An estimated 1030 virioplankton (viruses) are present in the world’s oceans, the majority of which are phages. These virioplankton are dynamic members of the community which have been shown to have significant influence on prokaryotic abundance, community structure, genetic exchange and global biochemical cycles, as well as being a vital genetic resource. They are also incredibly abundant with concentrations of 107 viruses ml-1 and an average ratio of 10 viruses to every bacteria. This abundance is typically highest in coastal regions and in the euphotic zone, which is thought to reflect the decrease in host abundance towards the deeper, more oceanic waters. Here, the dynamics of virioplankton are less well understood and so this study presents the first depth-resolved (300m) multi-year time series of oceanic virioplankton abundance and explains the dynamics in the context of water column stability, and bacterioplankton abundance and community composition.
Data were collected over a 10 year period as part of the Bermuda Atlantic Time-Series (BATS) program. The site was sampled for virio- and bacterioplankton monthly, and also biweekly between February and April, at a range of depths from 0-300m between January 2000 and December 2009. These samples were then fixed with formalin and gluteraldehyde, respectively, and stored until processing. Picophytoplankton samples were also taken, however only for the depths up to 250m. The mixed layer depth (MLD) was also determined. This is the depth of the largely homogenous layer of water found in seasonal systems caused by turbulent mixing.
The virioplankton samples were filtered then stained with SYBR Green I and enumerated using epifluorescence microscopy. A similar procedure was adopted for the bacterioplankton samples, although a 4’, 6-diamidino-2-phenylindole dihydrochloride stain was used. Flow cytometry was used to enumerate Prochlorococcus and Synecococcus samples; and to quantify the abundance of the SAR11 and Rhodobacteraceae clades.
The study reveals recurring annual patterns in virioplankton abundance which coincide with patterns in water column stability, represented by the MLD. The MLD reached depths of 150-300m in late winter/early spring due to mixing, followed by summer stratification events resulting in a low MLD (<50m). During such stratification events virioplankton abundance accumulated below the MLD and surface abundances were at minimal values, possibly due to high levels of ultraviolet radiation exposure to the hosts or to the viruses themselves. However from July to December, during convective mixing, virioplankton abundances initially increased, potentially due to redistribution and enhanced mixing with hosts, but subsequently fell within the upper 300m. This implies that convective mixing cannot be the only driver of virioplankton dynamics.
Trends in bacterioplankton abundances were also found to be apparent and some held a correlation with those of the virioplankton community. In terms of total abundances, both virio- and bacterioplankton reached maxima below the MLD during stratification events; however, they did not coincide temporally or with depth. Although, when only data from the depth region of the virioplankton maxima (60-100m) were considered; a slow accumulation of bacterioplankton was found between January and July, followed by a decrease between July and December. This decrease coincided with the continual increase in virioplankton abundance from July to December and seemed to be consistent with the Lotka-Volterra model of predator-prey relationships. Consequently, it seemed to be a ‘kill the winner’ scenario.
However, the lack of any real correlation was unmistakeable and unsurprising given that many viruses are host-specific. Therefore it seemed more likely that virioplankton dynamics would be more strongly correlated to specific lineages within the bacterioplankton. The SAR11 clade, although one of the most ubiquitous eukaryotic groups in open ocean systems, showed little correlation with the virioplankton dynamics. A weak, negative correlation was found with the mean depth horizon (60-100m) but the virioplankton-to-SAR11 ratio was consistently within a narrow range and varied little according to season and the MLD, suggesting no real relationship between the SAR11 clade and virioplankton abundances. This may be due to SAR11 being a defence specialist or its being a suboptimal host due to its k-selected life history strategy.
Synecococcus was also considered to be an unlikely driver of virioplankton abundance dynamics as they were completely decoupled from those of the virioplankton both temporally and with depth. Although the abundances were inversely correlated, suggesting a ‘kill the winner’ scenario, calculations showed that the predator-prey period ought to be shorter than observed. Furthermore, the virioplankton-to-Synecococcus ratio was very high and variable (1000- >2000) at the depth horizon between July and December; values which vastly exceed the burst size of lab cultured cyanophages which infect Synecococcus.
Conversely, Prochlorococcus was found to be very tightly coupled with virioplankton dynamics. Regular annual patterns demonstrated by the Prochlorococcus were very similar to those of the virioplankton including the annual maxima developing in the depth horizon of 60-100m and the variability of surface layer maxima during stratification events. Thus the two were highly correlated, with 44% of variance in virioplankton dynamics explained by those of Prochlorococcus. Also, in years when Prochlorococcus abundances were low, so too were those of the virioplankton. Although the link between these two groups is purely correlative, it seems highly likely that Prochlorococcus could be the driving force behind the oceanic abundance variations in the virioplankton at BATS, and that therefore the majority of the virioplankton community could be cyanophages.
Results for the Rhodobacteraceae clade also showed correlative evidence for potentially being the driver of virioplankton dynamics as it too had its annual maxima occurring within the 60-100m depth horizon and explained 23% of virioplankton variance in the data. However, members of the Rhodobacteraceae are typically found at low abundances in oligotrophic waters such as those at BATS and even with their r-selected strategy would not produce burst sizes large enough to carry the virioplankton dynamics found in the study. Therefore it is unlikely that the Rhodobacteraceae have much real influence on the virioplankton abundance dynamics.
As a consequence of these results, it is probable that virioplankton dynamics at BATS are influenced mostly by host availability and that the host appears to be Prochlorococcus. This availability is however, influenced by the seasonal mixing cycles found within the BATS area and so I think that further research ought to be done here to confirm the ecological processes at work. Having said that, this paper alone does give a good insight into the ecological role that viruses play within oceanic systems; which was a previous unknown for the area.
A review of:
Rachel J. Parsons, Mya Breitbart, Michael W. Lomas and Craig A. Carlson (2011) Ocean time-series reveals recurring patterns of virioplankton dynamics in the northwestern Sargasso Sea; The ISME Journal [online] doi:10.1038/ismej.2011.101
No comments:
Post a Comment