Conflicts of interest: the authors have declared no conflicts of interest.
DNA profiling of host–herbivore interactions in tropical forests
Article first published online: 4 JAN 2010
© 2010 The Authors. Journal compilation © 2010 The Royal Entomological Society
Special Issue: Insect Evolution Below the Species Level
Volume 35, Issue Supplement s1, pages 18–32, January 2010
How to Cite
NAVARRO, S. P., JURADO-RIVERA, J. A., GÓMEZ-ZURITA, J., LYAL, C. H. C. and VOGLER, A. P. (2010), DNA profiling of host–herbivore interactions in tropical forests. Ecological Entomology, 35: 18–32. doi: 10.1111/j.1365-2311.2009.01145.x
- Issue published online: 4 JAN 2010
- Article first published online: 4 JAN 2010
- Accepted 11 October 2009
- Molecular identification;
- insect–plant interactions;
- Top of page
- Materials and methods
- Supporting Information
1. The diversity of insects in tropical forests remains poorly known, in particular regarding the critical feeding associations of herbivores, which are thought to drive species richness in these ecosystems.
2. Host records remain elusive and traditionally require labour-intensive feeding trials. A recent approach analyses plant DNA ingested by herbivorous insects; direct PCR amplification from DNA extracts from weevils (Curculionoidea) using chloroplast (trnL intron) primers was successful in 41 of 115 cases, resulting in 40 different sequences.
3. The resulting trnL intron sequences were identified against public databases to various hierarchical levels based on their position in phylogenetic trees and shown to be members of 26 plant families from different major groups of angiosperms.
4. Among the trnL intron sequences, seven pairs or triplets of close relatives (0–2 bp difference) were found which may represent intraspecific variation in the respective host plants.
5. Molecular clock calibrations of mitochondrial cox1 sequences of weevils established great distances of lineages obtained (all splits estimated >20 Mya). Distant taxa were found to feed on the same or similar hosts in some cases, showing low evolutionary conservation of host associations among deeper levels.
6. The technique provides a new means of studying species diversity and plant–herbivore interactions in tropical forests, and removes the constraints of the need for actual observations of feeding in ecological and evolutionary studies.
- Top of page
- Materials and methods
- Supporting Information
A means of rapidly and effectively assessing insect–plant feeding associations is urgently needed. The huge species richness of insects in tropical forests is widely attributed to their interactions with an equally puzzling diversity of angiosperm host plants (Janzen, 1970; Farrell, 1998). The major hypotheses either invoke the complex interactions of the insect herbivores and plants, e.g. due to high host specificity and niche partitioning associated with plant defence systems (Coley & Barone, 1996; Novotny et al., 2006) or the correlation with host plant phylogenetic diversity, which is highest in tropical regions (Coley & Barone, 1996; Novotny et al., 2006). A further possible factor ratcheting up total species richness in the tropics is turnover (beta-diversity) in host associations among local assemblages (Lewinsohn & Roslin, 2008). In all cases, testing these hypotheses of tropical insect–host plant diversification requires solid data on host associations and diet breadth. Such host records are equally important in estimations of the magnitude of species richness itself, which rely heavily on extrapolations from the better-known plant diversity (e.g. Erwin, 1982; Odegaard, 2000).
The methodological approach adopted for establishing feeding associations may affect the inferences drawn about host specificity and herbivore species richness. Diverse approaches have been used to obtain data on host associations and host preferences of insects, but all of them are time-consuming and have various limitations. Classical methods include observations of host use either in situ (Barone, 1998, 2000) or in laboratory tests (De Boer & Hanson, 1984; Lee & Bernays, 1988; Barone, 1998; Novotny et al., 2002; Novotny et al., 2006; Dyer et al., 2007), transplantation experiments (Eichhorn et al., 2008) or behavioural tests by exposure to plant volatiles (Schneider, 1957; Chen & Fadamiro, 2007; Fernandez et al., 2007). Other studies have attempted the direct identification of the feeding source, either through morphological analysis of the gut content (Otte & Joern, 1976; Fry et al., 1978), diet plant tissue-specific staining techniques (Schlein & Jacobson, 1999), or diet plant isotope analysis from gut contents (Post, 2002). Most analyses of insect host specificity in rainforests have been much less detailed. Early studies did not test feeding directly but used capture sites (individual rainforest trees) to establish host associations (e.g. Erwin, 1982). Studies based on direct feeding trials generally led to estimates of tighter host specificity and generally better founded data on host associations than simple presence on a host plant (Lewinsohn & Roslin, 2008). This would make a strong case for feeding studies to establish the true nature of herbivore interactions, but the huge expense in manpower required (e.g. Weiblen et al., 2006) is prohibitive in most cases. In addition, feeding studies of this kind usually concentrate on common herbivore species (Barone, 2000; Novotny et al., 2007), although most species (of plants and herbivores) in tropical forest assemblages are rare (Novotny & Basset, 2000), while artificial breeding conditions may alter insect behaviour and therefore result in inaccurate conclusions on host breadth. A further problem is that the taxonomic uncertainty in hyperdiverse insect groups due to numerous unnamed species which requires that specimens rather than names must be cross-checked among samples, which in turn adds great difficulties to analyses of herbivore data among independent studies (Lewinsohn et al., 2005). Finally, where species delimitation is incomplete and relies on preliminary morphospecies approaches, the finer details of host associations, including recently diverged races or species, may be missed altogether (Hebert et al., 2004; Condon et al., 2008).
DNA-based techniques can potentially solve the dual problem of imprecise insect taxonomy and incomplete host plant data in a single step. Specimens of folivorous leaf beetles have been shown to contain a ‘molecular record’ of their feeding source in the form of ingested plant material. Consequently, host plant DNA can be PCR amplified from a standard whole-body DNA extraction of herbivorous insects and identified against existing taxonomic DNA databases (Jurado-Rivera et al., 2009). This same DNA extraction is used to amplify diagnostic insect DNA fragments for a sequence-based identification of the herbivore. This procedure opens up entirely new approaches to the study of tropical herbivore–host plant interactions overcoming the existing limitations, as it both provides a sensitive means of comparison and potentially obviates the need for taxonomic identification to species or below of either plant or insect. If applied on a large scale, the technique promises to provide the elusive host data for complete herbivore communities in tropical forests and elsewhere, including rare species currently left out of the analyses. The approach has been shown to perform in practice, through comparison of hosts suggested by DNA analysis and hosts observed in the field (Jurado-Rivera et al., 2009).
