Article | Open

Consequences of biodiversity loss diverge from expectation due to post-extinction compensatory responses

Published online:


Consensus has been reached that global biodiversity loss impairs ecosystem functioning and the sustainability of services beneficial to humanity. However, the ecosystem consequences of extinction in natural communities are moderated by compensatory species dynamics, yet these processes are rarely accounted for in impact assessments and seldom considered in conservation programmes. Here, we use marine invertebrate communities to parameterise numerical models of sediment bioturbation – a key mediator of biogeochemical cycling – to determine whether post-extinction compensatory mechanisms alter biodiversity-ecosystem function relations following non-random extinctions. We find that compensatory dynamics lead to trajectories of sediment mixing that diverge from those without compensation, and that the form, magnitude and variance of each probabilistic distribution is highly influenced by the type of compensation and the functional composition of surviving species. Our findings indicate that the generalized biodiversity-function relation curve, as derived from multiple empirical investigations of random species loss, is unlikely to yield representative predictions for ecosystem properties in natural systems because the influence of post-extinction community dynamics are under-represented. Recognition of this problem is fundamental to management and conservation efforts, and will be necessary to ensure future plans and adaptation strategies minimize the adverse impacts of the biodiversity crisis.


Species extinction and the associated degradation of ecosystems are proceeding at an accelerating pace1,2,3, but the consequences of the current decline in biodiversity for socio-ecological systems represent a critical knowledge gap for policy-makers4. Consensus reached from experiments that have manipulated biodiversity and measured various ecosystem functions, including primary productivity, nutrient cycling and decomposition, predict an accelerating decline in ecosystem properties with increasing biodiversity loss5, yet it is not clear how appropriate it is to apply this general relationship at the landscape scale6. A major difficulty is that the complexities of natural communities7,8, including the role of rare species9 and the occurrence of co-extinctions10, have been poorly articulated in controlled experiments, and very few studies have focussed on realistic trajectories of species loss that factor in a predisposition to community dynamics11. Indeed, the role of population dynamics in moderating the consequences of extinction have received little attention12,13,14, despite direct evidence that communities undergo fundamental shifts in the relative abundance of taxa and the dominance of species in each successive assemblage that follows an extinction event15,16. Such community turnover resembles that of recovering post-disturbance communities17, where a variety of compensatory interactions amongst surviving species18 develop and offset, wholly or in part, the functional contributions made by species that have been extirpated14.

Compensatory responses tend to be asynchronous within a perturbed community and can lead to to partial19,20, complete21,22,23, or over24 compensation in ecosystem functioning. Species can also switch behaviour25, make physiological regulation adjustments26, exhibit elevated growth (biomass compensation27) or increase reproductive investment (numeric compensation28) in response to perturbation, especially following release from competition/predation29 or during niche expansion30. These responses may not be immediate, in some instances taking months31 to years32 to develop, but once expressed they can be critical in influencing further species interactions33 and can dramatically affect ecosystem properties34,35,36,37. Furthermore, where extinction events are localized and communities are interconnected38, immigration and re-colonization from the surrounding area39, as well as recruitment from the wider species pool40, can play a pivotal role in stabilizing local population decline and ecological processes41. Ultimately, however, the degree of functional compensation will depend on the amount of functional redundancy in the surviving community, which, in turn, will reflect the spatial extent and consistency of the perturbation compromising biodiversity and the level of covariation between the extinction driver and the traits that mediate functioning42.

Despite the range of compensatory mechanisms and variety of ways in which species interactions can affect the functional properties of natural communities7,18, few studies have explored how surviving species moderate the consequences of extinction42,43,44,45 and the relative role of different compensation mechanisms remain largely unexplored. Here, we use probabilistic numerical simulations to test how the loss of sediment dwelling marine invertebrates may affect the sediment mixing depth, an important mediator of biogeochemical cycling46,47. Our simulations assume that the sequence of species loss is either random or ordered by body size or rarity to reflect likely sources of extinction risk. We compare these probabilistic distributions to further simulations in which populations of surviving species maintain total abundance (numeric compensation) or total biomass (biomass compensation) sourced from different components (common or rare species, within or between bioturbation functional groupings, same or different/lower or higher level of activity) of the surviving community. In doing so, our objective is to establish the extent to which alternative compensatory dynamics alter biodiversity-function relationships.


