A continental-scale study—the VEMAP project (VEMAP members, 1995)—demon-strated the interdependence of biogeographical and biogeochemical aspects of the ecological response to climate change; assessing how global change will affect ecosystems and must therefore include both aspects. Such a combined approach is exemplified by the process-based equilibrium terrestrial biosphere model BIOME-3 that simulates vegetation distribution and biogeochemistry (Haxeltine and Prentice, 1996). BIOME-3 predicts plant functional type (PFT) dominance based on environmental conditions, ecophysiological constraints, and resource limitations. The model uses the inputs of temperature, precipitation, cloudiness (Crammer and Leemans, 1991), soil texture, atmospheric pressure, and [CO2]atm (Figure 6.6). The level of [CO2]atm prescribed to BIOME-3 has a direct influence on gross primary productivity via a photosynthetic algorithm and competitive balance between C3 and C4 plants (Haxeltine and Prentice, 1996). Combining these inputs, a coupled carbon and water flux model calculates leaf area index (LAI) and net primary productivity (NPP) for each PFT. The NPP is translated to a series of prescribed PFTs, which then combine to form biomes (Figure 6.6). Although developed as a global vegetation model, BIOME-3 allows simulating changing environmental conditions on vegetation at regional and local scale (Jolly and Haxeltine, 1997; Marchant et al., 2002, 2004a). BIOME-3 can be modified to represent vegetation change within a single pixel that can be used to isolate and manipulate environmental variables—such as temperature, precipitation, seasonal variations of these, and changes in CO2 concentration (Foley et al., 1996; Haxeltine and Prentice, 1996). BIOME-3 model output has been tested and compared against maps of potential vegetation at a global (Prentice et al., 1992; 1993), continental (Jolly et al., 1998; Williams et al., 1998), and regional scale (Marchant et al., 2001a, 2004b). Often, there are discrepancies between model-based
BIOME model structure
Temperature Precipitation Insolation
Soil physical properties
Absolute minimum temperature
2 C3 herbaceous types 1 C4 herbaceous type
3 woody shrubs
2 C3 herbaceous types 1 C4 herbaceous type
3 woody shrubs
Reduction of PFT set
Absolute climatic limits for all PFTs
NPP, autotrophic and heterotrophic respiration and water fluxes on reduced PFT set
Optimization of carbon
Ranking of PFTs and determination of biome
Figure 6.6. A basic summary of the steps taken to calculate biomes on the basis of climatic input variables.
reconstructions and potential vegetation; these discrepancies result from numerous reasons, such as over-simplified soil hydrology (Marchant et al., 2001b).
Of particular interest for understanding vegetation dynamics is how climate system-biosphere interactions have developed since the last glacial maximum (LGM) (Cowling and Sykes, 1999). With this LGM focus in mind, BIOME-3 is run to demonstrate the impact of reducing [CO2]atm to levels (200p.p.m.V) ambient at the LGM (Petit et al., 1999). Reduced concentrations of [CO2]atm have a very significant impact on pan-tropical vegetation (Jolly and Haxeltine, 1997; Boom et al., 2002), as we can see from the modeled output of vegetation (Figure 6.7, see color section). Under low [CO2]atm (Figure 6.7, bottom) the amount of grassland (short and tall), xeric woodland, and scrub increase dramatically, particularly at the expense of Tropical Seasonal Forest (Figure 6.7). Interestingly, the amount of Tropical Rain Forest remains relatively constant as a consequence of largely being under control of changes in temperature and precipitation; the later component needing to change significantly to effect notable changes in Tropical Rain Forest distribution. There are likely to be major within-biome composition dynamics, both in terms of the importance of individual taxa and structure of the vegetation (not portrayed in the biome output). As gaseous exchange at the leaves involves both H2O and CO2 (Figure 6.8), changes in [CO2]atm not only impact on the processes of photosynthesis and photorespiration but also water-use efficiency (WUE) (Cowling and Sykes, 1999). WUE is linearly related to the level of [CO2]atm; under low [CO2]atm plants have to transpire more to achieve the same level of photosynthesis and hence NPP—in other words, halving the [CO2]atm is comparable to halving the rainfall (Farquhar, 1997). Thus, it is likely to be the hydrological impact rather than physiological impacts of lowered CO2 that causes the vegetation to change. Although it has been shown that some C3 plants
can respond to decreased [C02]atm by increasing the amount of stomatal area on the leaf lamella (Wagner et al., 1999), this is difficult to apply to the late glacial as the main impact—rather, the physiological response to low [C02]atm—appears to be reduced WUE (Cowling and Sykes, 1999). Thus, if the stomata have a wider aperture, or are more frequent, this will result in more water being evaporated. Therefore, no matter how the stomata compensate for the variation in [C02]atm, C4 or CAM plants will always have a competitive advantage over C3 plants in warm environments subject to water stress (Ehleringer et al., 1997; Boom et al., 2002).
An interesting application of vegetation models is to use them as a vehicle to display output from climate (Claussen, 1994, 1997) and biogeochemical models (Prentice et al., 1993; Peng et al., 1998). This use allows model output to be translated into maps of potential natural vegetation (Claussen and Esh, 1994; Foley et al., 1996; Prentice et al., 1996; Williams et al., 1998) used for the coupling of biosphere, atmosphere, and oceanic components (Claussen, 1994; Texier et al., 1997), and testing of biogeochemical dynamics (Peng et al., 1998). Climate shifts are displayed as shifts in biome boundaries and areal extent, which can be used to investigate the feedback between atmosphere, biosphere, and oceanic systems under changing boundary conditions (Figure 6.9). However, it must be stressed that, as there are a number of different scenarios available to drive the vegetation model, the results will vary depending on the model output used and feedbacks to the climate system (Figure 6.1). Such differences, particularly when outputs are compared with independent data, can be used to assess model performance and determine the importance of model components—such as land-ocean feedback, or dynamic rather than fixed ocean temperatures.
6.2.2 Dynamic global vegetation models
Extreme environmental events—such as droughts and fires—are important factors in global vegetation processes, yet data on them are sparse, unreliable, or completely lacking. Furthermore, how these are parameterized within models is even worse! Techniques to estimate the rates, extent, and magnitudes of extreme events, and the ability to quantify uncertainty in the results need to be developed. The role of biogeophysical vegetation feedback is not considered, but can be an important aspect of the tropical climate by altering the wet season and thus amplifying the response to orbital forcing (Doherty et al., 2000). These changes affect local, regional, and global climates, which feed back to the biogeography and physiology of the vegetation (Cao and Woodward, 1998). The influence of such dynamics and impact on vegetation can be assessed within a vegetation model developed to run over time (Thonicke et al., 2001). Similarly, internal feedbacks (Figures 6.3 and 6.9) can be incorporated by developing dynamic global vegetation models (DGVMs); of course, this results in a more complex model with associated computing limitations. The potential applications of DGVMs can be summarized within three main areas: simulating the transient changes in global vegetation patterns under future climate change, investigating human disturbance scenarios and estimating the transient behaviors of carbon pools and fluxes to provide fully interactive representation of biosphere within
GCMs to investigate potential vegetation-climate feedback mechanisms (Peng, 2000). To express these changes, the vegetation model needs to run to equilibrium—taking into account vegetation development and impacts of factors, such as fire. Such an approach will have advantages over equilibrium models in driving our understanding of ecosystem dynamics and the impact of a suite of forcing factors.
Was this article helpful?