Termites as Bioindicators of Habitat Quality in the Caatinga, Brazil: Is There Agreement Between Structural Habitat Variables and the Sampled Assemblages?

ECOLOGY, BEHAVIOR AND BIONOMICS

Termites as Bioindicators of Habitat Quality in the Caatinga, Brazil: Is There Agreement Between Structural Habitat Variables and the Sampled Assemblages?

W de F Alves^{I, II}; AS Mota^II; RAA de Lima^II; R Bellezoni^II; A Vasconcellos^II

^IPPG em Ciências Biológicas;

^IIDepto de Botânica, Ecologia e Zoologia, Centro de Biociências, Univ Federal do Rio Grande do Norte, Natal, RN, Brasil

^{Correspondence}

ABSTRACT

The composition of termite assemblages was analyzed in three caatinga sites of the Estação Ecológica do Seridó, located in the municipality of Serra Negra do Norte, in the state of Rio Grande do Norte, Brazil. These sites have been subjected to selective logging, and cleared for pasture and farming. A standardized sampling protocol for termite assemblages (30h/person/site) was conducted between September 2007 and February 2009. At each site we measured environmental variables, such as soil pH and organic matter, necromass stock, vegetation height, stem diameter at ankle height (DAH) and the largest and the smallest crown width. Ten species of termites, belonging to eight genera and three families, were found at the three experimental sites. Four feeding groups were sampled: wood-feeders, soil-feeders, wood-soil interface feeders and leaf-feeders. The wood-feeders were dominant in number of species and number of encounters at all sites. In general, the sites were not significantly different in relation to the environmental variables measured. The same pattern was observed for termite assemblages, where no significant differences in species richness, relative abundance and taxonomic and functional composition were observed between the three sites. The agreement between composition of assemblages and environmental variables reinforces the potential of termites as biological indicators of habitat quality.

Keywords: Isoptera, semi-arid, feeding group, abundance, environmental variable

Introduction

The Caatinga covers an area of around 800,000 km² in Northeast Brazil. Its climate is semiarid, with high rates of potential evapotranspiration (1500-2000 mm/year) and low rainfall (300-1000 mm/year), often concentrated over a three-to-five month period (Sampaio 1995). This biome was traditionally stigmatized as being homogeneous and characterized by low species diversity and endemism. However, the results of recent studies contradict this belief and classify Caatinga as a very heterogeneous biome with a biodiversity composed of several endemic species (Leal et al 2003). Even though it is biologically significant, the Caatinga is one of Brazilian biomes that has been most altered by human activities (Leal et al 2005).

The effects of environmental degradation in the Caatinga on the biological communities are practically unknown. Thus, analysis of the effects of anthropic disturbance on all the taxa of a caatinga community poses practical obstacles. Termites exhibit several characteristics that underscore their potential as a bioindicator taxon of habitat quality suitable for use in monitoring programs in tropical ecosystems (Brown Jr 1997, Bandeira et al 2003, Vasconcellos et al 2010). Among the main attributes that justify their characterization as biological indicators of habitat quality are: (i) widespread geographic distribution; (ii) high abundance; (iii) low locomotor capacity; (iv) functional importance; (v) ease of sampling using standard protocols; (vi) short response time to anthropic disturbances; and (vii) it is taxonomically treatable (Brown Jr 1997, Constantino 2005).

In some arid and semiarid ecosystems, the termite pedologic activity significantly increases soil water retention, a fact that directly reflects in plant structure and local primary productivity. Thus, these insects are considered essential for maintaining the structural and functional integrity of these ecosystems (Holt & Conventry 1990, Whitford 1991).

The aim of the present study was to analyze the structure of termite assemblages at three sites with different land use histories. Eleven environmental variables were measured in each of these sites to determine if type of land use could result in different environmental characteristics and if these possible differences between sites could significantly affect the composition of termite assemblages. If termites are in fact good indicators of environmental quality, they would be expected to reflect structural variations in the habitat, that is, if sites have different environmental parameters, it would follow that termite assemblages would also exhibit different compositions, reflecting habitat parameters, as observed in humid tropical forest sites (Eggleton et al 2002, Jones et al 2003) and another caatinga site (Vasconcellos et al 2010).

