The guts of beetle larvae constitute a complex system where relationships among fungi, bacteria, and the insect host occur. In this study, we collected larvae of five families of wood-feeding Coleoptera in tropical forests of Costa Rica, isolated fungi and bacteria from their intestinal tracts, and determined the presence of five different pathways for lignocellulolytic activity. The fungal isolates were assigned to three phyla, 16 orders, 24 families, and 40 genera;
Plant cell walls are predominantly composed of lignin, cellulose, and hemicellulose. Together, these three polymers represent one of the most abundant sources of renewable energy on Earth [
Recently there has been an increasing interest in the gut microorganisms of wood-feeding Coleoptera, since this microbial-host interaction is highly relevant from several perspectives. For example, in natural ecosystems, beetles and their associated microorganisms perform important functions as prime contributors to the degradation of organic matter [
Most of the 300,000 beetle species described to date occur in tropical rainforests [
The substrates on which the insects feed are major determinants for the gut microbial diversity. However, it is also possible that certain microbes have adapted to the endointestinal lifestyle and have developed mutualistic relationships necessary for the host survival. Hence, a fraction of the endosymbiont microbial community could be vertically transmitted [
Most of the previous studies on the microbial diversity of the coleopteran gut highlighted either bacterial diversity or, to a lesser extent, fungal diversity, generally obtained from a small number of economically important beetle species. In addition, some studies used metagenomic profiling for the discovery of novel lignocellulolytic genes. The purpose of the present work was to isolate and describe the composition of the cultivable fungi and bacteria associated with the guts of wood-feeding larvae of five families of Coleoptera and to determine their lignocellulolytic activities. For this, we collected larvae in tropical wet forests of several national parks in Costa Rica, isolated both fungi and bacteria from their guts, performed bioinformatics analysis, and assayed for the presence of five enzymatic activities related to the degradation of lignocellulosic materials. This work represents an initial step toward understanding relationships existing among the fungi, the bacteria, and the beetle host and the surrounding environment in a country having a particularly large biodiversity of Coleoptera.
This study was conducted in tropical wet forests of 10 protected areas of Costa Rica with the respective permit resolutions R-CM-INBio-40-2008-OT and R-CM-INBio-48-2008-OT of the National Authority of the Ministry of Environment. At each site, approximately 3 km of natural trails was explored, looking for decaying fallen trees within 25 m on each side of the path. The sampling area represented nearly 150,000 m2 (or 0.15 km2) per national park. Every fallen tree found was exhaustively examined for the presence of galleries of wood-feeding beetle larvae, particularly from the Scarabaeidae, Passalidae, Elateridae, Cerambycidae, and Tenebrionidae families. The selected national parks covered most of the natural distribution of the five coleopteran families studied: wet tropical forest ranging from zero to 1,300 m altitude (Table
Description of the location and main environmental parameters of the 10 national parks of Costa Rica, where sampling was carried out. The selected environments are classified as tropical wet forest and cover most of the natural distribution of the five coleopteran families studied.
National park | Latitude |
Altitude (m) | Mean temperature (°C) | Annual precipitation (mm) |
---|---|---|---|---|
Arenal | 10°26′49′′ N |
589 | 24 | 4000–5000 |
|
||||
Barbilla | 9°58′43′′ N |
460 | 21 | 3000–4000 |
|
||||
Braulio Carrillo | 10°9′33′′ N |
507 | 24 | 3500–4500 |
|
||||
Carara | 9°46′41′′ N |
78 | 27 | 2500–3000 |
|
||||
Hitoy Cerere | 9°40′18′′ N |
150 | 25 | 3000–4000 |
|
||||
Piedras Blancas | 8°41′56′′ N |
198 | 28 | 5000–6000 |
|
||||
Rincón de la Vieja | 10°46′29′′ N |
782 | 22 | 2500–3000 |
|
||||
Tapanti | 9°44′40′′ N |
1287 | 19 | 6000–7000 |
|
||||
Tenorio | 10°42′25′′ N |
727 | 22 | 3000–4000 |
|
||||
Tortuguero | 10°32′5′′ N |
0 | 26 | 5000–6000 |
The observed frequency of fallen trees and the presence of insect galleries within them varied according to the conditions of each site. Therefore, sample collection and study results were normalized to unit area. Most of the insect galleries contained insects of only one coleopteran family, but in few cases, two or three different families were present. The insect larvae, in the late second or third larval instars, were collected, placed in polyethylene boxes together with pieces of their feeding wood, and kept at ambient temperature until being transported to the laboratory (Table
Distribution of the number of larvae samples according to the insect family and national park. At each site, an approximate area of 150.000 m2 was explored for the presence of wood-feeding larvae. The number of insect groups found varied according to the natural condition of each forest. Each sample was composed of one to three individuals of the same species.
