Shifts in Symbiotic Endophyte Communities of a Foundational Salt Marsh Grass following Oil Exposure from the Deepwater Horizon Oil Spill

Abstract

Symbiotic associations can be disrupted by disturbance or by changing environmental conditions. Endophytes are fungal and bacterial symbionts of plants that can affect performance. As in more widely known symbioses, acute or chronic stressor exposure might trigger disassociation of endophytes from host plants. We tested this hypothesis by examining the effects of oil exposure following the Deepwater Horizon (DWH) oil spill on endophyte diversity and abundance in Spartina alterniflora - the foundational plant in northern Gulf coast salt marshes affected by the spill. We compared bacterial and fungal endophytes isolated from plants in reference areas to isolates from plants collected in areas with residual oil that has persisted for more than three years after the DWH spill. DNA sequence-based estimates showed that oil exposure shifted endophyte diversity and community structure. Plants from oiled areas exhibited near total loss of leaf fungal endophytes. Root fungal endophytes exhibited a more modest decline and little change was observed in endophytic bacterial diversity or abundance, though a shift towards hydrocarbon metabolizers was found in plants from oiled sites. These results show that plant-endophyte symbioses can be disrupted by stressor exposure, and indicate that symbiont community disassembly in marsh plants is an enduring outcome of the DWH spill.

Fig 1. Map of study area.

(A) The location of S. alterniflora collections for oiled (yellow circles) and unoiled reference (red circles) areas at Bay Jimmy; and (B) the location of S. alterniflora collections for oiled and unoiled reference areas at Fourchon. Specific GPS coordinates for each area at both sites are provided in Table A in S1 File. Images were obtained from the Louisiana Oil Spill Coordinator's Office (LOSCO), 20000120, Louisiana Land and Water Interface, Geographic NAD83, LOSCO (2000) (landwater_interface_LOSCO_1992). Metadata for these maps are available at http://lagic.lsu.edu/data/losco/landwater_interface_losco_1992.html.

Fig 2. Rarefaction curves for the number of operational taxonomic units (OTUs) by group.

The dotted line corresponds to samples from oiled areas, the solid line represents samples from unoiled reference areas, and the dashed line corresponds to all samples combined.

Fig 3. Linear contrasts for foliar and root endophyte diversity and abundance according to treatment.

(A) Fungal endophyte diversity in leaves and roots; (B) bacterial endophyte diversity in leaves and roots; (C) fungal endophyte abundance in leaves and roots; and (D) bacterial endophyte abundance in leaves and roots. Black bars correspond to oiled areas and grey bars correspond to unoiled reference areas. * = significant differences. Error bars are standard errors.

Fig 4. Ordinations of endophyte community data.

(A) Nonmetric multidimensional scaling using centroids for foliar and root communities from oiled and unoiled reference areas in Bay Jimmy and Fourchon. Arrows indicate the direction of community change between oiled and unoiled reference conditions. (B) Canonical Analysis of Principal Coordinates on unconstrained data illustrating community differences according to study location, oiling and tissue type with superimposed vectors of OTUs driving community differences.

Fig 5. The results of in vitro growth assays showed that oil reduced the growth of some endophyte strains, but increased growth for others.

(A) Mean (±1 standard error) percent change in colony growth shows that bacteria tended to be more oil-philic while fungi were relatively oil-phobic. The dotted zero line represents no difference in growth on oiled versus unoiled media. (B) The top panel shows growth differences for an oil-phobic fungus isolated from a leaf, while the bottom panel shows an oil-phobic fungus isolated from a root.