New Approaches to Functional Diversity

While traditional functional groups and species concepts have identified problems, other approaches show promise, at least for identifying sensitivities. Global applications of the Century ecosystem model (Parton et al. 1994; Schimel et al., 1996) were carried out. The model was set up with a relatively large number

Decreasing species richness (log, scale)

Decreasing species richness (log, scale)

2 species ' 11 species

Decreasing functional group richness

Germany

Portugal

Switzerland

Greece

Germany

Portugal

Switzerland

Decreasing species richness (log2 scale)

FIGURE 8 A pan-European field experiment: (a) Overall log-linear reduction of above ground biomass with the simulated loss of plant species richness, (b) Linear reduction with the loss of functional group richness within species richness levels, (c) Biomass patterns at each site (displayed with species richness on a log, scale for comparison with panel (a)) (Hector et al., 1999).

Decreasing species richness (log2 scale)

FIGURE 8 A pan-European field experiment: (a) Overall log-linear reduction of above ground biomass with the simulated loss of plant species richness, (b) Linear reduction with the loss of functional group richness within species richness levels, (c) Biomass patterns at each site (displayed with species richness on a log, scale for comparison with panel (a)) (Hector et al., 1999).

Bayreuth, 1998

1600

1200

O without legumes • with Trifolium sp O with other legumes

Decreasing species richness (log2 scale)

1250

7 1000

ro 250 E

7 1000

ro 250 E

-

Forest

-

Grassland

Sirpt

rtated NPP

1 1 i i i

50 75 100 Actual "évapotranspiration rate- tern yr'1]

Decreasing species richness (log2 scale)

FIGURE 9 The German experiment: above ground biomass production as related to decreasing species richness on three different diversity plots: a plot without legumes, a plot with Trifolium pratensc, and a plot with no Trifolium but other legumes (Scherer-Lorenzen, 1999).

50 75 100 Actual "évapotranspiration rate- tern yr'1]

FIGURE 10 The relationships between évapotranspiration and net primary production as they emerge from the Century model, applied globally. The near-linear realtionship between evaporation and NPP, very similar to those observed in semi-arid lands, is an emergent property. The differences in slope between biomes are largely due to differences in the C:N ratio of different plant functional types, indicating that ecosystem composition has direct effects on biogeochemistry (Scholes et al., 1999).

of biomes, where biome-specific parameterizations were based, as much as possible, on site-level data. Biomes differed in functional plant physiology, and disturbance regime. The results were analyzed in terms of major functional relationships. Figure 10 shows ecosystem-level water use efficiency (NPP vs évapotranspiration) identifying that grasslands have substantially lower ecosystem water use efficiency, despite higher leaf-level water use efficiency deserved frequently (Scholes et al., 1999). This arises because grasslands tend to be more nitrogen limited because of narrower whole-plant C:N ratios and higher chronic N losses due to fires and grazing (which are included explicitly in the model). These model predictions of ecosystem water use efficiency are borne out by recent global data surveys (Parton personal communication). Figure 11 shows a range of nitrogen-related efficiencies (Schimel et al., 1999). In these plots the light gray indicates forest points, while black indicates savanna and grassland ecosystems. Note again, that grassland and savanna systems tend to differ systematically, and, within the two major types, there are further biome differences. The model is run with 28 different biomes, but clearly from a carbon-nitrogen-water point of view, 5 or 6 types emerge as functionally different on large scales. This result has two implications. First, functional diversity matters even on the global scale: the biosphere cannot be treated as a uniform black box. Second, the dimensionality of functional diversity in carbon-nitrogen-water processes is much less than that of the number of species or even biomes. As a caution, for other processes, the dimensionality could be higher than that even of species, e.g., for pathogen resistance where both species and population variability are high. While this analysis is based on a model, sufficient data are emerging from the IGBP and scientific community to perform these types of analyses on observations.

0 0

Post a comment