In the absence of compensatory dynamics, the form of the biodiversity function curve approximates expectations (accelerating loss of function with declining species richness, moderated by how extinctions are ordered), and our simulations include a previously documented42 prominent bifurcation of the mixing depth that reflects whether a burrowing brittlestar, Amphiura filiformis, is present (deeper mixing depth) or absent (shallower mixing depth) in the surviving community (panels a,l,w in Figs 1 and 2). The disproportionate contribution of this species is evident throughout all of our simulations (Supplementary Figures S1 and S2), although it is clear that the loss of other species can also result in stepped changes (abrupt shifts in mixing depth, most prominently featured in Fig. 1) that show a tendency to only partially compensate for the loss of the extirpated species (Fig. 1).

Figure 1: Predicted changes in sediment mixing depth following extinction of sediment-dwelling invertebrates and post-extinction numeric compensation.
Figure 1

Simulations (n = 1000 per panel) represent species losses that occur at random (panels a–k) or which reflect trait-based vulnerabilities to extinction governed by body size (panels l–v) or rarity (panels w–ag). We assumed that the surviving community shows either no compensatory response (a,l,w) or full numeric compensation by common (b,m,x) or rare (c,n,r) species, species from within (d,o,z), between (e,p,aa), lower (f,q,ab) or higher (g,r,ac) functional groups, or species with the same (h,s,ad), different (i,t,ae), lower (j,u,af) or higher (k,v,ag) level of activity to the species that have gone extinct. Colour intensity (cold to warm colouration; grey - blue – red - yellow) reflects an increasing density (low to high) of data points.

Figure 2: Predicted changes in sediment mixing depth following extinction of sediment-dwelling invertebrates and post-extinction biomass compensation.
Figure 2

Simulations (n = 1000 per panel) represent species losses that occur at random (panels a–k) or which reflect trait-based vulnerabilities to extinction governed by body size (panels l–v) or rarity (panels w–ag). We assumed that the surviving community shows either no compensatory response (a,l,w) or full biomass compensation by common (b,m,x) or rare (c,n,r) species, species from within (d,o,z), between (e,p,aa), lower (f,q,ab) or higher (g,r,ac) functional groups, or species with the same (h,s,ad), different (i,t,ae), lower (j,u,af) or higher (k,v,ag) level of activity to the species that have gone extinct. Colour intensity (cold to warm colouration; grey - blue - red - yellow) reflects an increasing density (low to high) of data points.

When compensatory mechanisms of the surviving community are based on numerical responses, the mixing depth is largely maintained as species richness declines (Fig. 1), when extinctions are random (panels b–k) or ordered by rarity (panels x–ag). However, when extinctions are ordered by body size (panels m–v), compensatory responses, independent of how they are expressed, are unable to fully mitigate the functional consequences associated with species loss, yet notable differences exist depending on which species drive the compensatory response. However, when compensatory mechanisms of the surviving community are based on biomass, with a few exceptions, there is a tendency for overcompensation to take place along the main species-function trajectory (colour intensity in Fig. 2). Despite implementing these fundamentally different compensatory mechanisms, our simulations reveal that most probabilistic trajectories overlap one another to form a narrow band of likely ecosystem functioning (colour intensity in Figs 1 and 2). The variation of the outcome under numeric versus biomass compensation increases at low and high levels of species richness; this is particularly evident for compensatory responses driven by lower and higher functional groupings and activity level, but at intermediate levels of species richness there is some evidence to suggest similar levels of mixing depth, irrespective of the type of compensation mechanism (Fig. 3). The same general patterns persist in the absence of A. filiformis, although variation is considerably reduced for numeric compensation (Supplementary Figure S3).

Figure 3: Variability of sediment mixing depth following extinction of sediment-dwelling invertebrates and post-extinction compensation.
Figure 3

Standard deviations are shown for the probabilistic distributions (n = 1000) at each level of species richness for the extinction scenarios and compensatory responses depicted in Figs 1 and 2, assuming no compensatory response (blue), full numeric compensation (red) or full biomass compensation (black).