Jurado-Rivera et al. (2009) applied their analysis to a group of Australian leaf beetles (Chrysomelidae), with the specific purpose of establishing herbivore–host plant co-evolution in the subfamily Chrysomelinae. Using the trnL (UAA) intron chloroplast marker (Taberlet et al., 1991) as a plant ‘DNA barcode’ (Taberlet et al., 2007), PCR amplification was achieved readily from freshly caught specimens. The resulting sequences were searched against a database of over 67 000 trnL intron sequences available at that time in GenBank and permitted identification to genus or tribe in most cases. PCR amplification was obtained in nearly every sample tested and revealed only a single host plant except in a small proportion (10 out of 76) of species (Jurado-Rivera et al., 2009). The latter could be separated with standard cloning techniques and resulted in successful identification of multiple hosts. A separate study obtained DNA from the gut of plant feeding individuals from various insect orders and readily produced authentic sequences of the rbcL locus (Matheson et al., 2007).
The Australian chrysomelids used for host identification were mainly taken from comparatively species poor plant assemblages in subtropical sclerophyll forest. At most study sites an inventory of local flowering plants was available and provided a means of confirming or refining the identification where the trnL sequence was inconclusive, either due to insufficient sequence variation or poor taxon coverage in the database (Jurado-Rivera et al., 2009). Host identifications in tropical rain forests constitute a greater challenge because of higher overall diversity of the assemblage, less complete floristic information at local sites, and highly variable database coverage among major host plant lineages. Here, we describe the results of a trial prior to any large-scale attempts, based on a set of weevils (Curculionoidea) reared from fruits and seeds of known hosts. DNA from beetle extractions was used for PCR amplification using plant trnL intron markers. In addition, we applied the procedure to specimens collected in flight intercept traps or by hand without the context of a host plant. Most of them were obtained from the tropical forest of Barro Colorado Island (BCI), Panama, whose plant composition is comparatively well known (Croat, 1978). The specimens were used to test factors critical for DNA-based surveys of host association, including (1) the proportion of individuals producing (single or multiple) plant sequences, and (2) the confidence of identifications achievable with this marker given the level of sequence variation and the richness of the existing reference database. Phylogenetic inferences from both insect and host plant sequences provided additional information on their level of co-evolutionary associations, despite the limited sampling available in this study.
Materials and methods
- Top of page
- Materials and methods
- Supporting Information
Taxon sampling and plant ‘barcodes’ from ingested DNA
Weevils were collected during April–November 2006 on Barro Colorado Island, Panama, in flight intercept traps and manually. Flight intercept traps were attached to 12 trees from different species in Barro Colorado Island Monument at an approximate height of 25–30 m from the ground. Two traps were also set in San Lorenzo National Park at similar heights. A few specimens were obtained by manual collecting at Parque Nacional Altos de Campana and Fortuna Dam Area (both Republic of Panama) and La Selva Biological Station (Costa Rica). In addition, several individuals were reared from seeds of known host plants, to test the success of the DNA-based host identification. Local experts identified the sampled plants, using morphological characters of fruits and leaves. Specimens of weevils were identified to the lowest taxonomic level possible using morphological characters and comparisons to the Natural History Museum collection. The species included here were members of three families of Curculionoidea: Brentidae, Curculionidae (six subfamilies) and Dryophthoridae.
A total of 115 individuals were selected for genomic DNA extraction (Table S1). The head and thorax of each individual were used for the extraction with a Promega 96-well plate kit. After extraction, beetles were prepared for morphological identification; vouchers will remain in the collections at Universidad de Panama and the Natural History Museum, London. To establish phylogenetic affinities of the beetles, a fragment of the 3′ end of the cytochrome oxidase I (cox1) gene was amplified using primers C1-J-2183 (Jerry) and TL2-N-3014 (Pat) (Simon et al., 1994) and resulted in 600–782 bp of sequence per individual. The same DNA template was used for PCR amplification of the plastid trnL intron using the plant-specific primers c-A49325 and d-B49863 (Taberlet et al., 1991; Jurado-Rivera et al., 2009). Sequencing was from both strands with the same primers used for PCR amplification, using BigDye 2.1 and an ABI PRISM3730 automated sequencer (Applied Biosystems, Foster City, CA, USA). Sequence chromatograms were assembled and edited with Sequencher 4.6 (Gene Codes Corp., Ann Harbor, Michigan) (Table S2).
Phylogenetic analysis and host species identification
Phylogenetic trees from cox1 sequences of weevils were obtained using maximum likelihood searches with RAxML 7.0.4 (Stamatakis, 2006) using two partitions (first and second codon positions together vs. third codon positions) and 1000 bootstrap replicates were performed (Stamatakis et al., 2008) under a GTR + I + Γ substitution model fixed in RAxML. Representatives of Brentidae were specified as outgroup during the tree search, as this family is closely related, but clearly separated, from the Curculionidae (Hunt et al., 2007; McKenna et al., 2009).
Node ages were calculated using an uncorrelated lognormal relaxed clock in BEAST 1.4.8 (Drummond & Rambaut, 2007). The brentid species were constrained as a monophyletic clade and 2.3% divergence My−1 (Brower, 1994) was used as a fixed substitution rate. Two independent runs of 100 million generations (sampling every 10 000th generation) were done, using the GTR + I + Γ model and a Yule prior under default parameters for all other settings. Convergence and mixing of MCMC chains were estimated by trace plots using Tracer 1.4.1 (Drummond & Rambaut, 2007), ensuring stationarity. Trees were summarised in Tree Annotator 1.4.8, keeping trees after burn-in of 100 000 generations.
Identification of trnL sequences was performed individually through similarity BLAST searches against GenBank (Altschul et al., 1990), and subsequent phylogenetic analysis of the query (diet) sequence together with the 100 top hits. The resulting topology was annotated for the higher taxa included at the level of genus, tribe and family. The position of the query in this tree was used to identify the unknown sequence as member of the lowest hierarchical group into which it was included with posterior probability >0.7 in Bayesian analysis (Jurado-Rivera et al., 2009). Prior to tree reconstruction, each query and its top hits, plus outgroup gymnosperm sequences of Cycas siamensis (AY651841), Gingko biloba (AY145323) and Pseudotsuga menziensii (AF327589) were aligned with MAFFT 5.0 using default parameters for the L-INS-I strategy (Katoh et al., 2005). Phylogenetic trees were obtained with MrBayes 3.1.2 (Ronquist & Huelsenbeck, 2003) using the GTR + I + Γ model, as chosen by Modeltest 3.7 (Posada & Crandall, 1998). Trees were obtained after two independent runs of three MCMC chains with varying numbers of generations, with a starting number of 3 million generations, sampling trees every 100th generation. Runs that converged only after 10 million generations were re-run using one cold and two incrementally heated Markov chains (λ = 0.1) and sampling every 1000 steps. Burn-in generations were calculated and excluded from the analysis as before and all-compatible consensus trees were obtained from the retained trees.
