A Critical Review of Procedures and Approaches Used for Assessing Pollution-Induced Community Tolerance (PICT) in Biotic Communities

Pollution-induced community tolerance (PICT) is used for the detection of minor effects of toxicants in biotic communities. Organisms survive in toxic environments only if they are tolerant to the chemicals present in their habitat. In the selection phase, toxicants hinder the success of sensitive individuals and species and replace them by more tolerant ones. The resulting increase in community tolerance is quantified in the detection phase by short-term toxicity tests. In this way PICT can establish causal linkages between contaminants and effects. An increase in community tolerance compared to the baseline tolerance at reference sites suggests that the community has been adversely affected by toxicants. PICT has been used in aquatic and terrestrial environments with communities of periphyton, phytoplankton, bacteria, nematodes and insects. A variety of methods have been used for quantification of community tolerance including photosynthesis, sulfolipid synthesis, respiration, thymidine and leucine incorporation, survival, and substrate utilisation patterns. PICT has been observed for copper, zinc, nickel, mercury, cadmium, arsenate, monomethylarsonic acid, diuron, tributyl tin, 4,5,6-trichloroguaiacol, irgarol 1051, seanine 211, atrazine, and trinitrotoluene. It is necessary to validate PICT, at least by showing that it is related to the preexposure concentration of the toxicants and that it is coupled to a toxicant-induced succession (TIS) in the community. Care must also be taken to ascertain that PICT interpretation is not confounded by co-tolerance or bioavailability differences. Co-tolerance patterns, which are indicative of the specificity of PICT, have only been investigated for arsenate, diuron and a few metals. For the further improvement of PICT methodology special attention should be given to co-tolerance patterns and development of new integrating short-term tests for quantification of tolerance. It is also important to broaden the scope of organisms and toxicants used. Properly validated, PICT is a powerful tool for detection of community effects and its use in monitoring and site-specific risk assessment should be encouraged.     

The Role of Soil Microbial Tests in Ecological Risk Assessment: Differentiating between Exposure and Effects

The use of soil microorganisms in ecological risk assessment is hampered by an unclear dose-response relationship for most contaminants. Establishing dose-response curves for soil microbial communities requires that one have a clear estimate of exposure at the site of toxic action and a response free of confounding environmental factors. It is not clear what methods can estimate toxicant dose at the site of toxic action or determine microbial response to a toxicant. Pollution-induced community tolerance (PICT) is one possible estimate of microbial toxicant exposure. The PICT hypothesis is that the tolerance of a microbial community is proportional to the in situ dose. This method automatically corrects for differences due to differences in soil physical-chemical variables between samples. Various components of the soil nitrogen cycle can act as microbial bioindicators of toxicant impacts. Estimating denitrifica-tion activity presents a number of advantages over other components of the nitrogen cycle. Denitrifying bacteria come from a diversity of habitats, can be autotrophic or heterotrophic, and denitrification is a well-defined enzymatic system, which allows the use of molecular tools. Determining denitrification may be a good estimate of effects of toxicants on microbial communities. However, given the state of our ignorance regarding soil microbial community structure and function, redundant estimates of exposure and effect are necessary to adequately characterize the response of microbial communities to toxicants.     

Determination of Field Effects of Contaminants—Significance of Pollution-Induced Community Tolerance

The concept of pollution-induced community tolerance (PICT) consists of the phenomenon that communities in an ecosystem exhibit increased tolerance as a result of exposure to contaminants. Although a range of ‘classic’ ecological principles explains the processes that increase tolerance of a community, the value of PICT for ecological risk assessment was recognized only recently (Blanck et al. 1988). The following issues are recognized: First, regarding the question on the role of suspect compounds causing ecological effects, the PICT approach covers the issue of causality better than ‘classical’ ecological community response parameters like species densities or species diversity indices. This relates to the fact that the level of PICT is assumed to be relatively constant (compared to density and diversity), whereas the suspect compound causing the observed effect can be deduced with relative clear inference from artificial exposure experiments. Second, PICT directly addresses a level of biological organization (the community), the level of concern for many ecological risk assessment methods. Other methods for risk assessment, like toxicity testing or bioassays, focus on individual or population-level effects, and need extrapolation of the results to the field. Such extrapolation step may pose problems regarding validity of the outcome of risk assessment. The occurrence of PICT is, however, not (yet) a community endpoint that is sufficiently underpinned to trigger risk mitigation activities. This paper especially focuses on the possibility to improve risk assessment approaches by incorporation of PICT assessments, especially focusing on the issue of causality and on the ecological meaning of PICT. Despite the advantages over ‘classical’ parameters, literature analysis suggests that the PICT approach may be strengthened by determining to which degree the PICT approach relates to ecological changes, like shifts in community structure, functioning, and stability. The aim of this paper is to summarize some literature, putting the emphasis on terrestrial studies, to get insights whether PICT is a sensitive and powerful tool to quantify ecological effects in field conditions, to link them to toxicant stress, and thus to determine whether PICT may be taken into consideration in risk assessment.     