When compensation is realized via the most common species in the surviving community, our models indicate that numeric compensatory mechanisms (Fig. 1, panels b,m,x) can reduce, maintain, or increase the mixing depth relative to when there is no compensation, whilst biomass compensatory mechanisms (Fig. 2, panels b,m,x) show a tendency to increase the mixing depth. In contrast, when compensation is realized by the rarest species in the surviving community (panels c,n,y in Figs 1 and 2), the mixing depth tends to increase relative to the no compensation scenarios under both numeric and biomass compensation, albeit with high variability at low levels of species richness (panels c,n,y in Fig. 3). When compensating species stem from the same or different functional group, or exhibit identical or contrasting levels of activity relative to the species that have been extirpated (compare panels d-e, o-p, z-aa and h-i, s-t, ad-ae between Figs 1 and 2), there is little influence of compensation because compensation is sourced from all possible functional groups. Hence, over multiple species losses, compensation does not occur disproportionately in any one functional group. When compensation is directed towards lower functional groups (Figs 1f,q,ab and 2f,q,ab) and/or activity levels (Figs 1j,u,af and 2j,u,af), the loss of functioning accelerates relative to the other functional group and/or activity level scenarios because bioturbation capacity is reduced. Alternatively, when compensation is directed towards higher functional groups (Figs 1g,r,ac and 2g,r,ac) or activity levels (Figs 1k,v,ag and 2k,v,ag), the mixing depth tends to be maintained at similar or higher levels than those observed when compensation stems from species with the same or different functional group/activity level because bioturbation capacity continues to be maintained or is increased. These observations suggest that certain traits may well be linked to the mediation of ecosystem functioning, but their role in determining observed levels of functioning will depend on how influential such traits are for underlying ecosystem processes48.


Using numerical models parameterised with data from a marine benthic community, we have demonstrated that incorporation of compensatory dynamics, irrespective of the mechanism of compensation (numeric or biomass) or how a species extinction risk is determined (random, ordered by rarity or body size), have the potential to lead to clear differences in aggregate community responses to species loss that do not always conform to expectations based on the generalised biodiversity-function relation curve5. Our models showed that the probabilistic distributions for sediment mixing depth were deeper when compensatory mechanisms were present relative to when they were absent, and when compensatory dynamics reflected post-extinction increases in biomass rather than abundance. Moreover, simulations indicate that the way in which numeric or biomass compensation is expressed within the surviving species pool is most influential. These observations highlight the importance of post-extinction compensatory mechanisms in determining how traits are expressed and mediate function, and suggest that the ecological status of soft-sediment benthic habitats is unlikely to conform to expectation as current visions of future scenarios of extinction lack the necessary sophistication.

Numerous studies investigating the consequences of environmental forcing for ecosystem functioning have focused on the decline of the number of species and their functional traits or attributes. Whilst the effect of various attributes of biodiversity on ecosystem functioning, including evenness49, dominance50, and functional traits51 have been studied, the relative importance of compensatory mechanisms in natural ecosystems and the concept of compensation as a whole has largely been ignored despite evidence for the occurrence of compensation in natural systems52. Importantly, our simulations reveal that the mixing depth of sediment-dwelling invertebrate communities will depend on how compensatory behaviour is expressed, and the extent to which the functional attributes of compensating species affect bioturbation. A difficulty with determining the latter is that recent work has shown that differences in how species interact with sediment biogeochemistry and other aspects of the environment can converge in terms of absolute effects of ecosystem properties48,51. This may explain previous inconsistencies in linking particular species traits to ecosystem functioning47 and why community compensation does not appear to be dominated by a limited number of species that host specific sets of traits; a conclusion consistent with recent studies that demonstrate the importance of common species in maintaining ecosystem functioning53,54,55. Whilst the differences between alternative compensatory scenarios may be subtle, they can form crucial differences. Consequently, the repercussions of species loss are more effectively offset by a subset of species that share the same functional group or level of activity, especially at low levels of species richness. Whether these mechanisms can be identified and operate in natural communities, however, remains an open empirical question56.

It is important to acknowledge that our study is an abstraction of community dynamics and to recognize that the model assumptions we adopted represent an oversimplification of community interactions under environmental forcing. Our focus was not to predict the depth of sediment mixing for specific biodiversity futures, but rather to explore the relative importance of compensatory mechanisms in determining ecosystem properties. Whilst we were unable to incorporate the occurrence of co-extinctions10, non-indigenous invasive species57 or other cascading effects that can have further consequences for community structure and ecosystem functioning, we were able to establish divergent patterns in response for alternative extinction scenarios that hold promise for exploring new strategies of ecosystem management and governance. An important next step in predicting future biodiversity change, however, is to quantify the prevalence of local extinction drivers in the ecological landscape58 and understand how these interact in natural systems59,60 to influence the risk of extinction, altering community dynamics and ecosystem properties, both locally and across regional scales.