Material and Methods

The study was conducted at the Estação Ecológica do Seridó (ESEC-Seridó), located in the municipality of Serra Negra do Norte, state of Rio Grande do Norte, Brazil (6°33'30" to 6°37'00"S and 37°14'30" to 37°16'30"W). ESEC-Seridó occupies 1166.4 ha and was formerly used for cattle breeding and logging. Its dominant vegetation is hyper-xerophilous, arboreal-shrubby caatinga. The local climate is hot and semiarid, with mean annual temperature of 27.5°C and mean annual rainfall of 744.7 mm. The highest rainfall occurs between February and May. With respect to soils, there is a predominance of luvisols, eutrophic inceptisols and vertisols, in addition to sparsely scattered rocky outcroppings (Ministério da Agricultura 1978).

The study was conducted between September 2007 and February 2009 in three caatinga sites, chosen based on reports of residents about the history of each site and the type of current land use. The sites were inserted in a matrix composed by caatinga vegetation and pasture, and the distance between them ranged from 2 km to 6 km. Thus, the sites were arbitrarily categorized into: (i) Site A₁, with secondary vegetation, subjected to different levels of cattle breeding for the last 30 years; (ii) Site A₂, with secondary vegetation and regeneration time similar to that of A₁, currently bordered by dirt roads. Furthermore, until mid-2006 part of its site was used for growing elephant-grass (Pennisetum sp.); (iii) Site A₃, located outside the ESEC. It shows signs of considerable anthropic disturbance and is sporadically used for cattle breeding.

Termite sampling

A rapid standardized sampling protocol of termite biodiversity (30h/person/site) was applied in each of the sites. This protocol consisted of delimiting six 65m transects. Five 5 x 2m parcels, spaced 10 m apart, were marked in each of the transects, for a total of 300 m² sampled per site. The termites were collected in several microhabitats (soil, tree trunks and branches, active and abandoned nests, undergrowth, under tree bark, dead roots, etc.). Species identification was based on taxonomic studies conducted by Constantino (1998) and species comparison relied on specimens from the Isoptera collection of the Bioscience Center, Universidade Federal do Rio Grande do Norte.

Feeding groups

The species were categorized into four food groups, according to in situ observations, mandibular morphology of the workers (Deligne 1966), analysis of intestinal content (Bandeira 1989) and studies carried out in ecosystems of Northeast Brazil (Bandeira et al 2003, Sena et al 2003, Mélo & Bandeira 2004, Vasconcellos et al 2008, 2010). The four feeding groups were: (i) wood-feeders, which feed on wood at different stages of decomposition and have no silt or sand particles in their intestine; (ii) soil-feeders, which feed predominantly on organic particles mixed with soil. The intestinal content of the specimens has a large amount of silt and sand mixed with the organic matter ingested; (iii) wood/soil-feeders, which consume predominantly, wood in advanced stages of decomposition, usually mixed with mineral soil. They generally carry soil into the trunk that is being consumed; and (iv) leaf-feeders, which feed mainly on dry leaves found in the undergrowth.

Nest density

To estimate termite nest density, five 100 x 20m parcels were delimited in each study site, for a total of 1 ha per site. All the nests with active populations were counted, irrespective of size. Vasconcellos et al (2008) showed that the composition of termite nests may also provide information on the disturbance level in a particular area.

Abiotic and vegetation variables

The following environmental variables related to the soil of each site were measured: (i) granulometry (sand-silt-clay ratio); (ii) organic matter in the soil; (iii) pH, using a pHmeter (Instrutherm pH 2500); and iv) necromass stock of 13 transects (four in A₁, five in A₂ and four in A₃) that contained five 0.5 x 0.5m parcels spaced 10 m apart.

The vegetation physiognomy of each site was described for the five 20 x 10m parcels, totaling 1000 m²/site. In each parcels all the live individuals with diameter > 5 cm that were 10 cm above soil level were sampled. The following tree parameters were then measured: (i) diameter at ankle height (DAH); (ii) tree height; (iii) largest crown width; and (iv) smallest crown width.