- Top of page
- Materials and methods
- Supporting Information
Diversity of weevils
Out of 781 nucleotide positions sequenced for the mitochondrial cox1, 454 positions were variable and 389 informative. The sample included 13, 99 and 3 individuals of Brentidae, Curculionidae, and Dryophthoridae, respectively, but more detailed morphological identification to species level proved difficult despite access to excellent reference collections. However, identification to genus or subfamily revealed a great taxonomic diversity. Only the genus Conotrachelus was represented by several species, including three species (four individuals) reared from fruits and seeds of plants of known identity. A maximum likelihood tree obtained from the cox1 sequences was generally consistent with the currently accepted classification (Alonso Zarazaga & Lyal, 1999), except for the paraphyly of the most widely sampled subfamily Cryptorhynchinae, and some inconsistencies at the tribal (Hylobiini) and genus (Teramocerus, Zygops, Eubulus) level (Fig. 1). Preliminary analysis of the sequences showed that in several cases individuals clustered closely together, suggesting that they were members of a single species (Hebert & Gregory, 2005; Pons et al., 2006). This was supported by their morphological similarity in each case where adults were available. Using only one individual per sequence cluster reduced the total from 115 to 87 individuals in the phylogenetic analysis of the weevils (see legend to Fig. 1 for details). Yet, beyond these intraspecific clusters the long terminal branches in the clock-constrained tree suggested a great phylogenetic divergence in this set of specimens.
Diversity of host-plant sequences
PCR amplification with the trnL intron primers on the 115 extracts producing cox1 sequences was successful in 41 cases (35.6%). Failure of amplification did not show any apparent bias across the weevil groups represented, although was mostly associated with individuals collected in flight intercept traps, suggesting that host DNA in these individuals is not optimal for amplification. Sequencing in each of the 41 positive cases resulted in clear sequence reads that were confirmed as authentic plant DNA sequences by high GenBank matches with angiosperm trnL sequences. The sequences obtained ranged in length from 379 to 609 bp (including terminal missing data) and showed several long indels that were generally synapomorphic for groups at the family level. After alignment the data matrix included 854 positions, of which 514 were variable and 356 parsimony informative. Average divergence (uncorrected p distances) between any pair of trnL sequences was 17.5%, ranging from 0 (sequences obtained from two specimens of Conotrachelus sp. 44 reared from Eugenia galalonensis) to 31% (obtained from Brentidae sp. 120 associated with Poaceae and Brentidae sp. 119 feeding on Meliaceae). Several other host sequences differed by 1–2 bp, i.e. less than 1% (Table 2).
|Identification||bp difference (uncorr. P)||Herbivore 1||Herbivore 2||Uncorr. P (cox1)||Locality|
|Sister to Malleastrum sp.||1 (0.00174)||Brentidae sp. 119||Teramocerus sp. 118||0.13263||BCI|
|Bombacoideae clade||1 (0.00153)||Entiminae sp. 4||Molytinae sp. 146||0.17029||BCI|
|Celastraceae||2 (0.00337)||Curculionidae sp. 3*||Curculionidae sp. 3*||0||CAM|
|Sister to Albizia||2 (0.00348)||Coelosternus cf. sp. 39||Eubulus cf. sp. 41||0.13008||BCI|
|Sister to Leptospermum scoparium||8 (0.01427)||Conotrachelus sp. 9*||Conotrachelus sp. 44*||0.1402||BCI|
|Sister to Leptospermum scoparium||0||Conotrachelus sp. 44*||Conotrachelus sp. 44*||0||BCI|
|Sister to Leptospermum scoparium||9 (0.01586)||Conotrachelus sp. 9||Conotrachelus sp. 44*||0.13974||BCI|
|Sister to Arabis alpina||2 (0.00510)||Cryptorhynchinae sp. 69||Cryptorhynchinae sp. 83||0.22667||BCI|
|Sister to Arabis alpina||2 (0.00510)||Cryptorhynchinae sp. 83||Sicoderus cf. delusor||0.21284||BCI|
|Arabis alpina clade||2 (0.00512)||Sicoderus cf. delusor||Cryptorhynchinae sp. 69||0.16987||BCI|
Genetic distances between the trnL sequences and their closest GenBank hit ranged from 0 to 10.7% (K2P corrected distances) (Table 1). The phylogenetic position for each query sequence relative to those available from GenBank could be established to various levels of accuracy, as taxon coverage and marker informativeness were not universally good. Phylogenetic analyses of each trnL sequence together with their respective top 100 GenBank entries recovered these sequences as members of well-established clades. The lowest taxonomic level to which the sequences could be assigned was family, tribe, and genus level in 5, 14, and 23 cases, respectively (Table 1). In most of these searches focal taxa were monophyletic, demonstrating the power of the trnL locus to place sequences according to established taxonomic groupings. Where the taxonomic groups were not recovered as monophyletic, as in the genus Acacia (Fabaceae), their paraphyly had already been established previously (Maslin et al., 2003; Miller et al., 2003) (Fig. 2). The trnL intron based inference was consistent with the morphological host-plant identification in the reared specimens, although due to the incomplete host database the corroboration was to family level only in several cases (Table 1).