National park | Cer | Ela | Pas | Sca | Ten | Total |
---|---|---|---|---|---|---|
Arenal | 1 | 3 | 4 | |||
Barbilla | 6 | 6 | ||||
Braulio | 2 | 2 | 1 | 5 | ||
Carara | 1 | 3 | 1 | 5 | ||
Hitoy Cerere | 7 | 7 | ||||
Piedras Blancas | 1 | 2 | 1 | 4 | ||
Rincón de la Vieja | 1 | 2 | 3 | |||
Tapanti | 2 | 1 | 1 | 1 | 5 | |
Tenorio | 1 | 1 | 1 | 3 | ||
Tortuguero | 1 | 2 | 2 | 1 | 6 | |
|
||||||
Total | 8 | 5 | 16 | 16 | 3 | 48 |
Cer: Cerambycidae, Ela: Elateridae, Pas: Passalidae, Sca: Scarabaeidae, and Ten: Tenebrionidae.
The entire gut was removed from each larva and placed on a sterile Petri dish, crushed, and spread onto three different media plates. For isolation of fungi we used Potato-Dextrose Agar (PDA, Difco) amended with chlortetracycline (120 mg/L) and streptomycin (120 mg/L). For the isolation of bacteria we used one-third strength Luria-Bertani medium (3 g/L peptone, 5 g/L yeast-extract, 10 g/L NaCl, and 15 g/liter agar, pH 7.0) and the self-developed medium called LIGNO (1.5 g/L KH2PO4, 1.75 g/L K2HPO4, 0.8 g/L KNO3, 0.5 g/L MgSO4, 1 mL/L CaCl2 0.1 M, 4 g/L sawdust, 2 g/L bagasse powder, and 17 g/L agar, pH 7.0). Plates were incubated at 28°C for up to three weeks and checked every other day regularly for visible microorganism growths. Each emerging fungus was transferred into a fresh PDA plate amended with the antibiotics mentioned above, while each emerging bacterial colony was replated into Luria-Bertani medium (Difco). An initial screening, based on the morphological traits of the fungi, was performed to discard redundant isolates from the same sample (characteristics such as the color, shape, border type, mycelial density, presence-absence of secretions, and growth rate were compared). A second screening was based on molecular taxonomy. The resulting nonredundant isolates were included in a database with associated information and were preserved in the National Biodiversity Institute’s culture collection.
We screened all the isolates for the presence of five different pathways for lignocellulolytic activity possibly associated with degradation of structural wood components, including cellulose, lignin,
All the isolates were grown in Petri dishes containing the same media used for preservation. For the fungal DNA extraction, 400 mg of mycelia was ground with mortar and pestle in liquid nitrogen and further extracted using the DNeasy Plant kit (Qiagen, USA), including a pretreatment step consisting of the incubation at 60°C for one hour with 400
The taxonomic assignment of the bacterial sequences was performed by comparing the database against the 16S rRNA reference set 10 implemented in the Classifier tool of the Ribosomal Database Project, which assigned the 16S rRNA sequences to corresponding taxonomical hierarchy based on a naïve Bayesian rRNA classifier [
The analysis of the microbial communities was performed using the Vegan package implemented in the statistical programming environment and language R [
In this study, we isolated 92 fungal strains and 135 bacterial strains from larvae of five families of Coleoptera that were feeding on decaying wood in tropical wet forests of 10 national parks of Costa Rica. The 92 fungal isolates were assigned to three phyla, 16 orders, 24 families, and 40 genera (one different genus every 2.3 isolates). Within the phylum Zygomycota, we isolated members of the order Mucorales and within the phylum Basidiomycota members of the orders Agaricales, Polyporales, and Trichosporonales. Most of the fungi isolated from the gastrointestinal tracts of the larvae belonged to the phylum Ascomycota (89% of total). They were distributed in 12 orders, with Hypocreales being the dominant one; it comprised nearly 55% of the isolates (Table
Taxonomic distribution of the fungal isolates identified in this study. The number of isolates at the order and genera level is shown for each of the coleopteran families.