Effects of copper on periphyton communities assessed in situ using chemical-diffusing substrates

Chemical-diffusing substrates were designed to allow delivery of toxicants to mature periphyton communities under natural conditions without contaminating the surrounding environment. Artificial stream validation studies were conducted in which the effects of substrate-released copper (Cu) on periphyton communities were compared to those generated in a more conventional manner (via water column additions). Effects of copper on the following community parameters were assessed: total community biomass (measured as ash-free dry mass), relative chlorophyll a (chl a and adenosine triphosphate contents, and relative biomass of heterotrophic bacteria. Exposure of more laboratory periphyton communities to substrate-released Cu generated dose-response relationships and recovery models that were indistinguishable from those generated by the conventional route of exposure. The results of this study demonstrate the utility of chemical-diffusing substrates in field validations of laboratory toxicity tests and in investigations of the effects of stress history on periphyton tolerance to toxicants. This revised version was published online in September 2006 with corrections to the Cover Date.     

The periphyton as a multimetric bioindicator for assessing the impact of land use on rivers: an overview of the Ardières-Morcille experimental watershed (France)

Developing new biological indicators for monitoring toxic substances is a major environmental challenge. Intensive agricultural areas are generally pesticide-dependent and generate water pollution due to transfer of pesticide residues through spray-drift, run-off and leaching. The ecological effects of these pollutants in aquatic ecosystems are broad-ranging owing to the variety of substances present (herbicides, fungicides, insecticides, etc.). Biofilms (or periphyton) are considered to be early warning systems for contamination detection and their ability to reveal effects of pollutants led researchers to propose a variety of methods to detect and assess the impact of pesticides. The present article sought to provide new insights into the ecological significance of biofilm microbial communities and to discuss their bioindication potential for water quality and land use by reporting on 4 years of research performed on the French Ardières-Morcille experimental watershed (AMEW). Various biological indicators have been applied during several surveys on AMEW, allowing the characterisation of (i) the structure and diversity of biofilm communities [community level finger printing (CLFP) such as PCR–DGGE and pigment classes], (ii) functions associated with biofilm [community level physiological profiles (CLPP) such as extracellular enzymes, pesticides biodegradation or carbon sources biodegradation] and (iii) biofilm tolerance assessment (pollution-induced community tolerance, PICT) of the main contaminant in the AMEW (copper and diuron). Approaches based on CLFPs and PICT were consistent with each other and indicated the upstream–downstream impact due to the increasing land use by vineyards and the adaptation of algal and bacterial communities to the pollution gradient. CLPPs gave a contrasted bioindication because some parameters (most of the tested extracellular enzymes activities) did not detect a pollution gradient. Such CLPPs, CLFPs and PICT methods applied to biofilm could constitute the basis for a relevant in situ assessment both for chemical effects and aquatic ecosystem resilience.     

Cold-water coral reef frameworks,megafaunal communities and evidence for coral carbonate mounds on the Hatton Bank,north east Atlantic

Offshore banks and seamounts sustain diverse megafaunal communities, including framework reefs formed by cold-water corals. Few studies have quantified environmental effects on the alpha or beta diversity of these communities. We adopted an interdisciplinary approach that used historical geophysical data to identify topographic highs on Hatton Bank, which were surveyed visually. The resulting photographic data were used to examine relationships between megafaunal communities and macrohabitat, the latter defined into six categories (mud, sand, cobbles, coral rubble, coral framework, rock). The survey stations revealed considerable small-scale variability in macrohabitat from exposed Late Palaeocene lava flows to quiescent muddy habitats and coral-built carbonate mounds. The first reported evidence for coral carbonate mound development in UK waters is presented, which was most pronounced near present-day or former sites of topographic change, suggesting that local current acceleration favoured coral framework growth and mound initiation. Alpha diversity varied significantly across macrohabitats, but not between rock and coral rubble, or between smaller grain sized categories of cobbles, sand and mud. Community composition differed between most macrohabitats, and variation in beta diversity across Hatton Bank was largely explained by fine-scale substratum. Certain megafauna were clearly associated with particular macrohabitats, with stylasterid corals notably associated with cobble and rock habitats and coral habitats characterized by a diverse community of suspension-feeders. The visual surveys also produced novel images of deep-water megafauna including a new photographic record of the gorgonian coral Paragorgia arborea, a species not previously reported from Rockall Plateau. Further interdisciplinary studies are needed to interpret beta diversity across these and other environmental gradients on Hatton Bank. It is clear that efforts are also needed to improve our understanding of the genetic connectivity and biogeography of vulnerable deep-water ecosystems and to develop predictive models of their occurrence that can help inform future conservation measures. An erratum to this article can be found at     