Analyses

Once biological data, such as species richness, feeding groups, taxonomic groups and number of encounters of a given species, were obtained, the latter considered an indirect measure of relative abundance, the ecological measures of frequent use in the environmental monitoring were calculated using Primer 5 and Estimated softwares (Clarke & Warwick 2001, Colwell 2005). Species richness estimated for each site was calculated using nonparametric Jacknife 1 and Chao 2 estimators, considered two of the best nonparametric estimators of species richness (Walther & Moore 2005). Sample-based species accumulation curves were also built to compare species richness. Accordingly, 500 randomizations were performed from the initial data collected in each parcels.

The environmental variables of the sites were compared using ANOVA, with Turkey's test a posteriori, if there was a significant difference between the sites. The number of species found per transect inside each site was compared using the Kruskal-Wallis nonparametric test and the number of feeding and taxonomic groups found in each site were compared using the chi-square test. All of these tests were conducted using Statistica 7.1 software (StatSoft 2005).

Principal component analysis (PCA), based on the structural parameters of the three study sites, was used to determine if there was spatial segregation between the sites analyzed (Clarke & Warwick 2001). Abiotic and vegetation data were normalized before PCA was carried out. In PCA, the variables silt and smallest crown width were excluded because of their high correlation (> 0.95) with the variables sand and largest crown width, respectively. This procedure is recommended when there is a strong correlation between the variables analyzed, because no information is lost in the analysis (Clarke & Warwick 2001). ANOVA with the scores of the first PCA component was performed to check for the existence of a significant difference between the environmental attributes measured in the sites.

Results

Ten species of termites were found in the three study sites, seven in A₁, six in A₂ and seven in A₃. Five species occurred simultaneous in the three sites. Two were exclusive from A₁, one from A₂ and two from A₃ (Table 1). The species Anoplotermes sp. and Ruptitermes reconditus (Silvestri) are new records for ESEC-Seridó.

The family Termitidae was dominant in species richness in the three sites, with eight species, followed by Kalotermitidae and Rhinotermitidae, with only one species each. The most dominant subfamily of Termitidae was Termitinae, with four species, followed by Apicotermitinae and Nasutitermitinae, with two species each. The most abundant species in the three sites was Heterotermes sulcatus (Mathews), present in 73.3%, 66.7% and 50% of the parcels analyzed in sites A₁, A₂ and A_3, respectively. Moreover, the organic matter from cattle feces found in A₃ may be partially responsible for the presence of a number of species associated to this matter, such as Amitermes nordestinus (Mélo & Fontes).

According to nonparametric Jackknife 1 and Chao 2 estimators, estimated species richness in the three study sites was very close to the richness observed (Table 1). Additionally, the species accumulation curve demonstrated that the species richness of each site was not significantly different (Fig 1). There was also no significant difference between the number of encounters per transect (Kruskal-Wallis H = 5.39; P > 0.05).

The wood-feeders were dominant in the three sites analyzed in terms of number of species (three in each site) and relative abundance. On the other hand, only one species each of soil-feeder and leaf feeder (Anoplotermes sp. and R. reconditus, respectively) was found. A₁ was the only site to contain all the feeding groups. The number of encounters was not significantly different between the sites for the feeding groups (χ² = 11.49, gl = 6, P > 0.05) and taxonomic groups (χ² = 3.47, gl = 4, P > 0.05).

Estimated termite nest density in the three sites was 0.7 active nests/ha. All the nests were considered arboreal, given that they were built on live or dead trees, without direct contact with the soil. The nest-building species were Microcerotermes strunckii (Sörensen) and Nasutitermes corniger (Motschulsky). The only site with nests was A_1, with a density of 2 active nests/ha, whereas in the other sites no nests were found in the parcels analyzed. However, the presence of N. macrocephalus and M. strunckii nests in A₃ and of N. corniger nests in A₂ was observed, but not within the gradients.

Of the eleven environmental variables collected in the three sites (Fig 2), only mean plant density in A₃ showed a significant difference compared to the other sites (F_{2, 12} = 11.59; P < 0.05). Principal component analysis (PCA) showed no clear separation between the sites in terms of their abiotic attributes and vegetation structure (Fig 3). The sites were not significantly different according to the scores of the first principal component (F_{2, 11} = 2.96; P = 0.22).

