Robustness of the bacterial community in the cabbage white butterfly larval midgut

Abstract

Microbial communities typically vary in composition and structure over space and time. Little is known about the inherent characteristics of communities that govern various drivers of these changes, such as random variation, changes in response to perturbation, or susceptibility to invasion. In this study, we use 16S ribosomal RNA gene sequences to describe variation among bacterial communities in the midguts of cabbage white butterfly (Pieris rapae) larvae and examine the influence of community structure on susceptibility to invasion. We compared communities in larvae experiencing the same conditions at different times (temporal variation) or fed different diets (perturbation). The most highly represented phylum was Proteobacteria, which was present in all midgut communities. The observed species richness ranged from six to 15, and the most abundant members affiliated with the genera Methylobacteria, Asaia, Acinetobacter, Enterobacter, and Pantoea. Individual larvae subjected to the same conditions at the same time harbored communities that were highly similar in structure and membership, whereas the communities observed within larval populations changed with diet and over time. In addition, structural changes due to perturbation coincided with enhanced susceptibility to invasion by Enterobacter sp. NAB3R and Pantoea stewartii CWB600, suggesting that resistance to invasion is in part governed by community structure. These findings along with the observed conservation of membership at the phylum level, variation in structure and membership at lower taxonomic levels, and its relative simplicity make the cabbage white butterfly larval community an attractive model for studying community dynamics and robustness.

Figure 1

Average distribution of clones affiliated with the α-, β-, and γ-Proteobacteria subphyla and other phyla identified in cabbage white butterfly midgut bacterial communities. Graph excludes Actinobacteria and Chloroflexi, which were each detected once in one sample (0.13%)

Fig. 2

Comparisons of community structure as determined by Yue and Clayton (θ) indices at OTU_0.03. a Batch-to-batch and larva-to-larva structural variation. Unweighted pair-group method using arithmetic mean (UPGMA) clustering of communities from larvae reared on sterile artificial diet only. Library names are followed by date of sampling. b Effects of diet on community structure. UPGMA clustering of communities in larvae reared on sterile artificial diet (No Trt), Brussels sprouts, sinigrin, penicillin and streptomycin (PenStrep), or penicillin and streptomycin and then transferred to sterile artificial diet (PenStrep Recovered). Reference bar length corresponds to a distance of 0.10 (distance = 1−θ). No treatment control for antibiotics experiment (a); sequences from libraries constructed from individual larvae reared on sterile artificial diet were pooled and treated as one library (b); no treatment control for diet experiment (c)

Figure 3

Distribution of predominant γ-Proteobacteria in libraries constructed from larvae reared on unamended sterile artificial diet (BL-4), sterile artificial diet amended with penicillin and streptomycin (BL-5), or sterile artificial diet amended with penicillin and streptomycin and then transferred to unamended artificial diet (BL-6)

Figure 4

Effect of antibiotics on abundance of culturable bacteria. Midguts from cabbage white butterfly larvae fed untreated sterile artificial diet (n = 24) or penicillin and streptomycin (n = 29) were plated on one tenth strength tryptic soy agar. Error bars are standard error mean. Different letters indicate significant difference (Kruskal–Wallis and Mann–Whitney; P < 0.05)

Figure 5

Effects of antibiotics and diet on establishment of invaders in cabbage white butterfly midgut communities. a Larvae fed unamended sterile artificial diet (n = 36) or penicillin and streptomycin (100 units/ml and 100 μg/ml, respectively; n = 33) from hatching were fed Pantoea sp. CWB600. b Larvae fed unamended sterile artificial diet (n = 27) or sterile artificial diet amended with sinigrin (3.0 mg/ml; n = 18) or Brussels sprouts (n = 23) from hatching were fed Pantoea sp. CWB304. Error bars are standard error mean. Different letters indicate that values differ significantly (Kruskal–Wallis and Mann–Whitney; P values <0.05