Search for a project

The spatial scaling of biodiversity:function relationships

Prof William Kunin (SoB), Dr Wolfgang Buermann (SEE)

Contact email: w.e.kunin@leeds.ac.uk

Project summary

Species-rich ecological communities often show increased levels of function, but the scaling properties of this relationship are seldom explored.  There are good reasons to think that the relationship between biodiversity and function should be scale-dependent.  Biodiversity is "sub-additive", so that the total number of species in two plots combined is usually less than the sum of the plots examined separately (due to overlaps in species).  However, most ecosystem functions (e.g. production, carbon storage) are additive.  This implies that the relationship between biodiversity and function must shift with scale, although little is known about the nature of that scale shift.  This studentship will be devoted to exploring these issues, combining modelling, observational work on natural and experimental plant communities, and remote sensing work.  The work may focus in particular on the functional importance of rare species in plant communities, to test the degree to which ecosystem service arguments can be used to justify conservation efforts.  Depending on the skills and interests of the student, the project could include some mixture of the following:

(A) Mechanistic simulation modeling, testing how competition between functionally different plant species could produce scale-specific biodiversity function relationships, and exploring how such relationships would appear when assessed at inappropriate scales;

(B) Measurement of the relationship between one or more ecosystem functions (e.g. C fixation, total evapotranspiration, erosion control) and species richness or abundance-weighted diversity across multiple scales in natural vegetation differing in species richness;

(C) Testing ideas more mechanistically, using experimentally manipulated plant communities in the field or glasshouse differing in richness, relative abundance and spatial population structure, and potentially making use of existing vegetation diversity experimental plots.

(D) Assessing the potential of remotely sensed data (both from hyperspectral satellites and from UAV-mounted sensors) to be used in assessing vegetation diversity in natural and experimental plant communities.

The studentship would explore an important fundamental topic in ecosystem ecology, but also one of potential importance in conservation.  It would also provide a wide skills portfolio as a foundation for a research career.

Background

There is a large and growing body of evidence demonstrating links between the diversity of organisms in a site and various aspects of ecosystem function and service provision (e.g. Loreau et al. 2002, Kinzig et al. 2002).  This is particularly well-studied in plant communities: Species-rich vegetation generally shows increased levels of productivity (Duffy et al. 2017) and nutrient retention (Scherer-Lorenzen et al. 2003), reduced erosion (Berendse et al. 2015) resistance to invasion (Kennedy et al. 2002) and other valuable functions. 

Figure 1:  Plants provide valuable ecosystem functions: fixing carbon, stabilising soils, slowing runoff and moderating climates.  These functions may be affected by plant diversity, but the scaling propertied of such links are largely unknown. (sources: WEK, NASA)

There are, however, reasons to think that biodiversity:function relationships ought to be scale-specific. Species richness scales in a “sub-additive” fashion: if two adjacent plots of land each contain 10 species, the two combined are unlikely to have 20 species, as the sets of species are likely to overlap.  Many ecosystem functions, on the other hand, scale additively: if the two plots each fix 100 g of carbon, then together they will fix 200 g.  If biodiversity and function scale differently, then the relationship between them is virtually certain to shift with scale: one might expect a strong relationship at some specific scale (or range of scales) but little or no relationship at finer or coarser scales.  This in turn may help explain some of the variation in research findings between studies on biodiversity effects.

The role of rare species in community-level function is a particularly important topic for research.  In most natural communities the vast majority of species are rare, but most individuals represent common species, and such species probably dominate most functional roles. The biodiversity:function relationship has frequently been used as a justification for biodiversity conservation, but it is unclear to what extent rare species (in greatest need of conservation) can provide important ecosystem services (Lyons et al. 2005, Ridder 2008, Kleijn et al 2015).  However, rare species may harbour unusual functional trait combinations (Isbell et al. 2011, Mouillot et al 2013), or may help ensure resilience of function (Mori et al 2013).  Rare species differ dramatically in the spatial patterns of their populations (Rabinowitz 1981), and this in turn may affect their role in ecosystem functioning.  Developing clear experimental and observational tests of rare species’ functional roles in natural or experimental communities would be an important goal.

