Coupling Landscape Patterns and Ecological Processes - Landscape Ecology for Sustainable Environment and Culture - page 9

the influences of ecological processes on landscape patterns (Jeltsch et al. 1999 ;
Paruelo et al. 2008 ), and models that couple landscape patterns and ecological

processes (Fig. 1.2 ).

1.3.1 Models on the Effects of Landscape Patterns
on Ecological Processes

To develop such models, the landscape pattern is taken as the input for the
mechanism model. The effect of landscape changes on ecological processes can be
tested by as setting different scenarios of landscape pattern alterations e.g.,
removing certain landscape components or changing the location or area of certain
parts of the landscape. Typical models include SWIM (Soil and Water Integrated
Model; Krysanova et al. 1998 ), THMB (Terrestrial Hydrology Model with Bio-
geochemistry; Li et al. 2007 ) and ANSWERS (Areal Non-point Source Watershed
Environment Response Simulation; Beasley et al. 1980 ). The SWIM model is used
to simulate water flow and nutrient retention processes in riparian and wetland
ecosystems, the underground water and nutrient transfer of plants, as well as the
influence of riparian zones and wetlands on hydrological processes. The THMB
model is used to quantify the influence of forest and grassland on hydrological
processes. The ANSWERS model can assess the influence of vegetation on runoff
and sediment yields. The deficiency of such models lies in the fact that the pattern
change is analyzed through scenario simulation, whereas the factors affecting the
landscape pattern and the potential landscape evolution are not given due con-
sideration, thus compromising their accuracy in predicting the landscape pattern
change and ecological process responses.

1.3.2 Models on the Influences of Ecological Processes
on Landscape Patterns

Modeling the effects of ecological processes on landscape pattern change can be
conducted by assessing the corresponding landscape pattern changes after altering
or removing some input data pertaining to the ecological process, i.e., landscape
pattern is taken as an indicator to reflect the ecological process variation. The
difficulties in establishing such models lie in the chronicity and subtlety of the
inherent traits of ecological processes' effects on landscape patterns, the abundant
driving factors underlying landscape patterns, and insufficiency in basic data and
mechanism. Ambitious trials have been made for such models. For example,
Jeltsch et al. ( 1999 ) simulated the effects of precipitation, fire, and grazing on tree
distribution in savannas by using a spatially-explicit, grid-based model, and found
that higher precipitation increased tree numbers characterized by an enhancement
in tree clumping, whereas lower precipitation, fire, and grazing decreased tree