Agriculture Reference
In-Depth Information
and the PhenologyMMS decision support
tool (http://arsagsoftware.ars.usda.gov), where
the number of leaves or thermal time between
developmental events from emergence through
maturity is adjusted due to water and N
levels.
Beginning in the mid-1980s, some research
efforts directed greater attention at incorporating
the developmental concepts discussed in this
chapter into process-based simulation models
(e.g., leaf appearance and tillering) and explora-
tion of canopy architecture. Developmental pro-
cesses other than phenology have been incorporated
to varying degrees in whole-plant simulation
models. The AFRCWHEAT1/2 model added
two interesting components generally missing in
existing wheat models at that time: a population
ecology element and greater developmental detail
derived from European scientists. For example,
detailed tillering and leaf dynamics (e.g., appear-
ance, growth, and senescence-abortion) and the
effect on canopy LAI were simulated and then
used to estimate biomass.
Simultaneously and independently another
effort was underway in the US that resulted
in the developmentally driven SHOOTGRO
(McMaster et al., 1991, 1992a,b; Wilhelm et al.,
1993; Zalud et al., 2003) and MODWht3
(Rickman et al., 1996) models. SHOOTGRO is
slightly more developmentally detailed than
MODWht3, but less detailed in the root system
and simulating biomass production. SHOOT-
GRO provides the foundation to simulate the
development and growth of each morphologi-
cally identifi ed shoot (main stem and tillers) on
the median plant of up to six age classes, or
cohorts, based on time of seedling emergence
(http://arsagsoftware.ars.usda.gov). All pro-
cesses in Fig. 2.2 except for leaf primordia initia-
tion and fl oret primordia differentiation are
simulated. Soil water content determines the
thermal time required for germination, and seed-
ling emergence rates are simulated to establish
the cohorts. Following germination, sequential
developmental events are simulated using the
number of leaves produced (e.g., phyllochron)
between developmental events up to anthesis,
and thermal time after anthesis. SHOOTGRO
explicitly includes the effect of water and N
availability on all developmental and growth pro-
cesses. As shoots appear, the appearance, growth
or size, and senescence or abortion of each leaf
blade and sheath, internode, and spike compo-
nents on each shoot are simulated. Spike devel-
opment and growth is simulated by the appearance
of spikelets and spikelet differentiation into
fl orets, fertilization of fl orets, and subsequent
growth of each kernel.
The Sirius model has one of the most devel-
oped leaf appearance submodels of any wheat
simulation model (Jamieson et al., 2007). As with
the SHOOTGRO model, the assumption used is
that the developmental “clock” from emergence
to anthesis is best represented by the rate of leaf
appearance and fi nal number of leaves. Based on
vernalization requirement and photoperiod sensi-
tivity of the cultivar being simulated and on leaf
ontogeny, the fi nal leaf number is determined
(Brooking et al., 1995; Brooking 1996; Robertson
et al., 1996). This allows for an elegant quantita-
tive description of both spring and winter wheat
leaf appearance and integration with developmen-
tal events.
Canopy architectural, or functional-structural,
modeling for a variety of species has increased
dramatically since the mid-1990s, in part a result
of faster and low-cost computational resources.
Modeling efforts have tended to focus more on
the functional aspect of the plant such as simulat-
ing the biophysical environment of the canopy
and resource allocation, and a model usually
requires either “setting” the canopy architecture
or relatively simple attempts to create the struc-
ture (Norman and Campbell 1983; Grant 2001;
Dingkuhn et al., 2005; Evers et al., 2005; Renton
et al., 2005). Use of L-systems (Prusinkiewicz
1998) or the phyllochron in many of these models
has successfully created the plant architecture.
Models such as MODWht3, SHOOTGRO,
Sirius, and AFRCWHEAT2 might provide
further opportunities for simulating greater
canopy architectural detail.
Crop simulation modeling is beginning to
benefi t from the advent of object-oriented design
and programming languages such as C++ and
Java. Initial efforts have tended to view the plant