Order | Genus | Cer | Ela | Pas | Sca | Ten | Total |
---|---|---|---|---|---|---|---|
Botryosphaeriales |
|
1 | 1 | ||||
|
|||||||
Capnodiales |
|
1 | 1 | ||||
|
|||||||
Chaetothyriales |
|
2 | 2 | ||||
|
1 | 1 | |||||
|
1 | 1 | |||||
|
|||||||
Diaporthales |
|
1 | 1 | ||||
|
|||||||
Eurotiales |
|
1 | 1 | ||||
|
1 | 1 | 1 | 3 | |||
|
3 | 1 | 1 | 5 | |||
|
|||||||
Helotiales |
|
1 | 1 | ||||
|
|||||||
Hypocreales |
|
1 | 1 | ||||
|
1 | 1 | 2 | ||||
|
1 | 1 | |||||
|
1 | 1 | |||||
|
1 | 1 | |||||
|
1 | 1 | |||||
|
1 | 1 | 2 | ||||
|
1 | 1 | |||||
|
1 | 1 | 2 | ||||
|
1 | 1 | |||||
|
1 | 2 | 3 | ||||
|
1 | 1 | |||||
|
2 | 3 | 2 | 7 | |||
|
2 | 2 | |||||
|
1 | 1 | |||||
|
6 | 4 | 4 | 8 | 1 | 23 | |
|
|||||||
Microascales |
|
1 | 1 | 2 | |||
|
1 | 1 | 2 | ||||
|
1 | 1 | |||||
|
|||||||
Ophiostomatales |
|
2 | 2 | 4 | |||
|
|||||||
Pleosporales |
|
1 | 1 | ||||
|
|||||||
Saccharomycetales |
|
1 | 1 | 2 | |||
|
|||||||
Xylariales |
|
1 | 1 | ||||
|
1 | 1 | 2 | ||||
|
|||||||
Agaricales |
|
2 | 1 | 3 | |||
|
|||||||
Polyporales |
|
1 | 1 | ||||
|
1 | 1 | |||||
|
|||||||
Trichosporonales |
|
1 | 1 | 2 | |||
|
|||||||
Mucorales |
|
1 | 1 | 2 | |||
|
1 | 1 | |||||
|
|||||||
Total |
|
|
|
|
|
|
Cer: Cerambycidae, Ela: Elateridae, Pas: Passalidae, Sca: Scarabaeidae, and Ten: Tenebrionidae.
Most fungal orders and genera were sparsely represented, with 68% of the orders and 55% of the genera found associated with a specific coleopteran family at a particular site. All the insect families also presented unique fungal isolates per site, with Tenebrionidae being the only coleopteran family in which all isolates were phylogenetically distinct. Regarding the sampling sites, Tenorio National Park showed the highest number of unique phylotypes, but Piedras Blancas National Park contained a more phylogenetically diverse array of isolates.
The 135 bacterial isolates were classified within five phyla, 13 orders, 22 families, and 35 genera (one different genus from every 3.8 isolates), including members of Actinobacteria, Proteobacteria, Firmicutes, Flavobacteria, and Fusobacteria. Approximately 82% of the bacteria belonged to
Taxonomic distribution of the bacterial isolates obtained in this study. The number of isolates at the phylum and genera level is shown for each of the coleopteran families.