Extensive uncertainties exist in the responses of species and communities to environmental forcing, hence the use of empirically-based scenarios of the future to explore the potential consequences of species loss will continue to be a necessity for ecological advancements. We have shown, that the incorporation of important aspects of post-extinction community dynamics can lead to sharp contrasting forecasts of future ecosystem properties. Such information will help advance the predictability of community responses to change, provided that regionalised vulnerability assessments that determine the response of functionally important species under realistic future environmental conditions become available61,62. However, it is unlikely that all of these details will be available and incorporated into next generation models in the short-term, nor is this likely to be necessary. Consistency in community responses in the presence of compensation reveal patterns that may be general. Based on the available evidence, we should expect that the loss of species will be compensated by less efficient species over the long term63, resulting in alterations to ecosystem properties.

However, the discrepancy in ecosystem consequences between biomass and numeric compensatory responses emphasise the need to identify which, or whether both, compensatory processes prevail in natural assemblages. Estimates of the functional consequences of biodiversity loss that incorporate the error associated with such variation are needed, and will allow more confidence in simulations of the future and provide improved levels of certainty on the consequences of future global change.


Sampling and study site

Field data were collected at station Margaretta (22 m water depth, 53 ° 13.50′N, 09 ° 6.50′W) in Inner Galway Bay on the central west coast of Ireland. Samples of macro-invertebrates (retained on a 500 sieve; n = 5, 0.1 m2 van Veen grab) were collected approximately on a monthly basis over a one-year period (December 1996–November 1997, n = 11), returning a total of 139 invertebrate species42. Measurements of the sediment mixing depth were obtained using sediment profile imaging (SPI; n = 10) camera system42.

Extinction simulations

Using a comprehensive study of the macrofaunal assemblages of Galway Bay, west coast of Ireland42, we predict how species extinction is likely to affect the mixing depth, an indicator of invertebrate bioturbation. We established the relationship between an index of bioturbation potential that uses per capita bioturbation potential (BPi, Supplementary Equation S1) to estimate population-level (BPp, Equation Supplementary S2) and community-level bioturbation potential (BPc, Supplementary Equation S3), which accounts for each species body size, abundance, activity level (4 levels, scored on a categorical scale reflecting increasing activity, 1 = in a fixed tube, 2 = limited movement, 3 = slow movement through sediment profile, 4 = free movement via burrow system42), and mode of sediment mixing (5 levels, scored on a categorical scale reflecting increasing impact on sediment turnover, 1 = epifauna, 2 = surfical modifiers, 3 = upward or downward conveyorbelt feeders, 4 = biodiffusers, 5 = regenerators42), and measurements of mixing depth obtained from sediment profile images. This relationship (Supplementary Equation S4) was used to parameterise probabilistic, numerical simulations that test how alternative extinction scenarios might affect sediment mixing depths. As environmental forcing in natural systems can target different components of the community64, we consider simulations in which species go extinct at random (, where n = the number of species) versus extirpations ordered by body size (largest expire first) or rarity (least abundant expire first). As the functional consequences of extinction also depend on the response of surviving species, we developed models in which species either do not exhibit compensatory responses or in which the abundance (numeric compensation) or biomass (biomass compensation) of the surviving community are held constant following extinction. In doing so, we recognized that compensating species are not randomly assigned, rather they represent different components of the species pool:

Compensation by common species

Compensatory responses by common species is arguably one of the most likely pathways of compensation in a community, both probabilistically in terms of their relative proportion of abundance, and ecologically as numeric success reflects their disproportionate share of resources and competitive advantage over less numerous species19. In addition, this type of compensation has been observed in natural communities52.

Compensation by rare species

The majority of community species are rare, but some species are equipped with unique functional traits65 and may become important if they increase in abundance66,67.

Compensation by species from within/outwith or lower/higher functional grouping or which exhibit similar/distinct or lower/higher levels of activity

We assumed that species within the same functional group (e.g. sediment reworking mode42) will have similar functional traits and thus employ a similar ecological role, and that functional buffering can also be carried out by species from a different but adjacent functional group (e.g. grass versus forbs19, local versus non-local distribution of sediment particles by bioturbators42,68), or species selected from any lower or higher functional group. Compensating species can also exhibit similar or contrasting levels of activity (mobility categorisation42) relative to the species that has been extirpated, such that functional buffering can also be carried out by species from a different but adjacent mobility group, or species selected from any lower or higher mobility group. In the absence of an adjacent functional group, compensation stems from the next available functional group of greater/lesser or equal standing.