|Subfamily||BMNH||Species||Site||Collection method||Host inferred|
|Family||Lowest level||PP||% K2P||Hypothesis based on local records and plant systematic literature|
|Brentidae||747250||Brentidae sp. 119||BCI||Manual||Meliaceae||Sister to Malleastrum sp.||0.54||0.8||Trichilia spp. (9 spp.)|
|747443||Brentidae sp. 120||BCI||Manual||Poaceae||Sister to Agrostis capillaris||0.92||0||Agrostis spp. (4 spp.)|
|747323||Brentidae sp. 124||BCI||Manual||Euphorbiaceae||Croton clade||1||1.4||Croton billbergianus|
|Trachelizinae||747322||Teramocerus sp. 118||BCI||Manual||Meliaceae||Sister to Malleastrum sp.||0.65||0.6||Trichilia spp. (9 spp.)|
|Dryophthoridae||796478||Dryophthoridae sp. 127||BCI||FIT||Sapindaceae||Sister to Serjania altissima||0.95||1.6||Serjania spp. (9 spp.)|
|Curculionidae||793058||Curculionidae sp. 1||SEL||Manual||Moraceae||Brosimum clade||0.99||1.8||Dorstenia (2 spp.) or Brosimum (3 spp.)|
|796412||Curculionidae sp. 2||BCI||Manual||Theaceae||Sister to Camellia japonica||0.6||Gordonia (2 spp.)|
|826571||Curculionidae sp. 3*||CAM||Reared||Celastraceae||Sister to Loeseneriella africana||1.9||Celastraceae|
|826574||Curculionidae sp. 3*||CAM||Reared||Celastraceae||Sister to Loeseneriella africana||1.6||Celastraceae|
|Baridinae||793067||Solaria curtula||SEL||Reared||Sister to Dilleniaceae||Sister to Tetracera asiatica||1||5.2||Sister to Dilleniaceae|
|Conoderinae||826540||Lechriops sp. 1||FOR||Reared||Myrsinaceae||Sister to Ardisia speciosa||0.83||1.6||Ardisia spp.|
|Cryptorhynchinae||747260||Coelosternus cf. sp. 39||BCI||Manual||Fabaceae: Mimosoideae||Sister to Albizia||1.1||Tribe Ingeae|
|747622||Cryptorhynchinae sp. 4||BCI||FIT||Rubiaceae||Sister to Faramea multiflora||1||2.9||Faramea (10 spp.) or Coussarea (4 spp.)|
|747286||Cryptorhynchinae sp. 24||BCI||Manual||Urticaceae||Pilea ternifolia, P. melastomoides clade||1||3.2||Pilea spp.|
|796424||Cryptorhynchinae sp. 34||BCI||Manual||Musaceae||Basal to Musa clade||0.98||0.2||Musa sapientum|
|747320||Cryptorhynchinae sp. 69||BCI||Manual||Brassicaceae||Arabis alpina clade||0.6||Cardamine flexuosa|
|826691||Cryptorhynchinae sp. 71||BCI||Manual||Rubiaceae||Basal to Hallea and Mitragyna||0.3||Naucleeae species|
|826713||Cryptorhynchinae sp. 71||BCI||Manual||Piperaceae||Piper aequale, P. urophyllum clade||1||2||Piper aequale clade|
|826712||Cryptorhynchinae sp. 77||BCI||Manual||Fabaceae||Ingeae tribe||1.3||Tribe Ingeae|
|747370||Cryptorhynchinae sp. 83||BCI||Manual||Brassicaceae||Sister to Arabis alpina||0.3||Cardamine flexuosa|
|796489||Cryptorhynchus sp. 111||BCI||Manual||Melastomataceae||Sister to Adelobotrys||0.96||2.5||Merianieae, Miconieae, Bertolonieae or Blakeeae|
|747380||Elytrocoptus sp. 87||BCI||Manual||Sapindaceae||Basal to Matayba and Scyphonychium||1.2||Allophylus, Paullinia, Sapindus, Thinouia|
|747261||Eubulus cf. sp. 41||BCI||Manual||Fabaceae: Mimosoideae||Sister to Albizia||1.4||Tribe Ingeae|
|Curculioninae||826065||Sibinia sp.*||BCI||Reared||Fabaceae: Mimosoideae||Basal to Mimosa pigra and M. tweedieana||0.95||0.1||Mimosa pellita var. pellita (=Mimosa pigra)|
|747414||Sicoderus cf. delusor||BCI||Manual||Brassicaceae||Sister to Arabis alpina||0.9||Cardamine flexuosa|
|Entiminae||796343||Entiminae sp. 2||CAM||Manual||Melastomataceae||Basal to Adelobotrys, Melastoma and Tibouchina||1||6.5||Memecylaceae, Crypteroniaceae, Alzateaceae, Rhynchocalycaceae, Penaeaceae, or Oliniaceae|
|796276||Entiminae sp. 3||FOR||Manual||Asteraceae||Asteroideae||1||0.7||Asteroideae|
|747283||Entiminae sp. 4||BCI||Manual||Malvaceae||Bombacoidae clade||0.9||1||Cavanillensia, Pseudobombax, Ceiba, Gyranthera|
|Molytinae||793141||Cholini sp. 195||SEL||Manual||Marantaceae||Basal to Ataenidia, Marantocloa and Phacelophrynium||0.95||2.9||Marantaceae|
|826663||Conotrachelus inexplicatus Faust||BCI||FIT||Rubiaceae||Basal to Calycophyllum||1||Cinchonoideae|
|793105||Conotrachelus sp. 1†||SEL||Manual||Sapotaceae||Sister to Pouteria vernicosa||0||Pouteria spp.|
|826553||Conotrachelus sp. 8*||BCI||Reared||Basal to Salicaceae||Basal to Salicaceae||0.7||1.7||Lacistemataceae, Turneraceae, Passifloraceae|
|826118||Conotrachelus sp. 9*||BCI||Reared||Myrtaceae||Sister to Leptospermum scoparium||2.4||Chamelaucieae, Lindsayomyrteae, Leptospermeae, Syncarpieae and Myrteae|
|825983||Conotrachelus sp. 44*||BCI||Reared||Myrtaceae||Sister to Leptospermum scoparium||3.6||Chamelaucieae, Lindsayomyrteae, Leptospermeae, Syncarpieae and Myrteae|
|825984||Conotrachelus sp. 44*||BCI||Reared||Myrtaceae||Sister to Leptospermum scoparium||3.6||Chamelaucieae, Lindsayomyrteae, Leptospermeae, Syncarpieae and Myrteae|
|796425||Conotrachelus sp. 158||BCI||Manual||Lauraceae||Sister to Persea americana||1.1||sister to Persea americana|
|796287||Conotrachelus sp. 162||BCI||Manual||Arecaceae||Sister to Satakentia liukiuensis||0||Arecoideae|
|793143||Hylobiini sp. 1||SEL||Manual||Urticaceae||Basal to Cecropia||1||6.1||Urticaceae|
|747365||Hylobiini sp. 137||BCI||Manual||Anacardiaceae||Sister to Anacardium occidental E||1||1||Anacardium excelsum, A. occidentale or Mangifera indica|
|796415||Hylobiini sp. 139||BCI||Manual||Bignoniaceae||Sister to Arrabidaea pubescens||1||0.6||Arrabidaea clade|
|747437||Molytinae sp. 146||BCI||Manual||Malvaceae||Sister to Chorisia speciosa||0.7||Cavanillensia, Pseudobombax, Ceiba, Gyranthera|
We were specifically interested in the utility of the trnL intron to discriminate between close relatives. The literature on DNA barcoding has already established that different species may exhibit identical sequences (Lahaye et al., 2008; Hollingsworth et al., 2009). This was confirmed here, e.g. in reference sequences of the family Fabaceae the GenBank entries for up to six species were identical. In our sample, only a single case of the weevil Sibinia sp. 1 reared from fruits that were morphologically identified as Mimosa pigra L. showed complete identity with a GenBank entry, also designated as M. pigra. The closest relative in GenBank, M. tweedieana was different by two base pairs (Fig. 2), indicating the high discriminatory power of this locus and the precision of the DNA-based identification in the genus Mimosa. In others, e.g. the closely related Conotrachelus sp. 44 and C. sp. 9 reared from Eugenia nesiotica Standl. and E. galalonensis (Wright), respectively, host sequences were distinguishable by nine nucleotide changes, while the two records from E. galalonensis showed identical sequences. However, two specimens of Curculionidae sp. 3 that were reared from fruits of the same species (an unclassified species in the family Celastraceae) differed by two base pairs, suggesting intraspecific variation of the trnL intron. Similarly, several wild-caught individuals (i.e. of unknown host plant) produced closely related sequences also (1 or 2 bp divergences; Table 2). These slightly divergent sequences may reflect variation within a single species, or indeed constitute evidence for feeding on closely related (sympatric) species that differ slightly at this locus.
