Species and Interaction Diversity in Plants and Phytophagous Insects

Principal investigator: Thomas Michael Lewinsohn <thomasl@unicamp.br>omasl@unicamp.br
Laboratório de Interações Insetos-Plantas
Instituto de Biologia, Unicamp
13081-970 Campinas, SP, Brasil
Phone +55-19-7887022 Fax +55-19-2893124

PROJECT

INTRODUCTION

Most inventories and censuses of biodiversity in terrestrial systems are conducted with vascular plants and vertebrates, and directed at particular taxa. Considerable expertise and substantial data sets are available for plants (especially angiosperms) and various classes of vertebrates for most habitats and ecoregions on the planet. However, the bulk of the world’s biodiversity belongs to other taxa, notably arthropods, which are far less known and surveyed. In recent years, many authors have stressed the need to increase and organize efforts of surveying, inventorying, describing arthropods and other taxa. This entails a major capacity-building effort in taxonomy as well as the development of new methods to obtain samples of various taxa in different ecological settings, as well as new procedures to handle and process these samples and to analyse the ensuing data.

Arthropods offer a number of advantages for inventorying or monitoring diversity (Kremen et al., 1993): they occur in virtually any terrestrial habitat; their distribution, geographical and among habitats, is often finer-grained than that of vertebrates and they respond readily to a number of ecological processes; thus, they are potentially highly suitable as indicators of geographical or ecological uniqueness or to give early warning of various forms of disturbance. Moreover, their small body sizes and often large numbers make mass collecting feasible which, together with high species numbers, in turn make them suitable for statistical assessment of total species richness and its breakdown in various ways, notably into local and regional components and their turnover (e.g. Lewinsohn, 1991; Colwell and Coddington, 1994).

Inventories of arthropods will often concern whole taxa, at the family level or higher (Coddington et al., 1991) but recent proposals have advocated the advantages of more restricted sampling, in two different ways: either by concentrating on certain well-circumscribed habitats, such as single tree canopies sampled by fogging (e.g. Basset, 1996) or by the restriction to a smaller, "ecologically informative" taxon, for instance a single butterfly genus (Kremen, 1994).

In the present project we intend to build on our previous experience in investigating the diversity of a set of terrestrial insects defined by a common resource set rather than by a purely taxonomic criterion (Lewinsohn, 1991; Müller-Schärer et al., 1991; Lawton et al., 1993). Since plants are the resource base on which we survey and record phytophagous insects, plant diversity is both an objective in itself and a potential determining factor for the diversity of insects associated to these plants.

Our approach is somewhat similar to studies of canopy arthropod assemblages. However, we concentrate on endophagous insects that are collected within plants and reared out from them. Thus, while the variety of taxa and the size of the assemblages we study are much smaller than that encountered by external collecting methods such as canopy fogging, our approach ensures that every insect is feeding in the plant and that they are indeed trophically linked. Two consequences follow on that. First, such species assemblages are parts of food webs and therefore can be examined and evaluated for food-web patterns and traits. We are particularly concerned, for instance, with the degree of compartmentation of local assemblages, which should affect their stability and resilience to disturbance or their response to new additions, be it introduced plants or insects (Pimm, 1991; Polis and Winemiller, 1996). Beyond the number and species makeup of a local biotic assemblage, its structure of trophic associations – its interaction diversity (Thompson, 1997) – is potentially useful as indicator of biological diversity and to monitor responses its changes.

The second consequence of targeting assemblages of endophagous insects as a study system is their evolutionary relationship to host plants. Host associations evince long-term, phylogenetic patterns (Farrell et al., 1992) as well as short-term adjustments to host distributions and combinations and their variation in time and space (Thompson, 1994). Thus, endophagous insect diversity can be studied for their phylogenetic structure and constraints, but also as to their short-term response to shifting host conditions and combinations among localities and in time. In this regard, intraspecific genetic differentiation among host plants or among localities is an important question which we intend to address. Genetic studies of congeneric phytophagous species are essential for an assessment of the importance of phylogenetic restraints and radiation in establishing them (or not) on different hosts. Furthermore, studies of genetic similarity of the host plants themselves, both among conspecific populations and among species, provide further insights into the evolutionary establishment and maintenance of insect-plant associations.

The approach which we have been developing in previous work and which we propose for the current project, allows us to address biological diversity simultaneously at the level of species diversity and genetic diversity. Diversity at these two levels of organization is evaluated at different spatial scales, from the local host plant patch up and across different habitats and ecoregions. Local and regional components of biological diversity and their organization and dynamics at these different scales, are one of the main concerns of current ecological research (e.g. Lewinsohn, 1991; Ricklefs and Schluter, 1993; Ricklefs et al., 1995).