Discussion

The species richness found at ESEC-Seridó was relatively low, when compared to other caatinga sites in Northeast Brazil (Mélo & Bandeira 2004, Vasconcellos et al 2010). Including the present study, 30 species of termites have been recorded in the literature for the Caatinga (Martius et al 1999, Mélo & Bandeira 2004, Vasconcellos et al 2010). Specifically for these studies, species richness may be higher, since the samples collected by Kalotermitidae and Apicotermitinae were at morphospecies level and were not compared among the sites. In addition, these studies were conducted only in the ecoregions of the "Depressão Sertaneja Setentrional" (Martius et al 1999, Mélo & Bandeira 2004) and on the edge of the Borborema Plateau (Vasconcellos et al 2010) even though there are six other ecoregions with different climatic, geomorphological and vegetation characteristics in the Caatinga (Velloso et al 2002).

The assemblages of ESEC-Seridó, composed of few species and trophic groups and low relative species abundance, were considered structurally simple. Moreover, most species, except A. nordestinus, are widely distributed geographically, occurring in at least two distinct biomes. The low species richness of termites at ESEC-Seridó may be related to the low mean annual rainfall, usually concentrated into three months (Ministério da Agricultura 1978), and low concentration of organic matter in the soil, owing to the low productivity of this ecosystem and the anthropic disturbances that have been imposed on this region for decades (Vasconcellos et al 2010).

The greater relative abundance of H. sulcatus in the three sites examined suggests that the species is one of the most important in wood cycling at ESEC-Seridó and the most resistant to anthropic disturbance. Its colonies can resist high soil temperatures (mean = 34°C) and consume around 98.8 kg of dry wood/ha/year. It exhibits a preference for exotic wood species, such as the mesquite tree (Prosopis juliflora) (Mélo & Bandeira 2007). These attributes may explain the dominance of this species in all the parcels, mainly in A₃.

At ESEC-Seridó in the late 1990s, Martius et al (1999) recorded the presence of a species of Constrictotermes, later identified as Constrictotermes cyphergaster (Silvestri), and Inquilinitermes spp., with a density of two active nests of C. cyphergaster/ha. Species of Inquilinitermes are obligatory inquilines of nests built by the species Constrictotermes (Mathews 1977). In the present study, none of these termite colonies were found, suggesting a significant decrease in the abundance of these species at ESEC-Seridó.

The causes of the drastic local reduction or extinction of these termite populations are unknown. However, the absence of C. cyphergaster nests negatively affects many other populations of animals, since their active and abandoned nests serve as sites of refuge, predation and nest building for several invertebrates and even vertebrates such as a number of psittaciform birds that build their nests in the abandoned structure (Mathews 1977, Cunha & Brandão 2000, Barreto & Castro 2007). The reduced populations of C. cyphergaster at ESEC-Seridó could also affect nutrient cycling and energy flow in the site. Moura et al (2008) estimated consumption of 44.5 ± 14.70 kg of wood/ha/year only for populations of C. cyphergaster in a site of caatinga located in the state of Paraíba, Northeast Brazil. In this same site, a density of 59.0 ± 22.5 active nests/ha was estimated only for C. cyphergaster (Vasconcellos et al 2007).

At ESEC-Seridó, Martius et al (1999) estimated a density of one nest/ha for Nasutitermes and Microcerotermes. A similar result was obtained in site A₁, where one active nest per hectare was recorded for the species N. corniger and M. strunckii. Vasconcellos et al (2010), in turn, found a density that was sixteen times higher than the value obtained for the three sites analyzed at ESEC-Seridó, equivalent to 11 nests/ha, in a caatinga site submitted to selective timber cutting (± 30 years ago) located at Almas Farm, a Reserva Privada do Patrimônio Natural (RPPN) in the state of Paraíba.

The composition of feeding groups was not different between the sites. It is known that the anthropic disturbance of habitats can decrease nest-building sites and resource availability for a number of termite populations (Bandeira & Vasconcellos 2002, Junqueira et al 2008). In humid tropical forests there is a tendency for soil-feeding species to be more affected by habitat disturbance (Souza & Brown 1994, Bandeira et al 2003, Jones et al 2003, Vasconcellos et al 2008). On the other hand, wood-feeding termites were the most affected by habitat disturbance in a caatinga site located in the state of Paraíba (Vasconcellos et al 2010).

