Multivariate and phylogenetic analyses assessing the response of bacterial mat communities from an ancient oligotrophic aquatic ecosystem to different scenarios of long-term environmental disturbance

Abstract

Understanding the response of bacterial communities to environmental change is extremely important in predicting the effect of biogeochemical modifications in ecosystem functioning. The Cuatro Cienegas Basin is an ancient oasis in the Mexican Chihuahuan desert that hosts a wide diversity of microbial mats and stromatolites that have survived in extremely oligotrophic pools with nearly constant conditions. However, thus far, the response of these unique microbial communities to long-term environmental disturbances remains unexplored. We therefore studied the compositional stability of these bacterial mat communities by using a replicated (3x) mesocosm experiment: a) Control; b) Fluct: fluctuating temperature; c) 40C: increase to 40 ºC; d) UVplus: artificial increase in UV radiation; and f) UVmin: UV radiation protection. In order to observe the changes in biodiversity, we obtained 16S rRNA gene clone libraries from microbial mats at the end of the experiment (eight months) and analyzed them using multivariate and phylogenetic tools. Sequences were assigned to 13 major lineages, among which Cyanobacteria (38.8%) and Alphaproteobacteria (25.5%) were the most abundant. The less extreme treatments (Control and UVmin) had a more similar composition and distribution of the phylogenetic groups with the natural pools than the most extreme treatments (Fluct, 40C, and UVplus), which showed drastic changes in the community composition and structure, indicating a different community response to each environmental disturbance. An increase in bacterial diversity was found in the UVmin treatment, suggesting that protected environments promote the establishment of complex bacterial communities, while stressful environments reduce diversity and increase the dominance of a few Cyanobacterial OTUs (mainly Leptolyngbya sp) through environmental filtering. Mesocosm experiments using complex bacterial communities, along with multivariate and phylogenetic analyses of molecular data, can assist in addressing questions about bacterial responses to long-term environmental stress.

Fig 1. Mesocosm experiment with “synthetic” microbial mats from CCB.

A: Poza Azul; B: “microbial mat catchers” (trays with frosted glass slides) in the natural pools; C: mesocosm experiment.

Fig 2. Relative abundance of bacterial taxonomic groups from the clone libraries data of the microbial mats in the treatments and the pools at the end of the experiment.

The number of clones obtained from each environment is indicated below in its respective bar. The photo above each bar represents the “microbial mat catchers” in each environment.

Fig 3. Heatmap showing the abundance of the 39 bacterial orders in the pools and treatments at the end of the experiment based on the 16S rRNA gene libraries.

The Bray-Curtis dissimilarity matrix and Ward’s hierarchical clustering algorithm were applied in this analysis. Higher abundance is indicated by increased blue color intensity. The value shown in the color key scale represents the number of sequences detected for each environment. Rows indicate bacterial orders and columns indicate the different environments.

Fig 4. Non-metric multidimensional scaling (NMDS) ordination plot of Bray–Curtis community dissimilarities based on OTUs from the 16S rRNA gene sequences at the end of the experiment (2D stress value = 0.19).

Color symbols represent the mesocosms from each treatment and pools, while grey crosses represent OTUs.

Fig 5. Box-and-whisker plots comparing the phylogenetic mean pairwise distance (MPD) and the mean nearest taxon distance (MNTD) values for the bacterial OTUs between pools and treatments.

Asterisks indicate communities that are significantly structured at p< 0.05 level.