Among other potential correlates and predictors of insect biological diversity we single out the secondary chemistry of their host plants. Chemical diversification of plants is intimately related to insect diversification and a major candidate to explain both long-term, phylogenetically structured patterns of association, and short-term preferences or host shifts in local assemblages (e.g. Cates, 1980). Chemical screening of plants is one of the main thrusts of biodiversity prospecting, but usually with little concern for the ecological implications of chemical diversity. The Asteraceae, on which our current work is targeted, are well-known for their high chemical diversity; however, most neotropical species have never been investigated for their secondary chemistry. Thus, chemical diversification of the Asteraceae is an important potential explanatory factor for the diversity and patterns of association within our target assemblages; but the chemical survey of plants within this family is a goal of interest in itself.

ASTERACEAE AND FLOWERHEAD INSECTS AS A STUDY SYSTEM

The main system for this project are the Asteraceae (Compositae) and their associated phytophagous insects, especially the flowerhead feeding insects, in Southern and Southeast Brazil. Asteraceae fulfill several prerequisites and offer a number of advantages:

• nearly ubiquitous, with a rich gamut of wide-ranging species up to strict habitat endemics
• cosmopolitan, allowing comparisons across localities up to different continents and life zones
• taxonomically well studied; major subdivisions, especially tribes, fairly well-established, with proposed phylogenies for many major sections within the family (Bremer, 1994)
• they include serious weeds extensively canvassed for biocontrol agents; in South America, these include Chromolaena odorata (Crutwell, 1974), Baccharis, and several Heliantheae (Daniel Gandolfo, USDA/Argentina, unpublished).
• they also include several "classical" systems for the study of the ecology and evolution of insect-plant interactions, such as Senecio, Solidago, Centaurea, Haplopappus, their herbivores and parasitoids (Louda, 1982; Zwölfer, 1988; Root and Capuccino, 1993; Abrahamson and Weis, 1997).

Most endophagous insects associated with Asteraceae flowerheads are restricted to this family, and several (e.g. Diptera Tephritidae, Agromyzidae and Cecidomyiidae; Lepidoptera Pterophoridae and Tortricidae) count Asteraceae among their major, if not the foremost, host group. Thus, these insect taxa have evolved in close association with Asteraceae (Zwölfer, 1988; Lewinsohn, 1991).

Other study systems may be included in latter phases of the project, to validate results from the Asteraceae – flowerhead insect system.

PREVIOUS RESULTS

Results from previous grants are summarized in a later section (p.*). Here we give a brief history of this line of investigation within our research group.

The Asteraceae – flowerhead insect system has been studied by us since 1985. Initial samples established the general pattern of associations and partitioned local and regional diversity in various habitats of Southeast Brazil (Lewinsohn, 1991). Further study has concentrated on a local assemblage of Vernonieae and Tephritidae (Prado and Lewinsohn, 1994; ms in prep) and the genetics of Tomoplagia, the main tephritid genus in this assemblage, has also been studied in detail, showing a close match of host plant and genetic similarities (Solferini et al., in prep).

In an ensuing project from 1995 to 97 we collected and processed 1510 samples of more than 400 species of Asteraceae (over 20% of known species in Brazil) in two montane areas: the southern Espinhaço range in the State of Minas Gerais, and the coastal ranges and highlands in Santa Catarina and Rio Grande do Sul. Almost 17 000 insects were reared, sorted and (except for gall midges) mounted from these samples. Till now, Tephritidae, Tortricidae, Pterophoridae and Pyralidae have been wholly sorted, and Agromyzidae are near completion.

In all, 167 plant species were sampled at least three times in at least two localities, and 65 species have been sampled in five or more localities, thus being adequate for an analysis of local and regional richness. The whole database provides an extensive matrix of plants, insects and their association across various localities. This represents a unique resource for analyzing the distribution of diversity within and across localities, regions, and host taxa of different levels.

Genetic studies are being completed on two tephritid genera: Tomoplagia and Xanthaciura and also on two endemic plant genera. Chemical studies have investigated the structure and interspecific diversity and similarity of pyrrolizidine alkaloids within two genera, Senecio and Trichogonia (sensu lato, including Trichogoniopsis Robinson). Further studies are being directed at other genera of the Eupatorieae, particularly in the Eupatorium complex itself.

CERRADO

Conservation of the Cerrado biome has been of less concern than Amazonian and Atlantic rain forest or the Pantanal; yet it is under immediate threat, and disappearing at a faster rate than any other major terrestrial biome in Brazil. Cerrado is also under-represented in conservation areas. This situation was recently revised at the national level (PROBIO-MMA, 1998) and a synthesis of proposed priorities for conservation within the biome is forthcoming. Within São Paulo, a major study integrating existing information has resulted in a comprehensive survey of existing Cerrado areas and remnants and a proposal for priorities in their conservation and management (São Paulo - SMA, 1997).