Each of these model scenarios (i.e. 3 extinction orders, 2 compensation mechanisms and 11 compensation types, n = 66) was run for 1000 simulations (from 139–1 species). We provide the code for executing each of these simulations in Supplementary Material (Code S1).

Additional Information

How to cite this article: Thomsen, M. S. et al. Consequences of biodiversity loss diverge from expectation due to post-extinction compensatory responses. Sci. Rep. 7, 43695; doi: 10.1038/srep43695 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    et al. Scenarios for global biodiversity in the 21st century. Science 330, 1496–1501 (2010).

  2. 2.

    et al. The biodiversity of species and their rates of extinction, distribution, and protection. Science 344, 1246752 (2014).

  3. 3.

    Accelerating extinction risk from climate change. Science 348, 571–573 (2015).

  4. 4.

    et al. Spatial and temporal changes in cumulative human impacts on the world’s ocean. Nat. Commun 6, 7615; 10.1038/ncomms8615 (2015).

  5. 5.

    et al. Biodiversity loss and its impact on humanity. Nature 486, 59–67 (2012).

  6. 6.

    et al. A general biodiversity-function relationship is mediated by trophic level. Oikos, 10.1111/oik.03652 (2016).

  7. 7.

    , & Complex species interactions and the dynamics of ecological systems: Long-term experiments. Science 293, 643–650 (2001).

  8. 8.

    et al. Marine biodiversity and ecosystem function relationships: The potential for practical monitoring applications. Estuar. Coast. Shelf S. 161, 46–64 (2015).

  9. 9.

    , & Rarity in mass extinctions and the future of ecosystems. Nature 528, 345–351 (2015).

  10. 10.

    et al. Species coextinctions and the biodiversity crisis. Science 305, 1632–1634 (2004).

  11. 11.

    , , & Functional consequences of realistic biodiversity changes in a marine ecosystem. Proc. Natl Acad. Sci. USA 105, 924–928 (2008).

  12. 12.

    & Food-web interactions govern the resistance of communities after non-random extinctions. Nature 429, 174–177 (2004).

  13. 13.

    & The functional consequences of random vs. ordered species extinctions. Ecol. Lett. 8, 409–418 (2005).

  14. 14.

    & The causes and consequences of compensatory dynamics in ecological communities. Ann. Rev. Ecol. Evol. Syst. 40, 393–414 (2009).

  15. 15.

    , & Abundance distributions imply elevated complexity of post-Paleozoic marine ecosystems. Science 314, 1289–1292 (2006).

  16. 16.

    Life in the aftermath of mass extinctions. Curr. Biol. 25, 941–952 (2015).

  17. 17.

    et al. Depletion, degradation, and recovery potential of estuaries and coastal seas. Science 312, 1806–1809 (2006).

  18. 18.

    , & Models for biotic survival following mass extinction. Geol. Soc. Spec. Publ. 102, 41–60 (1996).

  19. 19.

    & Dominant species maintain ecosystem function with non-random species loss. Ecol. Lett. 6, 509–517 (2003).

  20. 20.

    , , , & Extirpation-resistant species do not always compensate for the decline in ecosystem processes associated with biodiversity loss. J. Ecol. 100, 1475–1481 (2012).

  21. 21.

    , , , & Ecosystem stability and compensatory effects in the Inner Mongolia grassland. Nature 431, 181–184 (2004).

  22. 22.

    & Compensatory responses to loss of warming sensitive plant species. Ecology 88, 740–748 (2007).

  23. 23.

    , , & Species compensation maintains abundance and raid rates of African Swarm-raiding army ants in rainforest fragments. Biol. Conserv. 142, 668–675 (2009).

  24. 24.

    et al. Five Years of Experimental Warming Increases the Biodiversity and Productivity of Phytoplankton. PLoS Biol. 13, 10.1371/journal.pbio.1002324 (2015).

  25. 25.

    , & Switching behaviour, coexistence and diversification: comparing empirical community-wide evidence with theoretical predictions. Ecol. Lett. 11, 802–808 (2008).

  26. 26.