Risk Assessment,Microbial Communities,and Pollution-Induced Community Tolerance

Until recently, parameters from microorganisms were generally not included in risk assessment at a comparable level to animals and plants. However, the major part of global biomass, biodiversity, and ecosystem processes is present in the microbial world and microbiological techniques applicable to risk assessment are becoming available. Two microbial indicators are described based on the usage of multiwell plates with different substrates and a redox indicator for monitoring mineralisation. With both techniques autochthonous microbial communities are analysed. Producing functional fingerprints of the microbial community gives insights into the composition of different functions. This is equivalent to observations of ecological abundance and species composition. When lack of reference sites or reference data renders risk assessment difficult, measurement of the pollution-induced community tolerance (PICT) can provide useful information.     

Fungal-Fungal Associations Affect the Assembly of Endophyte Communities in Maize (Zea mays)

Many factors can affect the assembly of communities, ranging from species pools to habitat effects to interspecific interactions. In microbial communities, the predominant focus has been on the well-touted ability of microbes to disperse and the environment acting as a selective filter to determine which species are present. In this study, we investigated the role of biotic interactions (e.g., competition, facilitation) in fungal endophyte community assembly by examining endophyte species co-occurrences within communities using null models. We used recombinant inbred lines (genotypes) of maize (Zea mays) to examine community assembly at multiple habitat levels, at the individual plant and host genotype levels. Both culture-dependent and culture-independent approaches were used to assess endophyte communities. Communities were analyzed using the complete fungal operational taxonomic unit (OTU) dataset or only the dominant (most abundant) OTUs in order to ascertain whether species co-occurrences were different for dominant members compared to when all members were included. In the culture-dependent approach, we found that for both datasets, OTUs co-occurred on maize genotypes more frequently than expected under the null model of random species co-occurrences. In the culture-independent approach, we found that OTUs negatively co-occurred at the individual plant level but were not significantly different from random at the genotype level for either the dominant or complete datasets. Our results showed that interspecific interactions can affect endophyte community assembly, but the effects can be complex and depend on host habitat level. To our knowledge, this is the first study to examine endophyte community assembly in the same host species at multiple habitat levels. Understanding the processes and mechanisms that shape microbial communities will provide important insights into microbial community structure and the maintenance of microbial biodiversity.     

Diversity of Dominant Bacterial Taxa in Activated Sludge Promotes Functional Resistance following Toxic Shock Loading

Examining the relationship between biodiversity and functional stability (resistance and resilience) of activated sludge bacterial communities following disturbance is an important first step towards developing strategies for the design of robust biological wastewater treatment systems. This study investigates the relationship between functional resistance and biodiversity of dominant bacterial taxa by subjecting activated sludge samples, with different levels of biodiversity, to toxic shock loading with cupric sulfate (Cu[II]), 3,5-dichlorophenol (3,5-DCP), or 4-nitrophenol (4-NP). Respirometric batch experiments were performed to determine the functional resistance of activated sludge bacterial community to the three toxicants. Functional resistance was estimated as the 30 min IC₅₀ or the concentration of toxicant that results in a 50% reduction in oxygen utilization rate compared to a referential state represented by a control receiving no toxicant. Biodiversity of dominant bacterial taxa was assessed using polymerase chain reaction-terminal restriction fragment length polymorphism (PCR-T-RFLP) targeting the 16S ribosomal RNA (16S rRNA) gene. Statistical analysis of 30 min IC₅₀ values and PCR-T-RFLP data showed a significant positive correlation (P < 0.05) between functional resistance and microbial diversity for each of the three toxicants tested. To our knowledge, this is the first study showing a positive correlation between biodiversity of dominant bacterial taxa in activated sludge and functional resistance. In this system, activated sludge bacterial communities with higher biodiversity are functionally more resistant to disturbance caused by toxic shock loading.     

