; Oliveira 1997).
The association between ants and extrafloral nectaries may be responsible for the reduction of damage by chewing insects on the reproductive parts of C. brasiliense plants as observed for Crotalaria pallida (Leguminosae) (Guimarães et al. 2006). Species of Holopothrips sp. scratch leaves (Cavalleri and Kaminski 2007), create galls (Cabrera and Segarra 2008), or are predators (Almeida et al. 2006) as are the spiders, and bugs of the genus Zelus spp. and those of the subfamily Asopinae (Molina-Rugama et al. 1998). These predators are important in different ecosystems (Landis et al. 2000; Almeida et al. 2006; Mizzel 2007; Oberg et al. 2008; Venturino et al. 2008).
The chewing insects started damaging C. brasiliense trees in October at the beginning of the rainy season when there were lower populations of ants. The damage by wood borers and fruit scrapers was higher in the spring and in the summer (October-March) during the period of fruit formation (Leite et al. 2006). The formation of flowers of C. brasiliense before the period of higher abundance of chewing insects may reduce the probability of damage by the latter, suggesting phenological escape mechanism of the plants by defoliators (Sloan et al. 2007). Higher ant visitation to extrafloral nectaries can favor the production of flowers or fruits of this plant and reduce damage to C. brasiliense trees by T. spinipes. Sprouting of leaves and flower development before the rainy period is common in perennial plants of the Brazilian savanna (Almeida et al. 1998; Felfini et al. 1999; Pedroni et al. 2002; Almeida et al. 2006; Leite et al. 2006). This allows plants to increase photosynthetic area when the efficiency of predation by insects is lower. In addition, there is no heavy rain during this period, and the low quantity of leaves facilitates the ability of pollinators to find C. brasiliense flowers such as observed for Felfini et al. (1999) with Stryphnodendron adstringens (Mart.) Coville (Fabaceae).
Footnotes
Acknowledgements
To Drs. Antônio Domingos Brescovit (Instituto Butantã) (Aracnidae), Ayr de Moura Bello (Coleoptera), Ivan Cardoso Nascimento (EMBRAPA-ILHÉUS-Centro de Pesquisas do Cacau, CEPLAC, Itabuna, BA) (Formicidae), Paulo Sérgio Fiuza Ferreira (Hemiptera) (UFV) and Renata C. Monteiro (Thysanoptera) for the identification of the specimens. To Mr. Célio Barbosa, Oscar Madureira da Silva and Fabrício Maurílio Ruas for supplying climate data. To the Brazilian agencies "Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)", "Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG)" and "Secretaria de Ciência e Tecnologia do Estado de Minas Gerais". This manuscript has been edited by Asia Science Editing.