Given the current effort in improving conservation and management efforts within the Cerrado region, the considerable information (especially floristic) already available, and the fair representation of Asteraceae, especially among herbs and shrubs, the cerrado is a sensible choice for a major target of this study, noting also that the cerrado biome does include the highland fields and montane meadows (campos rupestres) which are among the richest and most diverse areas in Asteraceae in Brazil, and which have been the main study areas of our recent work on this system.

OBJECTIVES

We propose to:

• Investigate the structure of biological diversity in subsets of natural communities composed of plants and their associated phytophagous insects, using Asteraceae and flowerhead insects as a base system.
• Sample Asteraceae and their associated phytophages in extensive areas and remnants of cerrado in the State of São Paulo and, for comparative purposes, in cerrado core areas in adjoining states.
• Develop or experiment existing methods to separate total biological diversity in components of different spatial, temporal and structural scales, e.g. within-host diversity and among host turnover, and local diversity versus species richness.
• Correlate diversity of Asteraceae and their endophages with that of other taxa or community subunits in cerrado areas.
• Extend the existing database on Asteraceae and phytophagous insects by sampling other areas, especially the Mantiqueira range on the northeastern border of São Paulo; its transition into the Espinhaço range in Southern Minas Gerais; and various open habitats in montane and coastal areas within the Atlantic rainforest biome.
• Investigate the contribution of genetical, phylogenetic, chemical, and ecological factors in determining species diversity or interaction diversity and its variation on different spatial, structural and time scales.
• Investigate the response of species diversity, interaction diversity and species composition to various regimes or histories of disturbance or impact, through comparisons of assemblages in preserved Cerrado areas with those in nearby sites.
• Investigate the structure of species interactions in these systems by applying existing analyses and theory or by developing new approaches.

RATIONALE

The aims of this project are in accordance with the objectives of the Special Research Program "Conservation and sustainable use of the biodiversity of the State of São Paulo - BIOTA/SP" (see p.*).

Within the general goals of the BIOTA-SP initiative, the present proposal fits in most clearly with the following: (1) "to study and know the biodiversity of the State of São Paulo and to disseminate this knowledge and its importance"; (2) "to understand the processes that generate, maintain and impact this biodiversity"; (5) "to develop methodological bases and reference standards for studies of environmental impact." We expect this project to contribute especially to goals 2 and 5.

The proposed study may also help to attain several other goals of this initiative, but we deem goal 13 to be exceptionally important: "to contribute to implant the Convention of Biological Diversity in the State of São Paulo, joining national and international efforts for the knowledge, conservation and sustainable use of biodiversity". Our proposed effort to investigate biodiversity jointly at the genetic and the species level and at different spatial scales so that results may be interpretable up to the landscape level as well, may represent its most evident contribution to attain this general goal.

Prioritization of the Cerrado biome among the several ecosystems to be studied has been discussed in the Introduction. This should also fit other projects concerning Cerrado in the BIOTA-SP initiative (such as proposals 13 and 17).

METHODS

FIELD SAMPLING AND REARING

Our basic information unit is the use of a plant species by an insect species in a given locality, derived from our field records and samples. Thus, the project will produce biogeographical information on Asteraceae and their endophagous insects, and on the associations among them as well. The minimal information set for our records comprises the locality name, GPS-recorded coordinates, altitude, habitat, collector name, and date. These data will be made available to the Biodiversity Data Base of the State of São Paulo, and to that end they conform to the minimum standards established for all projects under the BIOTA-SP initiative.

The general sampling procedure within a chosen locality consists of:

In the laboratory, flowerheads are maintained in containers with a meshed cover to rear out adult insects. Samples are inspected periodically and emerged adults are stunned with a stream of CO₂ , stored in vials and later mounted on entomological pins, except for cecydomyiid flies, which are preserved in alcohol or will be mounted in microscope slides (Gagné, 1994). After adult emergence ceases, within 60 days after collecting, samples are dried and weighed. A fraction of flowerheads from each sample is counted, dried and weighed to establish individual flowerhead weights, used to estimate total numbers of flowerheads in samples from their dry bulk weight.

SORTING AND IDENTIFICATION

Mounted specimens are sorted into morphospecies based on usual characters for each group; in some cases, e.g. Agromyzidae, this entails dissection and mounting of genitalia of specimens from each sample. Sorting is aided by reference collections ("morphotypes") and by photographs and photomicrographs of whole specimens (habitus) and relevant details.

Identifications are based on specimens previously identified by specialists and on published descriptions. Plant identifications are checked by participating and collaborating investigators (João Semir, Roberto Esteves, Nicholas Hind, John Pruski; others to be added). Subsets of reared insects are sent to collaborating investigators for identification or verification (Allan Norrbom, Diptera Tephritidae; Vitor Becker, Lepidoptera Tortricidae; Cees Gielis, Lepidoptera Pterophoridae) and occasionally to other taxonomists.