Contrasting primary successional trajectories of fungi and bacteria in retreating glacier soils

Early community assembly of soil microbial communities is essential for pedogenesis and development of organic legacies. We examined fungal and bacterial successions along a well‐established temperate glacier forefront chronosequence representing ~70 years of deglaciation to determine community assembly. As microbial communities may be heavily structured by establishing vegetation, we included nonvegetated soils as well as soils from underneath four plant species with differing mycorrhizal ecologies (Abies lasiocarpa, ectomycorrhizal; Luetkea pectinata, arbuscular mycorrhizal; Phyllodoce empetriformis, ericoid mycorrhizal; Saxifraga ferruginea, nonmycorrhizal). Our main objectives were to contrast fungal and bacterial successional dynamics and community assembly as well as to decouple the effects of plant establishment and time since deglaciation on microbial trajectories using high‐throughput sequencing. Our data indicate that distance from glacier terminus has large effects on biomass accumulation, community membership, and distribution for both fungi and bacteria. Surprisingly, presence of plants rather than their identity was more important in structuring bacterial communities along the chronosequence and played only a very minor role in structuring the fungal communities. Further, our analyses suggest that bacterial communities may converge during assembly supporting determinism, whereas fungal communities show no such patterns. Although fungal communities provided little evidence of convergence in community structure, many taxa were nonrandomly distributed across the glacier foreland; similar taxon‐level responses were observed in bacterial communities. Overall, our data highlight differing drivers for fungal and bacterial trajectories during early primary succession in recently deglaciated soils.     

Microbial communities acclimate to recurring changes in soil redox potential status

Rapidly fluctuating environmental conditions can significantly stress organisms, particularly when fluctuations cross thresholds of normal physiological tolerance. Redox potential fluctuations are common in humid tropical soils, and microbial community acclimation or avoidance strategies for survival will in turn shape microbial community diversity and biogeochemistry. To assess the extent to which indigenous bacterial and archaeal communities are adapted to changing in redox potential, soils were incubated under static anoxic, static oxic or fluctuating redox potential conditions, and the standing (DNA‐based) and active (RNA‐based) communities and biogeochemistry were determined. Fluctuating redox potential conditions permitted simultaneous CO₂ respiration, methanogenesis, N₂O production and iron reduction. Exposure to static anaerobic conditions significantly changed community composition, while 4‐day redox potential fluctuations did not. Using RNA : DNA ratios as a measure of activity, 285 taxa were more active under fluctuating than static conditions, compared with three taxa that were more active under static compared with fluctuating conditions. These data suggest an indigenous microbial community adapted to fluctuating redox potential.     

A phylogenetic approach to detect selection on the target site of the antifouling compound irgarol in tolerant periphyton communities

Using DNA sequence data for phylogenetic assessment of toxicant targets is a new and promising approach to study toxicant-induced selection in communities. Irgarol 1051 is a photosystem (PS) II inhibitor used in antifouling paint. It inhibits photosynthesis through binding to the D1 protein in PS II, which is encoded by the psbA gene found in genomes of chloroplasts, cyanobacteria and cyanophages. psbA mutations that alter the target protein can confer tolerance to PS II inhibitors. We have previously shown that irgarol induces community tolerance in natural marine periphyton communities and suggested a novel tolerance mechanism, involving the amino acid sequence of a turnover-regulating domain of D1, as contributive to this tolerance. Here we use a large number of psbA sequences of known identity to assess the taxonomic affinities of psbA sequences from these differentially tolerant communities, by performing phylogenetic analysis. We show that periphyton communities have high psbA diversity and that this diversity is adversely affected by irgarol. Moreover, we suggest that within tolerant periphyton the novel tolerance mechanism is present among diatoms only, whereas some groups of irgarol-tolerant cyanobacteria seem to have other tolerance mechanisms. However, it proved difficult to identify periphyton psbA haplotypes to the species or genus level, which indicates that the genomic pool of the attached, periphytic life forms is poorly studied and inadequately represented in international sequence databases.     