Phylogenetic structuring of host use and floristic implications
In total, the set of 41 sequences could be ascribed to 26 plant families. All trnL sequences were used to build a phylogenetic tree of the host plants, which showed general agreement with known angiosperm relationships (Bremer et al., 2003), including the basal split of monocots, magnoliids, and all other angiosperms, the asterids, and groups within the eurosids I and II (which both were split in two subclades) (Fig. 3). When mapped on the tree of curculionoids, the host-plant tree showed a high level of incongruence indicating a general lack of host conservation at deeper levels (Fig. 4). Nearer to the tips, there was a small clade of Conotrachelus sp. 44 and Conotrachelus sp. 9 that was congruent (Fig. 4), but major shifts in host association were evident even at the species level, as the two representatives of Cryptorhynchinae sp. 71 fed on the distant families Piperaceae and Rubiaceae. Vice versa, where closely related plant sequences were obtained from two or three weevils, these were highly divergent (Table 2) and their mean level of divergence did not differ greatly from the average distances of all weevils (Table 2).
Host plants inferred from the trnL intron sequences were encountered that had not been reported from the collecting site. For example, M. pigra is not known from BCI, although M. tenuiflora (Willd.) Poir. does occur in the area. The latter species was included in the analysis but was positioned in a distant clade, confirming the presence of M. pigra (or a very closely related species that is indistinguishable in trnL intron) existing in BCI. The same applies to Conotrachelus sp. 1, whose host sequence matched a GenBank entry attributed to Pouteria vernicosa T.D. Pennington. This specimen was collected in La Selva, Costa Rica where species plant lists show no record of this species but report 11 other species of the genus Pouteria. Likewise, the genus Pilea (Urticaceae), the inferred host of Cryptorhynchinae sp. 24, has not been reported for BCI, although it is known from the surrounding Panama Canal Area. The host of Brentidae sp. 120 was inferred to be a close relative of the genera Agrostis, Calamagrostis, or Poa (Poaceae), none of which has been cited as occurring at BCI, although they are known from the wider area.
- Top of page
- Materials and methods
- Supporting Information
The diversity of insect herbivores remains a major challenge to the understanding of species richness and functioning of tropical forests. However, the difficulty in the establishment of host associations hampers the study of plant–herbivore interactions and their role in promoting tropical species richness. A recent review concluded: ‘For establishing feeding association, we see no viable alternatives to experimental feeding trials or direct feeding and rearing records' (Lewinsohn & Roslin, 2008). The proof of DNA-based plant identification from herbivore tissue (Jurado-Rivera et al., 2009) now provides a novel method with great potential for the study of rainforests, as the most complex plant–herbivore assemblages on Earth, that will resolve the long-standing questions about the factors promoting species diversity. Although based on a very small sample of herbivores, the current study demonstrates the great potential of this procedure for determining host associations and ultimately diet breadth of tropical insects.
Identification success and phylogenetic information content of trnL
Accurate identification of the host is affected by the completeness of the reference DNA database, as well as the discriminatory power of the locus used for sequencing. We chose the trnL intron mainly because it has the highest level of coverage in GenBank among potential barcoding markers; the universality of PCR primers; its good PCR success with degraded DNA (Taberlet et al., 2007); and its power in phylogenetic analysis across a range of hierarchical levels (Bremer et al., 2002; Shaw et al., 2005). The latter is due to the presence of conserved and highly variable regions and the relative ease of alignment. Whereas no comparisons were conducted with other potential markers (Hollingsworth et al., 2009), the trnL intron sequences were sufficient to discriminate among all individuals found in these forests, with the single exception of two individuals of Conotrachelus sp. 44 which were reared from a single host species. The trnL locus discriminated between congenerics based on several nucleotide changes (e.g. E. nesiotica and E. galalonensis; Table 2), while there was no intraspecific variation (E. galalonensis–E. galalonensis). However, discrimination at the species level was not universal, e.g. in the genus Acacia where up to six species are indistinguishable using this marker (see Fig. 2). In turn, we observed trnL variation within a morphologically defined host species in the two host records for Curculionidae sp. 3 (both reared from an unidentified species of Celastraceae). This intraspecific variation for the locus is not unprecedented and has been reported previously (Taberlet et al., 2007; Tsai et al., 2008). However, it does mean that in examples of 1- or 2-bp differences in trnL sequences we do not know whether this is associated with intraspecific or interspecific variation without being able to identify the plants in another way. Hence, as there is no clear ‘barcoding gap’ as that seen in mtDNA of animals (Meyer & Paulay, 2005), the use of trnL intron to resolve taxonomic identifications near the species level remains limited (see also Chase et al., 2007). Possibly this problem can be overcome with other chloroplast markers exhibiting faster substitution rates (Lahaye et al., 2008) or a set of markers (Tsai et al., 2008; Hollingsworth et al., 2009). However, the discrimination of host races and very closely related isolates (e.g. Hebert et al., 2004) may remain problematic with the use of chloroplast markers.
