COMMUNITY AND ECOSYSTEM ECOLOGY Gall-Inducing Insect Species Richness as Indicators of Forest Age and Health G. WILSON FERNANDES, 1 EMMANUEL D. ALMADA, 2 AND MARCO ANTONIO A. CARNEIRO 3 Ecologia Evolutiva and Biodiversidade/DBG, C P 486, ICB/Universidade Federal de Minas Gerais, UFMG, 31270 901 Belo Horizonte, MG, Brazil Environ. Entomol. 39(4): 1134Ð1140 (2010); DOI: 10.1603/EN09199 ABSTRACT The changes in the plant community that occur during the process of succession affect the availability of resources for the community of herbivores. In this study, the richness of galling insects was evaluated in restored stands of Amazonian tropical rain forest of several ages (0 Ð21 yr), as well as in areas of primary forest in Brazil. The richness of gallers increased with the age of the restored stands. Fifty-eight percent of the variation in the richness of galling insects was explained by forest stand age, but an increase in richness was observed at intermediate stages of succession. The greatest similarity among groups was found between the initial successional stages and intermediate ones. The results indicate a recovery of both host plants and insect community and that succession directly affects the richness and composition of these herbivores. KEY WORDS Amazon, biodiversity, galling insects, mining, restoration ecology The change in plant species composition during the process of succession results in the replacement of the communities through time. While a particular climax community may not be reached, the species number has been argued to increase throughout the succes- sional process (Capers et al. 2005, Begon et al. 2006). In late successional stages, the number of trees and rare species increase, whereas the number of shrubs and lianas decrease (Guariguata et al. 1997, Pen ˜ a- Claros and de Boo 2002, Capers et al. 2005). Alterna- tively, many studies have shown that, during succes- sion, plant species achieve higher richness at intermediate stages (Rosenzweig 1995; Gaston 1992, 2000; Molino and Sabatier 2001). A quadratic relation- ship between plant species richness is expected, as reported by Grime (1973) and Connell (1978), who named it the intermediate disturbance hypothesis. In- deed, this hypothesis is supported by the Þndings that disturbed or managed systems present higher plant diversity than intact forests (Nagaike 2002, Webb and Sah 2003). Variation in plant community composition and structure strongly inßuence the trophic levels above, such as herbivores and their natural enemies. Numer- ical and qualitative responses of insects are expected to parallel that of the plant community (Price 2005, Novotny et al. 2006). The diversity of several taxa of free-feeding insect herbivores shows a positive cor- relation with succession stage (Parrish and Bazzaz 1982, Davies et al. 1999, DeWalt et al. 2000, Kennard 2002, Kalacska et al. 2004, Hilt and Fiedler 2005, Bar- low et al. 2007). The changes in the community struc- ture of herbivore insects in different successional stages may be driven by disturbances that help main- tain community composition across a wide variety of ecosystems. A simple increment or decrement of the frequency of disturbance can result in drastic changes in the community structure (Hobbs and Mooney 1991, Wootton et al. 1996, Floren and Linsenmair 2001, Foggo et al. 2001, Elderd 2006), although some taxa have been shown to be more or less susceptible to disturbance (Kruess and Tscharntke 2000, van Nou- huys 2005). The response of insect communities to changes in plant structure and composition has been widely stud- ied, especially in restoration programs (Jansen 1997, McGeogh 1998, Andersen et al. 2002, Nakamura et al. 2003, Moreira et al. 2007). Among the insects, some preference has been found for the use of Hymenop- tera, Lepidoptera, and Coleoptera as indicators of nat- ural regeneration of forests, perhaps because of their high diversity, habitat Þdelity, and taxonomical knowl- edge (Rodrõ ´guez et al. 1998, Ratchford et al. 2005). For instance, higher richness of these groups was reported on natural forest remnants than in reforested euca- lyptus areas in Australia (Cunningham et al. 2005). Grimbacher and Catterall (2007) reported that forest structure had a strong effect on the community of coleopterans, whereas forest age and distance from remnant primary forest was of second order of importance. 1 Corresponding author, e-mail: gw.fernandes@gmail.com. 2 Nu ´ cleo de Estudos e Pesquisas Ambientais, Universidade Estadual de Campinas, 35400 000, Campinas SP, Brazil. 3 Instituto de Cie ˆ ncias Exatas e Biolo ´ gicas, Universidade Federal de Ouro Preto, 35400 000, Ouro Preto, MG, Brazil. 0046-225X/10/1134Ð1140$04.00/0  2010 Entomological Society of America