    , & Short- and long-term behavioural, physiological and stoichiometric responses to predation risk indicate chronic stress and compensatory mechanisms. Oecologia 181, 347–357 (2016).

  27. 27.

    , , & The consequences of species loss on ecosystem nitrogen cycling depends on community compensation. Oecologia 149, 141–149 (2006).

  28. 28.

    & Factors Influencing Stream Fish Recovery following a Large-Scale Disturbance. T. Am. Fish. Soc. 126, 895–907 (1997).

  29. 29.

    & Compensation in resource use by foragers released from interspecific competition. J. Exp. Mar. Biol. Ecol. 185, 219–233 (1995).

  30. 30.

    The nature of niche expansion in west Indian Anolis Lizards I: Ecological consequences of reduced competition. Evolution 30, 659–676 (1976).

  31. 31.

    & Long-term effects of warming and ocean acidification are modified by seasonal variation in species responses and environmental conditions. Philos. Trans. R Soc. London [Biol.] 368, 1–11 (2013).

  32. 32.

    & Species loss, delayed numerical responses, and functional compensation in an antbird guild. Ecology 92, 1126–1136 (2011).

  33. 33.

    Individual behaviour and higher-order species interactions. Am. Nat. 140, S5–S32 (1992).

  34. 34.

    & Numerical and per capita responses to species loss: mechanisms maintaining ecosystem function in a community of stream insect detritivores. Oikos 93, 221–234 (2001).

  35. 35.

    , & Density-dependent body growth reduces the potential of marine reserves to enhance yields. J. of Appl. Ecol. 43, 61–66 (2006).

  36. 36.

    Fish population regulation beyond “Stock and recruitment”: the role of density-dependent growth in the recruited stock. B. Mar. Sci. 83, 183–196 (2008).

  37. 37.

    , & Trophic compensation reinforces resistance: herbivory absorbs the increasing effects of multiple disturbances. Ecol. Lett. 18, 182–287 (2015).

  38. 38.

    & Resilience to climate change in coastal marine ecosystems. Ann. Rev. Mar. Sci. 5, 371–392 (2013).

  39. 39.

    , & Habitat structure mediates biodiversity effects on ecosystem properties. Proc. R. Soc. Lond. [Biol.] 278, 2510–2518 (2011).

  40. 40.

    , , , & Estimating recruitment and survival in partially monitored populations. J. of Appl. Ecol. 53, 73–82 (2016).

  41. 41.

    & Regional zooplankton biodiversity provides limited buffering of pond ecosystems against climate change. J. Anim. Ecol. 81, 251–259 (2012).

  42. 42.

    et al. Extinction and ecosystem function in the marine benthos. Science 306, 1177–1180 (2004).

  43. 43.

    et al. Species loss and aboveground carbon storage in a tropical forest. Science 310, 1029–1031 (2005).

  44. 44.

    , , & Fish extinctions alter nutrient recycling in tropical freshwaters. Proc. Natl Acad. Sci. USA 104, 4461–4466 (2007).

  45. 45.

    , , & Biogeochemical implication of biodiversity and community structure across multiple coastal ecosystems. Ecol. Mon. 85, 117–132 (2015).

  46. 46.

    , , , & Impact of burrowing crabs on C and N sources, control, and transformations in sediments and food webs of SW Atlantic estuaries. Mar. Ecol. Prog. Ser. 293, 155–164 (2005).

  47. 47.

    , , & How biodiversity affects ecosystem process: implications for ecological revolutions and benthic ecosystem function. Aquat. Biol. 2, 289–301 (2008).

  48. 48.

    et al. Same pattern, different mechanism: Locking onto the role of key species in seafloor ecosystem process. Sci. Rep. 6, 26678, 10.1038/srep26678 (2016).

  49. 49.

    et al. Consistent effects of biodiversity on ecosystem functioning under varying density and evenness. Folia Geobot. 48, 335–353 (2013).

  50. 50.

    , & Specific arrangements of species dominance are more influential than evenness in maintaining ecosystem properties. Sci. Rep. 6, 39325, 10.1038/srep39325 (2016).

  51. 51.

    , , & Functional richness, functional evenness and functional divergence: the primary components of functional diversity. Oikos 111, 112–118 (2005).

  52. 52.

    et al. Effects of functional diversity loss on ecosystem functions are influenced by compensation. Ecology 97, 2293–2302 (2016).