If vegetation diversity plays an important role in ecosystem function and service provision, it would be useful to find efficient means of assessing it remotely.  There has been substantial progress in classifying British plant communities from remotely sensed imagery using machine-learning algorithms such as RandomForest (Bradter et al. 2011), and work in progress suggests that vegetation diversity can also be assessed with some confidence.   Further development of such methods would allow the mapping of diversity and potentially of function at a regional or national scale.

Methodology

Figure 2:  Island systems (such as this area in Mallorca) typically have much lower species richness than comparable mainland areas, providing an excellent test system for biodiversity:function relationships. (Source: WEK)

The project would ideally involve a mixture of modelling, observational and experimental research.  Some likely components include the following, although the breadth and depth of each will depend on the student’s skills and interests.

A) Ecosystem function modelling.  Students with an interest or aptitude for modelling might construct simple spatially explicit biodiversity: function models.  Communities could be build up using simple neutral or niche-based algorithms, with function values derived from species richness or diversity (e.g. Shannon index) values in a moving-window of specified scale.  The relationship between diversity and function could then be assessed at a range of scales, to test how inappropriate scale assays affect the perceived strength of biodiversity:function links.

B) Observational studies of biodiversity function relationships across scales. Natural communities differing in species richness (e.g. from islands and comparable mainland habitats; Fig. 2) can be surveyed for species richness at multiple nested scales, and specific ecosystem functions (e.g. carbon flux, evapotranspiration) measured over the same areas.  Pilot work on two sites in Germany shows signs of scale dependence in biodiversity effects, but the functional measures need further work.  This project would be an excellent opportunity to carry out such work.

Figure 3:  The project might take advantage of collaborative links with existing plant biodiversity experiments, such as this one in Germany. (source)

C) Experimental communities and mechanistic tests.   Observational work is useful, but stronger inferences about biodiversity effects on function can be made through experimental manipulation. The project could involve novel field or greenhouse experiments, in which the richness, relative abundances and/or spatial patterning of vegetation was manipulated at multiple scales (e.g. Gunton & Kunin 2007), and the functional consequences explored.  In addition, the project could take advantage of collaborative links to existing plant diversity experiments elsewhere (Fig. 3), conducting multi-scale biodiversity and function assays on contrasting treatment plots.

Figure 4.  Vegetation classes in the Yorkshire Dales National Park, assessed using remote sensing data and RandomForest algorithms (source: Bradter et al. 2011l)

D) Remote sensing of vegetation diversity. If vegetation diversity plays a key role in ecosystem functioning, it would be useful to have efficient methods for assessing it over wide areas.  Increasingly sophisticated and high resolution remote sensing and environmental data, combined with machine-learning algorithms such as RandomForest, have the potential to greatly improve our understanding of vegetation patterns.  Past work (e.g. Fig. 4) has demonstrated that vegetation communities can be inferred with substantial accuracy, but further developments could allow vegetation diversity patterns to be assessed.

Expected outcomes

The project should produce several publishable papers, with at least one of high impact, as several of the issues addressed here are little-explored, and of potentially wide scientific interest.

Requirements

A strong undergraduate (and ideally masters) degree in ecology, ecosystem science or physical geography is expected, although candidates with strong backgrounds in other relevant fields will be considered.  Experience with computer programming, remote sensing or other advanced analytical methods would be helpful.

Training

The project would provide students with a wide set of analytical tools, ranging from computer programming, ecosystem process assays, remote sensing experience, and experience in machine learning and other advanced statistical techniques.  The work would blend observational work on natural communities with experimentation and simulation, providing a broad foundation for an academic research career. 

