Applied Vegetation Science appendix to

Applied Vegetation Science

APPENDIX TO:

A new practical tool for deriving a functional signature for herbaceous vegetation

Additional methodological details

The conversion of floristic data into a C-S-R signature is a multi-stage process which is carried out automatically by the first, or ‘calculator’, part of our spreadsheet tool (main paper, Fig. 2).

1. 
   The user pastes-into a labelled, two-dimensional field a data matrix containing quantitative records from one vegetation sample. The format follows the usual conventions, having one named species per row together with a value of percent abundance (any consistent measure may be used). There are as many rows as there are species, up to a maximum of 100. In each use of the tool the original data represent one ‘snapshot’ of a vegetation, arranged into one combination of location, treatment type and/or time of recording. The data can be averaged across replicate samples, if any, or these can be treated separately.
   
2. 
   From internally-stored tables, covering approximately 1000 European species, the tool looks-up the C-S-R identity (functional type) of each of the species presented. If no match is found for a species that fact is flagged. The user then faces three options: (a) to estimate the unknown type from first principles by means of the simple literary and laboratory procedure described by Hodgson et al. (1999), (b) to substitute for the name of the species that of a presumptive C-S-R type chosen from Part 1 of the tool (worksheet ‘Standard coordinates’), (c) to group the taxon, and any others like it, under the single eliminator ‘unknown’.
   
3. 
   Within each of the (maximum nineteen) C-S-R functional types present as known, estimated or presumed components of the vegetation, the tool sums all the individual species abundances into a percent abundance of the whole type. The sample thus becomes redescribed, not in terms of the relative abundance of its plant species, but in terms of the relative abundance of its functional types. Put formally, the percent abundance P of functional type type i is calculated as
   

where the P_j are the percent abundances of the n species that are present both within the sample and within the type.

4. 
   For each type represented, the tool next imports the numerical, three-part, C,S,R coordinate from a copy of the standard list shown in Part 1 of the tool (worksheet ‘Standard coordinates’). The tool then chooses each of the three parts (ordinates) of the C,S,R coordinate in turn and multiplies it by the percent abundance of the type within the sample. Each of these three products is then summed across all types present and a net value is calculated. This provides a single, weighted value for each of the three ordinates, a ‘C-S-R signature’ for the whole community which specifies its mean position in C-S-R space. Thus, in the general case (assuming no unknown species), for types i = 1 to 19 where P_i is again the percent abundance of type i and L^D_i is the standard ordinate of type i in dimension D, then for D = C, S, R in turn the weighted average for each ordinate is calculated as:
   

 P_i L^D_i

100

5. 
   The tool plots this net position onto a triangular representation of C-S-R space (main paper, Fig. 1b). For reference, the tool also reports which of the nineteen standard C-S-R positions lies closest to the calculated signature. Finally, a variety of indices of diversity (Ludwig & Reynolds 1988) are reported, and this is done on the basis of both species and types.
   

Part 2. Comparing C-S-R signatures

Additional details of an example already in the published paper

The full results of the analysis of the Sheffield data appear in Appendix Table 1. In this Table, the equitability index represents the relative evenness of distance between individual abundances when these are arranged in order of abundance; the Shannon-Wiener index represents a composite of equitability and richness; the Simpson index measures degree of dominance; the ‘Number’ is a simple index of species- or type-richness.

Additional worked examples not included in the published paper

Buxton: Treatment effects in a stable community

Among the regimes imposed is one that maintains both elevated winter temperature and summer droughting. From November to April inclusive, computer-controlled heating cables raise soil surface temperatures to 3°C above ambient; throughout July and August, rainfall receipt is inhibited by automatically-operated, translucent rainshelters that deploy during every rainfall event. At the end of the growing season all plots are cut to 5cm. A point-quadrat analysis of the vegetation is done annually in late June/early July, at the time of maximum plant growth and immediately before the imposition of drought (thus the data collected in any one year do not reflect the immediate effects of that year’s droughting).

Results for the period 1994 to 2001, converted to C-S-R signatures, are shown in Appendix Figure 1. The control plots maintained a relatively stable composition over this period. Though year-to-year variations were present, little general trend with time was discernible in the C-, S- or R-components of the community. In contrast, the warm winter/dry summer treatment produced dramatic effects, particularly in the earlier years, with the net signature heading quickly towards the S-corner of C-S-R space (Appendix Fig. 1a). As a result, both the C- and the R-components within the vegetation were progressively reduced (Appendix Figs. 1b,d) and the S-component rose to a new and relatively stable plateau (Appendix Fig. 1c). It is likely that the winter warming promoted the C-component annually at the expense of the R- (Dunnett et al. 1998), but the accentuated summer drought then depleted the C-component, allowing the S-component to achieve a gradual increase.

Despite having almost unchanging C-S-R signatures, the control plots (Appendix Figs. 2a-c, closed symbols) nevertheless exhibited subtle changes over the years in the indices of diversity for both species and types (the scaling of values varies between these two because there are more species than types). Both species richness and species equitability increased in the controls (and also, therefore, the Shannon-Wiener index). Consequently, there was a slight decline in the degree of dominance within the vegetation (Simpson). The effect of the warm winter/dry summer treatment was to arrest, or even reverse, the trends seen in both species and types in the control plots: equitability declined (and the Shannon-Wiener index with it), and the dominance index (Simpson) increased.

In respect of numbers present (Appendix Fig. 2d), the divergence between control and treated plots is less marked in the case of functional types than it is in the case of species. However, the control plots shows a very slow upwards drift with time in both. The net effect of treatment is to arrest the increase in number of species while continuing the slow increase in number of types. Species being lost from the system are slowly replaced by other, fewer species of different types; this is sufficient to bring about the movement seen in the C-S-R signature (Appendix Fig. 1a).

This example shows how the C-S-R signature, together with the indices of diversity, can reveal and summarize by means of direction and velocity of change the effects of a treatment imposed upon a natural vegetation. The subtle interplay between the three components of the signature and the indices of diversity also suggests mechanisms that can explain the treatment’s impacts.

Fribourg: Treatment effects in a changing community

The C-S-R signatures reveal that the control community was not in a stable condition between 1999 and 2000, though it equilibrated more between 2000 and 2002. Despite considerable year-to-year variations (Appendix Fig. 3), opposing trends were discernible in the C- and the S-components of the control plots. There was a net drift in the controls towards the S-corner of C-S-R space, doubtless representing a continued adjustment to the recent cessation of sheep grazing. Annual cutting, which was the replacement regime, was clearly more deleterious towards the C- and R-components of the vegetation than was the grazing. This permitted a corresponding a rise in the S-component. The treatment with ozone appeared to accelerate all of these drifts, being differentially severe on CR-types and their near neighbours. Species with strong C- or R-type features all possess short-lived, fast-decomposing, palatable leaves of high specific leaf area (Grime et al. 1997, Diaz et al., submitted), a configuration which is especially sensitive to both physical and physiological damage by ozone (McKee 1994, BIOSTRESS 2003).

This example demonstrates that despite the short time-period and the initially unstable nature of the control comparison, the C-S-R signature is able to indicate at least approximately the direction and magnitude of possible treatment effects within the community. Statistical analysis is barely feasible under these circumstances, of course, but the signature still provides an early indication of the emergence of signal from noise.
Appendix References

BIOSTRESS 2003. Biodiversity in herbaceous semi-natural ecosystems under stress by global change components. EC FP5 project, reported at http://www.uni-hohenheim.de/biostress/.

Díaz, S., Hodgson, J.G., Thompson, K., Cabido, M., Cornelissen, J.H.C., Jalili, A., Montserrat-Martí, G., Grime, J.P., Zarrinkamar, F., Asri, Y., Band, S.R., Basconcelo, S., Castro-Díez, P., Funes, G., Hamzehee, B., Khoshnevi, M., Pérez-Harguindeguy, N., Pérez-Rontomé, M.C., Shirvany, F.A., Vendramini, F., Yazdani, S., Abbas-Azimi, R., Bogaard, A., Boustani, S., Charles, M., Dehghan, M., de Torres-Espuny, L., Falczuk, V., Guerrero-Campo, J., Hynd, A., Jones, G., Kowsary, E., Kazemi-Saeed, F., Maestro-Martínez, M., Romo-Díez, A., Shaw, S., Siavash, B., Villar-Salvador, P. & Zak, M.R. 2004. The plant traits that drive ecosystems: evidence from three continents. Journal of Vegetation Science 115: 295-304.

Dunnett, N.P., Willis, A.J., Hunt, R., & Grime, J.P. 1998. A 38-year study of relations between weather and vegetation dynamics in road verges near Bibury, Gloucestershire. Journal of Ecology 86: 610-623.

Grime, J.P., Brown, V.K., Thompson, K., Masters, G.J., Hillier, S.H., Clarke, I.P., Askew, A.P., Corker, D. & Kielty, J.P. 2000. The response of two contrasting limestone grasslands to simulated climate change. Science 289: 762-765.

Grime, J.P., Thompson, K., Hunt, R., Hodgson, J.G., Cornelissen, J.H.C., Rorison, I.H., Hendry, G.A.F., Ashenden, T.W., Askew, A.P., Band, S.R., Booth, R.E., Bossard, C.C., Campbell, B.D., Cooper, J.E.L., Davison, A.W., Gupta, P.L., Hall, W., Hand, D.W., Hannah, M.A., Hillier, S.H., Hodkinson, D.J., Jalili, A., Liu, Z., Mackey, J.M.L., Matthews,N., Mowforth, M.A., Neal, A.M., Reader, R.J., Reiling, K., Ross-Fraser, A.M., Spencer, R.E., Sutton, F., Tasker, D.E., Thorpe, P.C. & Whitehouse, J. 1997. Integrated screening validates primary axes of specialisation in plants. Oikos 79: 259-281.

Hodgson, J.G., Wilson, P.J., Hunt, R., Grime, J.P. & Thompson, K. 1999. Allocating C-S-R plant functional types: a soft approach to a hard problem. Oikos 85: 282-294.

Ludwig, J.A. & Reynolds, J.F. 1988. Statistical ecology: a primer in methods and computing. John Wiley & Sons, Chichester.

McKee, D.I. 1994. Tropospheric ozone. Lewis Publications, Boca Raton.

Volk, M., Geissmann, M., Blatter, A., Contat, F. & Fuhrer, J. 2003. Design and performance of a free-air exposure system to study long-term effects of ozone on grasslands. Atmospheric Environment 37: 1341-1350.

A
ppendix Fig. 2.

A
ppendix Fig. 3.