  53. 53.

    et al. Spatial turnover in the global avifauna. Proc. R. Soc. Lond. [Biol.] 274, 1618 (2007).

  54. 54.

    , , & Abundance of common species, not species richness, drives delivery of a real-world ecosystem service, Ecol. Lett. 18, 626–635 (2015).

  55. 55.

    , , & Key role in ecosystem functioning of scavengers reliant on a single common species. Sci. Rep. 6, 29641, 10.1038/srep29641 (2016).

  56. 56.

    , , & Functional consequences of realistic biodiversity changes in a marine ecosystem. Proc. Natl Acad. Sci. USA 105, 924–928 (2008).

  57. 57.

    , , & Global ecological impacts of invasive species in aquatic ecosystems. Glob. Change Biol. 22, 151–163 (2016).

  58. 58.

    et al. Recent trends in local-scale marine biodiversity reflect community structure and human impacts. Curr. Biol. 25, 1938–1943 (2015).

  59. 59.

    , & Synergies among extinction drivers under global change. Trends Ecol. Evol. 23, 453–460 (2008).

  60. 60.

    & Quantifying the evidence for ecological synergies. Ecol. Lett. 11, 1278–1286 (2008).

  61. 61.

    et al. Life history and spatial traits predict extinction risk due to climate change. Nat. Clim. Change 4, 217–221 (2014).

  62. 62.

    , , & Modelling the sensitivity of life history traits to climate change in a temporary pool crustacean. Sci. Rep. 6, 29451, 10.1038/srep29451 (2016).

  63. 63.

    , , , & Body size as a predictor of species loss effect on ecosystem functioning. Sci. Rep. 4, 4616, 10.1038/srep04616 (2014).

  64. 64.

    , , , & Ecological selectivity of the emerging mass extinction in the oceans. Science 353, 1284–1286 (2016).

  65. 65.

    & Rare species loss alters ecosystem function – invasion resistance. Ecol. Lett. 4, 358–365 (2001).

  66. 66.

    , , & Diversity loss with persistent human disturbance increases vulnerability to ecosystem collapse. Nature 494, 86–89 (2013).

  67. 67.

    et al. The importance of rare species: a trait-based assessment of rare species contributions to functional diversity and possible ecosystem function in tall-grass praries. Ecol Evol. 4, 104–112 (2014).

  68. 68.

    Organism activity levels predict marine invertebrate survival during ancient global change extinctions. Glob. Change Biol. 10.1111/gcb.13484 (in press).

Download references


Supported by a scholarship (awarded to M.S.T., grant reference NE/L002531/1) co-funded by the Southampton Partnership for Innovative Training of Future Investigators Researching the Environment (SPITFIRE, a NERC Doctoral Training Partnership) and the Centre for Environment, Fisheries and Aquaculture Science (Cefas, reference DP371C), UK. C.G., M.S., J.A.G., S.B. and R.P. were supported by Work Package 2 of the Shelf Sea Biogeochemistry Programme (NE/K001906/1, 2011–2017), jointly funded by the Natural Environment Research Council (NERC) and the Department for Environment, Food and Rural Affairs (Defra) in the UK. The views expressed are those of the authors and do not necessarily represent those of NERC or Defra.

Author information


  1. Ocean and Earth Science, National Oceanography Centre Southampton, University of Southampton Waterfront Campus, European Way, Southampton. SO14 3ZH, UK

    • Matthias S. Thomsen
    • , Jasmin A. Godbold
    •  & Martin Solan
  2. Cefas Laboratory, Pakefield Road, Lowestoft, Suffolk, NR33 0HT, UK

    • Clement Garcia
    • , Stefan G. Bolam
    •  & Ruth Parker
  3. Biological Sciences, University of Southampton, Highfield Campus, Southampton. SO17 1BJ, UK

    • Jasmin A. Godbold


  1. Search for Matthias S. Thomsen in:

  2. Search for Clement Garcia in:

  3. Search for Stefan G. Bolam in:

  4. Search for Ruth Parker in:

  5. Search for Jasmin A. Godbold in:

  6. Search for Martin Solan in:


M.S.T., M.S. and J.A.G. designed the study, C.G. and M.S.T. modified the R script from Solan et al.42. M.S.T., M.S., J.A.G., S.B. and R.P. contributed to the manuscript.

Competing interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to Matthias S. Thomsen.

Supplementary information


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Creative CommonsThis work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit