Fig 1.
Leaf-cutter ants concentrate decaying plant matter in refuse dumps.
(A) Leaf-cutter ants are dominant herbivores in Central and South American rain forests. An Atta worker carries a leaf fragment back to her nest in Costa Rica. (B) The leaf structure is still visible in the dump material tended by this Atta worker in the Currie lab at University of Wisconsin-Madison. (C) A vertically cross-sectioned A. colombica refuse dump in Costa Rica. Photo credits: Don Parsons (A, B), Gina Lewin (C).
Fig 2.
Electron microscopy of leaf-cutter ant refuse dumps.
(A and B) Scanning electron microscopy shows the ultrastructure of refuse dump leaf material and different bacterial morphologies. (C) Transmission electron microscopy shows leaf cells and surrounding bacteria. Red boxes indicate degraded plant cell wall and abnormal, clumped internal cell structure. Photo credits: Rolando Moreira Soto.
Fig 3.
Comparison of degradation ability across colonies and layers.
(A and B) Qualitative Assay Data. Test tubes containing carbon-free minimal media and a strip of cellulosic filter paper were used to enrich for cellulolytic communities. Failure plots, indicating when the filter paper broke apart in each culture, were fit with Kaplan Meier curves and analyzed using the Wilcoxon method to determine significant differences among colonies (indicated by letters A-C) and layers (no significant differences). (C and D) Quantitative Assay Data. Pre-weighed, submerged cellulosic filter paper allowed quantification of cellulose degradation after 10 days. Samples are grouped by colony or dump layer. Error bars represent one standard error from the mean. Significant differences were determined using Tukey’s HSD test and are indicated above the data. Photo credits: Gina Lewin (A, C).
Table 1.
Degradation, sequencing depth, and alpha diversity metrics for sequenced samples at a 97% OTU definition.
Fig 4.
Morisita-Horn Beta Diversity Clustering of Samples.
The corresponding percentage of cellulose degradation, colony, layer, and taxonomic classification of OTUs are shown for each sample.
Fig 5.
Correlation between number of reads and cellulose degradation by the community for OTU1 (Acidovorax) and OTU9 (Ferruginibacter) across sequenced samples.
Table 2.
Taxonomy of top 30 OTUs across all samples and reports of cellulose degradationa.
Fig 6.
PCoA clustering of Morisita-Horn Diversity Index.
Sample shape indicates colony. Sample color indicates degradation (A) or layer (B). Panel C shows the correlation analysis. The vectors indicate the correlation of each OTU and the percentage of cellulose degradation with the principal coordinates shown.