Even though disturbance time was different, no statistical support was found to substantiate the idea that the sites are different in relation to the environmental variables measured and disturbance levels. The same pattern was observed for termite assemblages, in that no significant difference was observed in species richness, relative abundance, and composition of feeding and taxonomic groups between the three sites. Vasconcellos et al (2010) found a significant difference between caatinga sites with different levels of habitat disturbance caused by extensive vegetation cutting and cattle/goat trampling. These disturbances resulted in the presence of structurally different termite assemblages in the sites, in accordance with the difference in environmental variables measured in each site.

In the present study, the habitats were quite similar, showing no significant differences between their structural attributes or termite assemblage composition. The similarity of these results reinforces the potential of termites as biological indicators of habitat quality in the Caatinga.

Acknowledgments

To Professors Adelmar G Bandeira, Márcio Z Cardoso and Adalberto A V Freire for suggestions during the writing of this article; to Ricardo Pacheco and João B Trindade for help in field collections; to the Fundação de Apoio à Pesquisa do Estado do Rio Grande do Norte - FAPERN, for supporting the project; to the management team at the Estação Ecológica do Seridó for access and support during the study.

Received 21 September 2009 and accepted 05 March 2010

Edited by Wesley A C Godoy - ESALQ/USP

• Bandeira AG (1989) Análise da termitofauna (Insecta: Isoptera) de uma floresta primária e de uma pastagem na Amazônia Oriental. Bol Mus Para Emilio Goeldi Ser Zool 5: 225-241.
• Bandeira AG, Vasconcellos A (2002) A quantitative survey of termites in a gradient of disturbed highland forest in Northeastern Brazil (Isoptera). Sociobiology 39: 429-439.
• Bandeira AG, Vasconcellos A, Silva M, Constantino R (2003) Effects of habitat disturbance on the termite fauna in a highland humid forest in the Caatinga Domain, Brazil. Sociobiology 42: 117-127.
• Barreto LS, Castro MS (2007) Ecologia de nidificação de abelhas do gênero Partamona (Hymenoptera: Apidae) na caatinga, Milagres, Bahia. Biota Neotropica 7: 137-142.
• Brown Jr KS (1997) Diversity, disturbance, and sustainable use of Neotropical forests: insects as indicators for conservation monitoring. J Insect Conserv 1: 25-42.
• Clarke KR, Warwick RM (2001) Change in marine communities: an approach to statistical analyses and interpretation. PRIMER-E: Plymouth, 91p.
• Colwell RK (2005) EstimateS: Statistical estimation of species richness and shared species from samples. Version 7.5. Persistent URL <http://viceroy.eeb.uconn.edu/estimates purl.org/estimates>
• Constantino R (1998) Catalog of the living termites of the new world (Insecta: Isoptera). Arq Zool 35: 135-231.
• Constantino R (2005) Padrões de diversidade e endemismo de térmitas no bioma Cerrado, p.321-333. In Felfili JM, Scariot A, Sousa-Silva JC (org) Cerrado: ecologia, biodiversidade e conservação. Brasília, Ministério do Meio Ambiente, 439p.
• Cunha HF, Brandão D (2000) Invertebrates associated with the Neotropical termite Constrictotermes cyphergaster (Isoptera: Termitidae, Nasutitermitinae). Sociobiology 37: 593-599.
• Deligne J (1966) Caractères adaptatifs au regime alimentaire dans la mandibula des termites (Insectes: Isoptères). C R Acad Sc 263: 1323-1325.
• Eggleton P, Bignell DE, Hauser S, Dibog L, Norgrove L, Madong B (2002) Termite diversity across an anthropogenic disturbance gradient in the humid forest zone of West Africa. Agric Ecosyst Environ 90: 189-202.