The only major group of flowerhead endophages for which we have no established procedure as yet are Diptera Cecidomyiidae. Genera commonly found in asteraceaous flowerheads can be identified fairly readily according to Gagné (1994 and references therein). However, identification at the species level is demanding (Raymond Gagné, personal communication) and we may presume that our samples include a fair number of undescribed taxa. Therefore, to saort these midges reared specimens and pupal cases from each sample will be mounted on microscope slides and then examined under interference microscopy, according to Gagné (1994 and personal communication). We intend to set up a reference collection of microscope slides and digitize pictures of adults (with special attention to antennal segments, legs, and genitalia) and larvae, especially the sternal spatula. These often present usable characters in this family, and discrete characters may offer a basis for an operational separation of morphospecies in the group. After the first year, we will evaluate our results and the effort involved, to decide whether we will continue to attempt to separate the family to species level.

Insect specimens will be stored in reference collections in our lab and in the permanent collections of the Museu de História Natural of Unicamp. Type specimens will be placed in the Museu de Zoologia of USP (MZUSP), São Paulo. Plant specimens will be stored in the Unicamp Herbarium. Selected specimens may be lent or donated to other collections in Brazil or overseas.

DATA STORAGE AND ANALYSIS

Both in field work and in rearing and sorting, we intend to use sample identification and data entry with barcoding, to reduce errors and improve efficiency. We also intend to use digital images of sampled plants and insects as aids to recognition and identification, incorporating them in the project data base. One of the intended products of the project is a visual identification system and guide to the organisms in this study, both plants and insects.

GENETIC STUDIES AND PHYLOGENETIC ANALYSES

Genetic studies focus on phytophagous genera comprising several species with well-established host ranges, varying from strict monophagy to several tribes of Asteraceae. Previous studies have concentrated on two genera of tephritids. Within this project, Tephritidae will still be the main study group but others will be added, starting with the genus Melanagromyza (Diptera Agromyzidae).

Genetic variability of populations will be evaluated through several analytical techniques, using nuclear and cytoplasm markers in combination with the sequencing of particular regions to produce phylogenies. Markers with different modes of inheritance allow better and more detailed inferences on observed patterns. Eletrophoretic analysis of isozymes, the main technique employed in previous work, will be used as a complementary technique to examine the extent of genetic differentiation among hosts and/or localities, and whether variability covaries with host range.

Variability and interspecific differentiation of host plants will also be examined through enzyme and DNA studies, concentrating at first on populations of endemic and restricted species and genera.

SECONDARY CHEMISTRY OF ASTERACEAE

Following on previous work, we intend to screen host plants for pyrrolizidine alkaloids. Work will span plants in various tribes but concentrate on Eupatorieae and Senecioneae. Fresh flowerheads are collected in ethanol and acid-base extracts are analyzed through gas chromatography - mass spectrometry. Other chemical groups will be added to the screening program if feasible.

SAMPLING PLAN

CERRADO STUDY

In project years 1-2, ten representative localities of Cerrados will be chosen within São Paulo according to their size, preservation status, plant species richness and distance from others sites. Each site will be sampled once at the beginning of the dry season and once in the rainy season, during two years. Five sites within the core area of the Cerrados (Minas Gerais, Goiás and Distrito Federal) will be sampled twice within a year, to compare their diversity and composition to the São Paulo sites, which are fairly close to the southern limit of cerrado distribution. These initial samples will serve to establish the general insect-plant association matrix for assemblages within the cerrado.

In years 2-4 of the project, a larger number of sites (30 or more) will be sampled, concentrating on commoner and widespread plant species (ca. 10 species). These sites will include areas with different kinds of use and, preferably, with a known history. They will serve to tease out the relationship between local and regional insect species richness and the response of insect assemblages to different forms of use.

Within the cerrado biome, further montane areas will be sampled to complete the insect-plant array, in one or two extended field trips per site.

MANTIQUEIRA AND OTHER AREAS

Asteraceae in the Serra da Mantiqueira and its transition into the Espinhaço will be sampled in years 1-3 in three different seasons: early wet season, late wet season, and early dry season. Mantiqueira sites will spread out from two foci: Campos do Jordão (SP) and Itatiaia (RJ). The transition to the Espinhaço will be sampled in a broad west-east arch from the Serra da Canastra up to the Parque do Ibitipoca (MG).

Montane and coastal areas within the Atlantic forest region will be sampled in two field trips per site. Sites will coincide if feasible with those of other projects in the BIOTA-SP initiative.

PERMANENT SITES

We intend to establish at least one permanent site to initiate a longer-term study of host populations and their associated insects, to be surveyed extensively and then monitored for at least five years. Since Unicamp has no field station, choice of site(s) will depend on their quality, guarantees that the study area will not be cleared or disturbed, and authorization to conduct field experiments.

LITERATURE CITED