The problem of low interspecific variation was relevant for identifications in the current sample only in a few cases, as coverage in GenBank was generally not sufficient to provide matches or near matches to sequences obtained here. Instead, phylogenetic inferences relative to the top 100 GenBank entries had to be used to place the query sequence. Frequently, host sequences could only be assigned to a plant family or subfamily, in particular when applying our rather stringent criteria for identification (based on the Bayesian inference) as a group member that required high support levels of a subtending node. The trnL intron marker proved surprisingly powerful for phylogenetic reconstruction of relationships of host plant lineages over a wide hierarchical range. The locus separated major groups of flowering plants (Fig. 3) and established genus-level relationships in accordance with recent DNA-based studies (Fig. 2). This was important to provide accurate high-level identifications (Jurado-Rivera et al., 2009), which were confirmed where these inferences were applied to reared specimens of known host associations (Table 1). With greater taxonomic coverage, the trnL locus provided increased precision of identification, as in the case of the densely sampled genus Mimosa that permitted species-level identification (Fig. 2). At the current state of the databases, low taxon coverage of trnL intron sequences, rather than short fragment length or limited sequence variability, had the greatest impact on identification success.
Host identification will gradually become more accurate, as database content accumulates from studies of host use (Jurado-Rivera et al., 2009), DNA barcoding (Taberlet et al., 2007), and phylogenetic analysis (Bremer et al., 2002; Shaw et al., 2005). Eventually, even species-level questions may be addressed with greater accuracy once DNA-based analyses of species boundaries and intraspecific variation have been conducted. In addition, a combined set of cpDNA and other markers (including single-copy nuclear markers) will be amplifiable from the herbivore tissue also. In addition to improving species discrimination, such multiple-marker system will also address discrimination for those lineages with poor taxon coverage in the trnL intron. Therefore, a narrow focus on levels of variation in developing DNA barcodes (Lahaye et al., 2008) should not ignore the need for phylogenetic power that ultimately will put sequences in the context of others. Phylogenetics will be important for identification of taxa not represented in the database as well as for evolutionary analyses of ecological, biogeographic, behavioural and other data that may be associated with the individuals from which the sequences were obtained, or for co-evolutionary analyses of plants and their herbivores or pathogens (also see Jurado-Rivera et al., 2009).
Composition of the sample and its relevance for larger DNA-based surveys of host plants
The trnL intron sequence provided a feeding record for a particular individual. While such individual records do not directly investigate diet breadth, as feeding records build up, they will provide the spectrum of potential host plants from which diet breadth can be derived. Quantitative sampling will also indicate food preferences within the host range. The DNA largely conveys information on the most recent feeding episode. It is known that ingested DNA decreases exponentially in abundance over a period of 4–50 h (Chen et al., 2000; Greenstone et al., 2007; King et al., 2008) and therefore obtaining an individual at different times during the life cycle or feeding cycle might have resulted in a different host association. This is also of significance for trap catches, as the amplification of plant DNA from trap catches failed frequently in this study. It is possible (although not tested) that the plant DNA in the insect gut degrades over the period (sometimes of several days) that the insect spends in the comparatively weak preservative in the trap.
The current study demonstrated the interest of findings from the technique. First, from each of our samples only a single sequence per individual was obtained, i.e. there was no intra-individual variation and hence these feeding records were unequivocal for that individual without the need for cloning of PCR products. Second, weevils exploited a broad range of host-plant families showing a high trophic diversity across most major groups of angiosperms. While some host sequences were closely related (Table 2) and might indeed be from the same host species, the sequence variation still indicates a diversity of host trees and populations. Third, host records may be highly variable for a species when more than one individual was available for analyses (e.g. results for Cryptorhynchinae sp. 71), but host conservation derived from high sequence similarity is typically found (e.g. Conotrachelus sp. 44 and its sister species Conotrachelus sp. 9, or Curculionidae sp. 3). Fourth, host specificity was not phylogenetically conserved in the studied group of weevils, as sister taxa usually feed on different plant families, and generally very distant ones phylogenetically, and reciprocally with similar host plants being used by highly divergent herbivores (Fig. 3).
How much these host records reflect the diet breadth of these herbivore species, and to what extent this demonstrates divergent larval and adult host preferences and diet breadth remains to be assessed based on the study of further specimens. In this respect, currently larval host associations are hypothesised to be more restricted than those of adults (Novotny & Basset, 2005). While our ecological knowledge is limited regarding the breadth of adult versus larval hosts and variation among subsequent food intakes, DNA-based feeding records will be able to build an increasingly complete image of feeding behaviour from comprehensive sampling of different stages representing particular species across seasonal and geographical ranges. The cases encountered here illustrate extremes in the spectrum of host specificities, which may be narrow in some species, and (phylogenetically) broad in others (Symons & Beccaloni, 1999). Increased taxonomic coverage and larger sample sizes per species will make it possible to quantify host specificity in the future, from monophagous to oligophagous or polyphagous and from evolutionarily conserved to labile, while also taking into account local differences in response to beta-diversity of host plant assemblages (e.g. Condon et al., 2008).
Finally, the DNA-based feeding records may reveal discrepancies in collecting site and actual food plants. For example, our sample of Conotrachelus sp. 1 was obtained from an individual walking on fruits of Inga alba (Fabaceae), a frequent host for this genus, but the DNA-based inference of feeding source was the distantly related asterid Sapotaceae (also a known host of Conotrachelus). In addition, the comparison of species encountered at a site may be compared with floristic inventories, which already pointed to the presence of unreported plant species in an area. The host records can potentially be used to discriminate between host trees of different genotypes at the population level, providing a measure of dispersal of the insects between rain forest sites.
Phylogenetic information content of cox1 and coevolutionary relationships
The use of cox1 in identification and species delimitation in insects is well established (Hebert et al., 2004; Pons et al., 2006). Levels of sequence divergence in cox1 near the species level were much higher than in the trnL intron, and detected intraspecific variation as clusters of closely related sequences. The cox1 marker was similarly powerful in separate deep-level groups (families and subfamilies) in Curculionoidea (Fig. 1) and resolved relationships within those, e.g. recovering the basal branching pattern in Conotrachelus. Sequence divergences in mitochondrial cox1 and chloroplast trnL were more similar at deeper levels, presumably because mtDNA is affected by saturation of nucleotide variation. This may also compromise its power at basal levels of the tree, e.g. resulting in the failure to recover Dryophthoridae outside of the Curculionidae [which is the main incongruence of the tree with the existing classification, although a multi-gene data set also found this position for Dryophthoridae (Hunt et al., 2007; McKenna et al., 2009)]. However, the precise resolution of basal relationships of Curculionoidea were not of great concern for the current study, as coevolutionary analyses (below) were mostly affected by host switches nearer the tips of the trees. Therefore, more critical than adding markers (which potentially could be obtained from existing phylogenetic datasets) the power of the current analyses would be improved with denser taxonomic sampling.