• Holt J A, Coventry RJ (1990) Nutrient cycling in Australian savannas. J Biogeogr 17: 427-432.
• Jones DT, Susilo FX, Bignell DE, Hardiwinotos S, Gillison AN, Eggleton P (2003) Termite assemblage collapse along a land-use intensification gradient in lowland central Sumatra, Indonesia. J Appl Ecol 40: 380-391.
• Junqueira LK, Diehl E, Berti Filho E (2008) Termites in eucalyptus forest plantations and forest remnants: an ecological approach. Bioikos 22: 3-14.
• Leal IR, Silva JMC, Tabarelli M, Lacher Jr TE (2005) Changing the course of biodiversity conservation in the caatinga of Northeastern Brazil. Conserv Biol 19: 701-706.
• Leal IR, Tabarelli M, Silva JMC (eds) (2003) Ecologia e conservação da Caatinga. UFPE, Recife, 804p.
• Martius C, Tabosa WAF, Bandeira AG, Amelung W (1999) Richness of termite genera in a semi-arid region (sertão) in NE Brazil. Sociobiology 33: 357-365.
• Mathews AGA (1977) Studies on termites from the Mato Grosso State, Brazil. An Acad Bras Cienc, An Acad Bras Cienc, 267p.
• Mélo ACS, Bandeira AG (2004) A qualitative and quantitative survey of termites (Isoptera) in an open shrubby caatinga, Northeast Brazil. Sociobiology 44: 707-716.
• Mélo ACS, Bandeira AG (2007) Consumo de madeira por Heterotermes sulcatus (Isoptera: Rhinotermitidae) em ecossistema de caatinga no Nordeste do Brasil. Oecol Bras 11: 350-355.
• Ministério da Agricultura (1978) Aptidão agrícola das terras do Rio Grande do Norte. Brasil. Brasília, Ministério da Agricultura.
• Moura FMS, Vasconcellos A, Araújo VFP, Bandeira AG (2008) Consumption of vegetal organic matter by Constrictotermes cyphergaster (Isoptera, Termitidae, Nasutitermitinae) in an site of caatinga, Northeastern Brazil. Sociobiology 51: 181-189.
• Sampaio EVSB (1995) Overview of the Brazilian Caatinga, p.35-58. In Bullock SH, Mooney HA, Medina E (eds) Seasonally dry forests. Cambridge, Cambridge University Press, 875p.
• Sena JM, Vasconcellos A, Gusmão MAB, Bandeira AG (2003) Assemblage of termites in a fragment of cerrado on the coast of Paraíba State, Northeast Brazil (Isoptera). Sociobiology 42: 753-760.
• StatSoft Inc. (2005) Statistica (data analysis software system), version 7.1. www.statsoft. com
• Souza OFF, Brown VK (1994) Effects of habitat fragmentation on Amazonian termite communities. J Trop Ecol 10: 197-206.
• Vasconcellos A, Araújo, VFP, Moura FMS, Bandeira AG (2007). Biomass and population structure of Constrictotermes cyphergaster Silvestri (Isoptera: Termitidae) in the dry forest of Caatinga, northeastern Brazil. Neotrop Entomol 36: 693-698.
• Vasconcellos A, Bandeira AG, Almeida WO, Moura FMS (2008) Térmitas construtores de ninhos conspícuos em duas áreas de mata atlântica com diferentes níveis de perturbação antrópica. Neotrop Entomol. 37: 15-19.
• Vasconcellos A, Bandeira AG, Moura FMS, Araújo VFP, Constantino R (2010) Termite assemblages in three habitats under different disturbance regimes in the semi-arid Caatinga of NE Brazil. J Arid Environ 74: 298-302.
• Velloso AL, Sampaio EVSB, Pareyn FG (orgs) (2002) Ecorregiões do bioma Caatinga. Recife, APNE / The nature conservancy, v.1, 76p.
• Walther BA, Moore JL (2005) The concepts of bias, precision and accuracy, and their use in testing the performance of species richness estimators, with a literature review of estimator performance. Ecography 28: 815-829.
• Whitford WG (1991) Subterranean termites and long-term productivity of desert rangelands. Sociobiology 19: 235-243.

Publication Dates

• Publication in this collection
  14 Mar 2011
• Date of issue
  Feb 2011

History

• Received
  21 Sept 2009
• Accepted
  05 Mar 2010