Research context and partners

Both Prof. Kunin and Dr. Buermann have active research groups and strong records of relevant research: in spatial ecology and environmental informatics respectively.  The student will be involved in weekly team meetings, as well as having access to both formal (Faculty) and informal (Ecology & Evolution group) seminar series through both host departments.

Further reading and bibliography

Berendse F, J van Ruijven, E Jongejans & S Keesstra (2015) Loss of plant species diversity reduces soil erosion resistance.  Ecosystems 18: 881-888.

Bradter U, TJ Thom, JD Altringham, WE Kunin & TG Benton (2011)  Prediction of National Vegetation Classification communities in the British uplands using environmental data at multiple spatial scales, aerial images and the classifier random forest. Journal of Applied Ecology 48: 1057-1065.

Duffy JE, CM Godwin & BJ Cardinale (2017) Biodiversity effects in the wild are common and as strong as key drivers of productivity. Nature 549: 261-264.

Gunton RM and WE Kunin 2007. Density effects at multiple scales in an experimental plant population.  Journal of Ecology 95: 435-445.

Isbell, F, V Calcagno, A Hector, J Connolly, WS Harpole, PB Reich, M Scherer-Lorenzen, B Schmid, D Tilman, J van Ruijven, A Weigelt, BJ Wilssey, ES Zavaleta & M Loreau (2011) High plant diversity is needed to maintain ecosystem services. Nature 477: 199-202

Kennedy TA, S Naeem, KM Howe, JMH Knops, D Tilman & P Reich (2002) Biodiversity as a barrier to ecological invasion. Nature 417: 636-638.

Kinzig AP, SW Pacala & D Tilman eds (2002) The functional consequences of Biodiversity. Princeton University Press, Princeton NJ.

Kleijn D. R Winfree, [55 others] & SG Potts (2015) Delivery of crop pollination services is an insufficient argument for wild pollinator conservation. Nature Communications, DOI: 10.1038/ncomms8414

Loreau M, S Naeem & P Inchausti eds (2002) Biodiversity and ecosystem functioning. Oxford University Press, Oxford.

Lyons KG, CA Brigham, BH Traut & MW Schwartz (2005) Rare species and ecosystem functioning.  Conservation Biology 19: 1019-1024.

Mori AS, T Furukawa & T Sasaki (2013) Response diversity determines the resilience of ecosystems to environmental change. Biological Reviews 88: 349-364.

Mouillot, D, DR Bellwood, C Baraloto, J Chave, R Galzin, M Harmelin-Vivien, M Kulbicki, S Lavergne, S Lavorel, N Mouquet, CET Paine, J Renaud & W Thuiller (2013) Rare species support vulnerable functions in high-diversity ecosystems.  PLoS Biology 11: e1001569

Rabinowitz, D (1981) Seven forms of rarity. In H Synge (ed.) The Biological Aspects of Rare Plant Conservation, pp. 205-218. John Wiley & Sons, Chichester.

Ridder B (2008) Questioning the ecosystem services argument for biodiversity conservation. Biodiversity Conservation 17: 781-790.

Scherer-Lorenzen M, C Palmborg, A Prinz & ED Schulze (2003) The role of plant diversity and composition for nitrate leaching in grasslands. Ecology 84: 1539-1552.

Violle C, W Thuiller, N Moquet, F Munoz, NJB Kraft, MW Cadotte, SW Livingstone & D Mouillot (2017) Functional rarity: the ecology of outliers. Trends in Ecology & Evolution 32:356-367.

Related undergraduate subjects:

  • Biodiversity
  • Biodiversity conservation
  • Bioinformatics
  • Biology
  • Botany
  • Computing
  • Conservation
  • Conservation biology
  • Earth system science
  • Ecology
  • Environmental biology
  • Environmental conservation
  • Environmental science
  • Geography
  • Hydrology
  • Physical geography
  • Remote sensing
  • Spatial ecology