Virtually all of the terminal branches of the curculionid tree were very long relative to the internal branches, indicating that clade diversity was not captured to any degree of completion. According to the dating procedure all pairs of sister taxa were separated by >20 Mya. The absolute calibration using the universal 2.3% divergence My−1 is problematic and its utility needs to be confirmed for dating older nodes, while confidence intervals may be large given the slow convergence of the MCMC chain. It is noteworthy, however, that the basal split of Brentidae and Curculionidae estimated to 150 Mya was in almost perfect agreement with a recent age estimation of this node considering multiple genes and fossil calibration across the Curculionoidea (McKenna et al., 2009). Therefore, deep branches of the tree obtained here, e.g. resolving the basal lineages in Cryptorhynchinae, are ancient and probably represent many thousands of species of that subfamily missing from the trees. Likewise, the genus Conotrachelus represented here by just eight species includes more than 1200 described from the Neotropics and many more undescribed species. Inevitably the incomplete taxon sampling limits the evolutionary analysis of character changes, e.g. to study the rate of host switches and conservation within clades. However, it is clear that host plant–herbivore interactions in the rainforest assemblage are evolutionarily fluid. Although the estimates for appearance of the earliest crown group angiosperms at >140 Mya (Friis et al., 2006) are similar to the estimated age of the weevils, host associations are unlikely to be ancient, and host shifts and lags in diversification of herbivores are well established (Lopez-Vaamonde et al., 2006; Gómez-Zurita et al., 2007; Hunt et al., 2007; McKenna et al., 2009). However, much denser taxon sampling is required to assess the frequency and step size of host shifts.
- Top of page
- Materials and methods
- Supporting Information
Ingested chloroplast DNA that is obtained from the insect specimens using standard extraction protocols provides the information on host records necessary for evolutionary and ecological analyses of herbivore–host plant interactions (Jurado-Rivera et al., 2009). The current study assesses the utility of this method when applied to a highly complex rainforest assemblage. Evident limitations for host identification are the insufficient local host database and the lack of clearly known species limits in the hosts, i.e. the trnL sequences do not always provide an unequivocal host record. While future increase of the database and use of additional chloroplast markers will improve the accuracy, some apparent limitations of host plant inferences are independent of the study method. Feeding studies of tropical forest assemblages (e.g. Barone, 2000; Novotny et al., 2002) to date have not questioned taxonomy and species limits of host plants, nor assessed population differentiation and geographic turnover. Due to their greater resolution, DNA-based analyses can contribute vital information on host populations and spatial differentiation of host use. The technique therefore permits the reinvestigation of pertinent hypotheses explaining tropical forest diversity, e.g. regarding the greater trophic specialisation and higher species diversity in the tropics (Coley & Barone, 1996), density-dependent factors to maintain high diversity (Janzen, 1970), or the correlation of herbivore diversity with phylogenetic diversity of host plants (Novotny et al., 2006). Solid data on the host-plant use and host specificity are a necessity for testing any of these hypotheses.
- Top of page
- Materials and methods
- Supporting Information
We are grateful to Anna Papadopoulou for her important input to this work. We thank Osvaldo Calderon and Salomon Aguilar for their help with botanical information and plant identifications. We also thank Hugh Loxdale and Michael Claridge for suggesting the topic of our symposium contribution. We are grateful to the STRI staff who helped with facilities and ANAM collection permits to S. P. N. We thank Dr Richard Thompson for his help with weevil identification. This work has been supported by the studentships from SENACYT-Panama, CONACYT-Mexico (196749) and the European Union-Alban (E05D058103PA) to S. P. N.
J. J. R. received support from the SYNTHESYS grant (GB-TAF-4394) of the European Community Research Infrastructure and J.G-Z. from the Spanish “Fundación BBVA” (project >IN2-BST-Nicaragua).
- Top of page
- Materials and methods
- Supporting Information
- 1999) A World Catalogue of families and genera of Curculionoidea (Insecta: Coleoptera) (excepting Scolytidae and Platypodidae). Entomopraxis, Barcelona, Spain. & (
- 1990) Basic local alignment search tool. Journal of Molecular Biology, 215, 403–410. , , , & (
- 1998) Host-specificity of folivorous insects in a moist tropical forest. Journal of Animal Ecology, 67, 400–409. (
- 2000) Comparison of herbivores and herbivory in the canopy and understory for two tropical tree species. Biotropica, 32, 307–317. (
- 2003) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society, 141, 399–436. , , , , et al. (
- 2002) Phylogenetics of asterids based on 3 coding and 3 non-coding chloroplast DNA markers and the utility of non-coding DNA at higher taxonomic levels. Molecular Phylogenetics and Evolution, 24, 274–301. , , , , , et al. (
- 1994) Rapid morphological radiation and convergence among races of the butterfly Heliconius erato inferred from patterns of mitochondrial DNA evolution. Proceedings of the National Academy of Sciences of the United States of America, 91, 6491–6495. (
- 2007) A proposal for a standardised protocol to barcode all land plants. Taxon, 56, 295–299. , , , , , et al. (
- 2007) Differential electroantennogram response of females and males of two parasitoid species to host-related green leaf volatiles and inducible compounds. Bulletin of Entomological Research, 97, 515–522. & (
- 2000) Identifying key cereal aphid predators by molecular gut analysis. Molecular Ecology, 9, 1887–1898. , , & (
- 1996) Herbivory and plant defenses in tropical forests. Annual Review of Ecology and Systematics, 27, 305–335. & (
- 2008) Hidden neotropical diversity: greater than the sum of its parts. Science, 320, 928–931. , , & (
- 1978) Flora of Barro Colorado Island. Stanford University Press, Stanford, California. (
- 1984) Foodplant selection and induction of feeding preference among host and non-host plants in larvae of the Tobacco Hornworm Manduca sexta. Entomologia Experimentalis et Applicata, 35, 177–193. & (
- 2007) BEAST: Bayesian evolutionary analysis by sampling trees. MBC Evolutionary Biology, 7: 214. doi:10.1186/1471-2148-7-214. & (
- 2007) Host specificity of Lepidoptera in tropical and temperate forests. Nature, 448, 696–699. , , , , , et al. (
- 2008) The influence of soil type on rain forest insect herbivore communities. Biotropica, 40, 707–713. , & (
- 1982) Tropical forests: their richness in Coleoptera and other arthropod species. Coleopterists Bulletin, 36, 74–75. (