Class | Genus | Cer | Ela | Pas | Sca | Ten | Total |
---|---|---|---|---|---|---|---|
Actinobacteria |
|
1 | 1 | ||||
|
1 | 1 | |||||
|
1 | 1 | |||||
|
3 | 3 | |||||
|
1 | 1 | |||||
|
2 | 5 | 7 | ||||
|
1 | 1 | |||||
|
|||||||
|
|
1 | 1 | ||||
|
2 | 2 | |||||
|
|||||||
|
|
1 | 1 | 2 | |||
|
1 | 1 | |||||
|
|||||||
|
|
2 | 4 | 1 | 3 | 10 | |
|
1 | 1 | |||||
|
1 | 1 | |||||
|
2 | 2 | |||||
|
1 | 1 | |||||
|
4 | 4 | 1 | 3 | 12 | ||
|
1 | 1 | |||||
|
1 | 1 | 2 | ||||
|
1 | 1 | |||||
|
2 | 1 | 1 | 4 | 1 | 9 | |
|
3 | 1 | 4 | ||||
|
2 | 2 | |||||
|
4 | 2 | 2 | 2 | 2 | 12 | |
|
1 | 1 | 2 | ||||
|
|||||||
Firmicutes |
|
4 | 4 | 6 | 10 | 3 | 27 |
|
3 | 2 | 5 | ||||
|
1 | 5 | 2 | 8 | |||
|
1 | 2 | 4 | 7 | |||
|
1 | 1 | 1 | 3 | |||
|
1 | 1 | 2 | ||||
|
|||||||
Flavobacteria |
|
1 | 1 | ||||
|
|||||||
Fusobacteria |
|
1 | 1 | ||||
|
|||||||
Total |
|
|
|
|
|
|
Cer: Cerambycidae, Ela: Elateridae, Pas: Passalidae, Sca: Scarabaeidae, and Ten: Tenebrionidae.
The remaining bacterial classes obtained in this study were less represented. For example, members of Actinobacteria accounted for 11% of the isolates whereas
Nearly 65% of the fungal genera and 48% of the bacterial genera presented positive results in at least one of the five lignocellulolytic activities evaluated, with carboxymethylcellulose degradation being the most common activity observed in both groups (Table
Results of the screening for lignocellulolytic activities. Fungal genera with positive results are shown in the upper group and bacterial genera in the lower group.
Genus | CMC | lignin |
|
|
celob |
---|---|---|---|---|---|
|
+ | + | + | + | |
|
+ | + | + | + | + |
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | + | |||
|
+ | ||||
|
+ | + | + | + | |
|
+ | ||||
|
+ | + | + | ||
|
+ | ||||
|
+ | + | + | + | |
|
+ | + | + | ||
|
+ | + | + | + | |
|
+ | + | + | ||
|
+ | + | |||
|
+ | + | |||
|
+ | ||||
|
+ | + | + | ||
|
+ | ||||
|
+ | + | + | ||
|
+ | + | |||
|
+ | ||||
|
+ | + | + | + | + |
|
+ | + | + | + | + |
|
+ | ||||
|
|||||
|
+ | + | + | + | |
|
+ | + | + | + | |
|
+ | ||||
|
+ | + | + | + | |
|
+ | + | + | ||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ | ||||
|
+ |
CMC: cellulase activity on carboxymethylcellulose, lignin: ligninolytic activity on Remazol Brilliant Blue R,
We performed community analysis with Vegan to gain insight into how the microbial gut composition of the beetle families related to one another. This approach clustered the environments according to Bray-Curtis distances of the abundance distribution of the isolates, considering also their phylogenetic relationships. The results showed that the fungal composition of the isolates associated with larvae of Cerambycidae, Scarabaeidae, and Passalidae clustered together; Cerambycidae and Passalidae shared one order of Basidiomycota and three orders of Ascomycota, while Scarabaeidae and Passalidae had in common four orders of Ascomycota. A second cluster was formed by the fungal microbiotas isolated from Tenebrionidae and Elateridae; they shared two orders of Ascomycota and one of Basidiomycota (Figure
Heatmap of the abundance distribution of fungal communities associated with the guts of five wood-feeding families of Coleoptera. The taxonomic relationship of the fungal genera is shown in the rows, while the clustering of the coleopteran families, determined by their composition similarities, is shown in the columns. Higher intensities of the color reveal higher abundances of the isolates.
Heatmap of the abundance distribution of bacterial communities associated with the guts of five wood-feeding families of Coleoptera. The taxonomic relationship of the bacterial genera is shown in the rows, while the clustering of the coleopteran families, determined by their composition similarities, is shown in the columns. Higher intensities of the color reveal higher abundances of the isolates.