Effects of arsenate and copper on the algal communities in polluted lakes in the northern parts of Sweden assayed by PICT (Pollution-Induced Community Tolerance)

The tolerance for arsenate and copper in the carbon dioxide fixation activity of phytoplankton communities coming from lakes around the smelter at Rönnskär at the Swedish east-coast were measured during three years (1989–1991). The smelter have for several decades discharged arsenic and heavy metals into the air, and their concentrations in the lakes were clearly correlated with the distance from the smelter. The tolerance of communities from the three most polluted lakes were higher than communities from reference lakes with background concentrations of arsenic and copper. In accordance with the PICT concept it is indicated that those communities have been affected by these substances. These communities also had lower diversity than the others, but no clear correlation could be done with phytoplankton species number, or between phytoplankton biomass and pollution levels.     

Carbon cycling by seafloor communities on the eastern Beaufort Sea shelf

Tight pelagic-benthic coupling on Arctic shelves suggests that resident benthic communities may be particularly important in the cycling of carbon and regeneration of nutrients. We sampled 16 stations in the eastern Beaufort Sea during Autumn 2003 and Summer 2004 to determine spatial patterns in sediment community carbon demand, and the manner in which that demand was partitioned among epifauna, macroinfauna, and meio-/microfauna. Sediment carbon demand in this relatively oligotrophic area was similar to that measured in more productive Arctic shelf sites, and was largely related to the distribution of phytodetritus in surface sediments. Epibenthic megafaunal communities were dominated by echinoderms and exhibited peak abundance (up to 240 ind. m^− 2) and biomass at stations in the 60-90 m depth range. Partitioning of the carbon demand revealed the local importance of megafauna, accounting for up to 41% of the community demand. Macrofauna accounted for on average between 25 and 69% of the carbon demand, while meio-/microfauna were responsible for 31-75% of the demand. Total community carbon demand by the benthos is estimated to account for approximately 60% of the annual new production in the region, suggesting the great ecosystem importance of benthic communities on the Beaufort shelf, and potentially across the Arctic. Our study region is strongly influenced by the Mackenzie River, and ongoing climate change is likely to result in altered productivity regimes, changes in quality and quantity of available food, and higher levels of sediment deposition. Impacts of these events on benthic community structure and function will likely have repercussions throughout the ecosystem.     

Recovery of a stream macroinvertebrate community from mine drainage disturbance

Recovery of aquatic macroinvertebrates from the effects of mine drainage was documented using a weight-of-evidence approach which included measures of physical, chemical, and biological data. Taxa richness; number of taxa in the orders Ephemeroptera, Plecoptera, and Trichoptera; and shredder taxa richness all increased downstream of the point source after water treatment was initiated. Cluster analysis of aquatic macroinvertebrate community data along with abundance of a metals sensitive mayfly (Rhithrogena hageni) also suggested recovery from metals effects. Response to decreased metal inputs was rapid and biological measurements of impacted sites attained levels comparable to upstream reference sites in two years. Our results suggest that aquatic communities impacted by metals, in the absence of degraded habitat and with nearby colonist pools, will recover quickly if low instream concentrations of toxicants are achieved.     

Functional trait assembly through ecological and evolutionary time

A classic community assembly hypothesis is that all guilds must be represented before additional species from any given guild enter the community. We conceptually extend this hypothesis to continuous functional traits, refine the hypothesis with an eco-evolutionary model of interaction network community assembly, and compare the resultant continuous trait assembly rule to empirical data. Our extension of the “guild assembly rule” to continuous functional traits was rejected, in part, because the eco-evolutionary model predicted trait assembly to be characterized by the expansion of trait space and trait/species sorting within trait space. Hence, the guild rule may not be broadly applicable. A “revised” assembly rule did, however, emerge from the eco-evolutionary model: as communities assemble, the range in trait values will increase to a maximum and then remain relatively constant irrespective of further changes in species richness. This rule makes the corollary prediction that the trait range will, on average, be a saturating function of species richness. To determine if the assembly rule is at work in natural communities, we compared this corollary prediction to empirical data. Consistent with our assembly rule, trait “space” (broadly defined) commonly saturates with species richness. Our assembly rule may thus represent a general constraint placed on community assembly. In addition, taxonomic scale similarly influences the predicted and empirically observed relationship between trait “space” and richness. Empirical support for the model’s predictions suggests that studying continuous functional traits in the context of eco-evolutionary models is a powerful approach for elucidating general processes of community assembly.     