- 1998) ‘Inordinate fondness' explained: why are there so many beetles? Science, 281, 555–559. (
- 2007) Electrophysiological responses of the blue willow leaf beetle, Phratora vulgatissima, to volatiles of different Salix viminalis genotypes. Entomologia Experimentalis et Applicata, 125, 157–164. , , & (
- 2006) Cretaceous anglosperin flowers: innovation and evolution in plant reproduction. Palaeogeography, Palaeoclimatology, Palaeoecology, 232, 251–293. , & (
- 1978) Grasshopper food web analysis: use of carbon isotope ratios to examine feeding relationships among terrestrial herbivores. Ecology, 59, 498–506. , & (
- 2007) Feeding mode and prey detectability half-lives in molecular gut-content analysis: an example with two predators of the Colorado potato beetle. Bulletin of Entomological Research, 97, 201–209. , , , & (
- 2007) Recalibrating the tree of Leaf Beetles (Chrysomelidae): independent diversification of angiosperms and their insect herbivores. PLoS ONE, 2, e360. , , & (
- 2005) The promise of DNA barcoding for taxonomy. Systematic Biology, 54, 852–859. & (
- 2004) Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. Proceedings of the National Academy of Sciences of the United States of America, 101, 14812–14817. , , , & (
- 2009) A DNA barcode for land plants. Proceedings of the National Academy of Sciences of the United States of America, 106, 12794–12797. , , , , , et al. (
- 2007) A comprehensive phylogeny of beetles reveals the evolutionary origins of a superradiation. Science, 318, 1913–1916. , , , , , et al. (
- 1970) Herbivores and number of tree species in tropical forests. American Naturalist, 104, 501–528. (
- 2009) DNA barcoding insect–hostplant associations. Proceedings of the Royal Society Series B, 276, 639–648. , , , & (
- 2005) MAFFT version 5: improvement in accuracy of multiple sequence alignment. Nucleic Acids Research, 33, 511–518. , , & (
- 2008) Molecular analysis of predation: a review of best practice for DNA-based approaches. Molecular Ecology, 17, 947–963. , , & (
- 2008) DNA barcoding the floras of biodiversity hotspots. Proceedings of the National Academy of Sciences of the United States of America, 105, 2923–2928. , , , , , et al. (
- 1988) Declining acceptability of a food plant for the polyphagous grasshopper Schistocerca americana–the role of food aversion learning. Physiological Entomology, 13, 291–301. & (
- 2008) Four ways towards tropical herbivore megadiversity. Ecology Letters, 11, 398–416. & (
- 2005) Insects on plants: diversity of herbivore assemblages revisited. Annual Review of Ecology, Evolution and Systematics, 36, 597–620. , & (
- 2006) Fossil-calibrated molecular phylogenies reveal that leaf-mining moths radiated millions of years after their host plants. Journal of Evolutionary Biology, 19, 1314–1326. , , , , & (
- 2003) Overview of the generic status of Acacia (Leguminosae: Mimosoideae). Australian Systematic Botany, 16, 1–18. , & (
- 2007) A PCR method for detection of plant meals from the guts of insects. Organisms Diversity and Evolution, 7, 294–303. , , , , , et al. (
- 2009) Temporal lags and overlap in the diversification of weevils and flowering plants. Proceedings of the National Academy of Sciences of the United States of America, 106, 7083–7088. , , & (
- 2005) DNA barcoding: Error rates based on comprehensive sampling. PLoS Biology, 3, 2229–2238. & (
- 2003) Molecular phylogenetics of the Australian acacias of subg. Phyllodineae (Fabaceae: Mimosoideae) based on the trnK intron. Australian Journal of Botany, 51, 167–177. , & (
- 2000) Rare species in communities of tropical insect herbivores: pondering the mystery of singletons. Oikos, 89, 564–572. & (
- 2005) Review–host specificity of insect herbivores in tropical forests. Proceedings of the Royal Society B: Biological Sciences, 272, 1083–1090. & (
- 2002) Low host specificity of herbivorous insects in a tropical forest. Nature, 416, 841–844. , , , , , et al. (
- 2006) Why are there so many species of herbivorous insects in tropical rainforests? Science, 313, 1115–1118. , , , , , et al. (
- 2007) Low beta diversity of herbivorous insects in tropical forests. Nature, 448, 692–695. , , , , , et al. (
- 2000) How many species of arthropods? Erwin's estimate revised. Biological Journal of the Linnean Society, 71, 583–597. (
- 1976) On feeding patterns in desert grasshoppers and the evolution of specialized diets. Proceedings of the Academy of Natural Sciences of Philadelphia, 128, 89–126. & (
- 2006) Sequence-based species delimitation for the DNA taxonomy of undescribed insects. Systematic Biology, 55, 595–609. , , , , , et al. (
- 1998) MODELTEST: testing the model of DNA substitution. Bioinformatics, 14, 817–818. & (
- 2002) The long and short of food-chain length. Trends in Ecology & Evolution, 17, 269–277. (
- 2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics, 19, 1572–1574. & (
- 1999) Sugar meals and longevity of the sandfly Phlebotomus papatasi in an arid focus of Leishmania major in the Jordan Valley. Medical and Veterinary Entomology, 13, 65–71. & (
- 1957) Elektrophysiologische untersuchungen von chemo- und mechanorezetoren der antenne des seidenspinners, Bombyx mori L. Zeitschrift fur Vergleichende Physiologie, 40, 8–41. (
- 2005) The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany, 92, 142–166. , , , , , et al. (
- 1994) Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Annals of the Entomological Society of America, 87, 651–701. , , , , & (
- 2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22, 2688–2690. (
- 2008) A rapid bootstrap algorithm for the RAxML web servers. Systematic Biology, 57, 758–771. , & (
- 1999) Phylogenetic indices for measuring the diet breadths of phytophagous insects. Oecologia, 119, 427–434. & (
- 2007) Power and limitations of the chloroplast trnL (UAA) intron for plant DNA barcoding. Nucleic Acids Research, 35, Article No.: e14.10.1093/nar/gk1938. , , , , , et al. (
- 1991) Universal primers for amplification of 3 noncoding regions of chloroplast DNA. Plant Molecular Biology, 17, 1105–1109. , , & (
- 2008) Bidens identification using the noncoding regions of chloroplast genome and nuclear ribosomal DNA. Forensic Science International: Genetics, 2, 35–40. , , , , & (
- 2006) Phylogenetic dispersion of host use in a tropical insect herbivore community. Ecology, 87, S62–S75. , , , & (
- Top of page
- Materials and methods
- Supporting Information
Additional Supporting Information may be found in the online version of this article:
Table S1. Specimens from the superfamily Curculionoidea tested for trnL intron, not yielding good PCR products.
Table S2. Voucher specimens and accession numbers of weevils cox1 and plant trnl intron.
Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
|EEN_1145_sm_tables1.pdf||63K||Supporting info item|
|EEN_1145_sm_tables2.pdf||68K||Supporting info item|
Please note: Neither the Editors nor Wiley Blackwell are responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.