We collected larvae of five families of wood-feeding Coleoptera in tropical forests of Costa Rica with the aim of estimating the species composition of cultivable fungi and bacteria inhabiting their guts and to identify microorganisms with relevant lignocellulolytic activities. The main limitation of this study is that the cultivation-dependent approach, based on artificial media, covers only a small proportion of the total microbial diversity present in this particular niche. The positive trade-off of this approach was the identification of several isolates with lignocellulose-degrading capabilities, which can be further used for the respective enzyme characterization, for direct degradation assays on residues from agriculture and forestry, for the treatment of industrial effluents, and for bioprospecting novel metabolites with other biotechnological applications. Despite the inherent bias of the isolation method, our results suggest that gut microbiota of wood-feeding tropical beetles presents a relatively high diversity in terms of microbial richness, phylogenetic composition, and lignocellulolytic activities.
The order Hypocreales represented about 60% of the total number of fungal isolates. Within this group, the genus
In addition, our data indicate that guts of wood-feeding larvae were from environments having a high representation of Hypocreales, as also observed in a similar study performed in other locations of Costa Rica [
It is difficult to know whether these fungal isolates are truly endosymbionts of the intestinal tracts of the coleopteran larvae or are transitory inhabitants associated with host feeding habits. Hence, it is also possible that some of these microorganisms could be commensals, parasites, and facultative endosymbionts. They might even be using the insect as a dispersal mechanism [
The analysis of the taxonomic composition of the bacterial isolates showed the presence of seven major phylogenetic classes, codominated by
When comparing the fungal and bacterial species composition among the beetle families, the plots of the Bray-Curtis distances and canonical correspondence analyses produced biologically meaningful clusters to group the environments that share similar microbial compositions. The first fungal cluster relates the microbiota associated with the guts of Cerambycidae, Passalidae, and Scarabaeidae. This is consistent with the observation of a high diversity of isolates from Cerambycidae that shared members of the fungal phyla Basidiomycota and Ascomycota with Passalidae and members of Zygomycota and Ascomycota with Scarabaeidae. The cluster formed by Tenebrionidae-Elateridae shared, in a lower proportion, members of the Basidiomycota and Ascomycota. The bacterial microbiota associated with Passalidae and Scarabaeidae also formed a cluster, sharing members of five major bacterial clades; microbiota of Cerambycidae, Elateridae, and Tenebrionidae shared members only of
The clustering analyses revealed that Cerambycidae presented a high diversity of fungi but not of bacteria, while Passalidae and Elateridae exhibited a high diversity of bacteria and moderate diversity of fungi. Scarabaeidae and Tenebrionidae contained a similar composition of both. These results suggest that the nature of the beetle host has an important effect on the phylogenetic diversity of its associated microbiota and that many factors can influence its configuration. These factors may include the biology of the host, the physical and chemical characteristics of the gut compartments, the feeding habits of the insects, and the microbial diversity associated with the environment in which the insect is living [
Our results consistently showed that both the fungal and bacterial populations associated with the guts of beetle larvae are highly diverse in terms of the number of species obtained and in their phylogenetic composition. These microbial inhabitants could be forming complex consortia that would be acting synergistically to provide many of the nutritional needs of the beetle host. Some of these functions include the degradation and fermentation of lignocellulosic materials, as shown by the high percentage of fungal and bacterial genera that presented positive activities or by the production of proteins and other metabolites necessary for the development of the insect [
The authors declared that there is no conflict of interests regarding this paper.
The authors acknowledge the contribution of the following people during the development of this project: Luis Guillermo Acosta for the field sampling and early insect identification; Angel Solis, Carlos Hernández, and Elena Ulate for the identification of some adult specimens; Jorge Blanco for the fungal isolation; Angelica Acuña and Beatriz Rivera for the DNA extraction and enzymatic assays; Manuel Ferrer and Cesar Mateo for their advices on the lignocellulolytic activity determination; Ana Lorena Guevara and Giselle Tamayo for the overall support; the editor and reviewers of this journal for critical comments on the paper. This research was funded by the support of the National Council of Science and Technology (CONICIT, FV-027-2007), the CSIC and CRUSA Foundation (2007 CR0034), and Florida Ice & Farm Co. They thank ACLAC, ACOPAC, ACOSA, ACTo, ACLAP, ACCVC, ACAT, ACAHN, ACG, and ACLAC National Conservation Areas and CONAGEBIO for granting the sample collecting permits (R-CM-INBio-40-2008-OT, R-CM-INBio-48-2008-OT).