Facilitation promotes invasions in plant‐associated microbial communities

While several studies have established a positive correlation between community diversity and invasion resistance, it is less clear how species interactions within resident communities shape this process. Here, we experimentally tested how antagonistic and facilitative pairwise interactions within resident model microbial communities predict invasion by the plant–pathogenic bacterium Ralstonia solanacearum. We found that facilitative resident community interactions promoted and antagonistic interactions suppressed invasions both in the lab and in the tomato plant rhizosphere. Crucially, pairwise interactions reliably explained observed invasion outcomes also in multispecies communities, and mechanistically, this was linked to direct inhibition of the invader by antagonistic communities (antibiosis), and to a lesser degree by resource competition between members of the resident community and the invader. Together, our findings suggest that the type and strength of pairwise interactions can reliably predict the outcome of invasions in more complex multispecies communities.     

Cedar Infestation Impacts Avian Communities along the Niobrara River Valley,Nebraska

Habitat modification has the potential to cause changes in structure and composition of bird communities. Our goal was to determine the response of Songbird community composition to eastern red cedar (Juniperus virginiana) removal in The Nature Conservancy's Niobrara Valley Preserve, Nebraska. We used point counts to survey birds in the riparian matrix of grassland and forest habitats. More than 60 species were recorded on surveys during 2004–2005. We also use the program PRESENCE to determine the response of five species to various habitat components, including cedar density: House Wren (Troglodytes aedon), Spotted Towhee (Pipilo maculates), Ovenbird (Seiurus aurocapillus), Red‐eyed Vireo (Vireo olivaceus), and Indigo Bunting (Passerina cyanea). Species richness estimates were highest in open and mixed habitat patches. Local populations of Ovenbirds and Red‐eyed Vireos responded positively to cedar density, whereas House Wren numbers declined as cedar density increased. Cedar abundance explained the most variation in bird community similarity between survey points; species richness increased as cedar density decreased. Habitat structure and composition drove variation in community composition and population abundance at fine, local scales within the Preserve. Fine‐scale management to remove cedar from local areas should increase diversity of avian species by maintaining a matrix of habitat types. Cedar removal at any scale will affect the composition of bird communities, and we encourage a structured approach to management decisions.     

Fungal diversity in ectomycorrhizal communities: sampling effort and species detection

A number of recent review articles on ectomycorrhizal (ECM) fungal community diversity have highlighted the unprecedented increase in the number of publications on this ecologically important but neglected area. The general features of these species-rich, highly dynamic and complex communities have been comprehensively covered but one aspect crucial to our assessment of diversity, namely the sampling of ECM communities has received less attention. This is a complex issue with two principal components, the physical sampling strategy employed and the life cycle traits of the ECM fungi being examined. Combined, these two components provide the image that we perceive as ECM diversity. This contribution will focus primarily on the former of these components using a recent study from a pine forest in central Sweden to highlight some sampling problems and also to discuss some features common to ECM communities. The two commonly used elements of diversity, species richness and community evenness, present rather different problems in the assessment of ECM diversity. The applicability of using current measures of abundance (number or percentage of root tips colonised) to determine community evenness is discussed in relation to our lack of knowledge on the size of individual genets of ECM fungi. The inherent structure of most ECM communities, with a few common species and a large number of rare species, severely limits our ability to accurately assess species richness. A discussion of theoretical detection limits is included that demonstrates the importance of the sampling effort (no. of samples or tips) involved in assessing species richness. Species area abundance plots are also discussed in this context. It is suggested that sampling strategy (bulk samples versus multiple collections of single tips) may have important consequences when sampling from communities where root tip densities differ. Finally, the need for studies of the spatial distribution of ECM on roots in relation to small-scale soil heterogeneity and of the temporal aspects of ECM community dynamics is raised.     

Community history affects the predictability of microbial ecosystem development

Microbial communities mediate crucial biogeochemical, biomedical and biotechnological processes, yet our understanding of their assembly, and our ability to control its outcome, remain poor. Existing evidence presents conflicting views on whether microbial ecosystem assembly is predictable, or inherently unpredictable. We address this issue using a well-controlled laboratory model system, in which source microbial communities colonize a pristine environment to form complex, nutrient-cycling ecosystems. When the source communities colonize a novel environment, final community composition and function (as measured by redox potential) are unpredictable, although a signature of the community''s previous history is maintained. However, when the source communities are pre-conditioned to their new habitat, community development is more reproducible. This situation contrasts with some studies of communities of macro-organisms, where strong selection under novel environmental conditions leads to reproducible community structure, whereas communities under weaker selection show more variability. Our results suggest that the microbial rare biosphere may have an important role in the predictability of microbial community development, and that pre-conditioning may help to reduce unpredictability in the design of microbial communities for biotechnological applications.