Overview of Ecophysiological Models

A simulation using a basic model might start with a set of initial conditions specifying where the crop is grown, the initial status of water and nutrients in the soils, and the parameters needed to represent the physiological characteristics of the crop.

Fig. 4.1 Flow diagram for a hypothetical ecophysiological model with a daily time step

For an annual crop, the model loops through a series of subroutines that estimate plant or soil processes on an hourly or daily basis, outputting intermediate values at specified intervals (Fig. 4.1). In each cycle, the model checks whether the crop has reached maturity or a harvest data, in which case the yield and a diverse range of summary data may be output. Normally, the output of these models consists of a series of continuous curves representing different plant or environmental variables that change over time. Figure 4.2 presents examples of such output from the Cropping System Model (CSM)-CROPGRO model for a single crop of common bean grown near Cali, Colombia.

Mathematically, a model integrates a system of differential equations that describe rates that vary over time. The predicted (state) variables for the crop may include the dry mass of organs, leaf area, root length and vertical distribution in the soil, developmental progress, and soil water and nutrient concentrations of individual soil layers or horizons. In practice, the equations are far too complex for analytical solutions, so they are integrated numerically using time steps of a few seconds in very detailed models or hourly to daily, as found in most models.

Hundreds of ecophysiological models have been created. Many of these were developed by either a single scientist or small teams for a single research purpose. Most of these models can now only be found in the literature, although their algorithms may persist in newer models. There is no formal system of nomenclature, and in some cases, a single model has been modified independently by different groups, resulting in confusion over versions referred to in publications. Table 4.1 lists four families of models that have seen widespread use in climate change research.

The basic processes represented in ecophysiological models are described here mainly with reference to a hypothetical average plant. Most models actually report outputs on a land area basis, which corresponds to a community of identical "average" plants. A few models can simulate genetic mixtures, either of the same species or different species, including weeds.

-Tops

iffim

f V - -«. - —* -1

■■■Transpiration ■ - - Soil evap. — Potential ET

■■■Transpiration ■ - - Soil evap. — Potential ET

200 150 100 50 0

200 150 100 50 0

-Plant uptake

Fig. 4.2 Simulation of growth of common bean at Palmira, Colombia using the CSM-CROPGRO-Drybean model. (a) Change in dry matter for tops (total aboveground material), leaves and grains. Symbols indicate observed data from individual replicates. (b) Daily maximum and minimum temperatures. (c) Transpiration, evaporation from soil, and potential evapotranspiration. (d) Nitrogen taken up by plants or lost through leaching

20 40 60

Days after Planting

-Plant uptake

Fig. 4.2 Simulation of growth of common bean at Palmira, Colombia using the CSM-CROPGRO-Drybean model. (a) Change in dry matter for tops (total aboveground material), leaves and grains. Symbols indicate observed data from individual replicates. (b) Daily maximum and minimum temperatures. (c) Transpiration, evaporation from soil, and potential evapotranspiration. (d) Nitrogen taken up by plants or lost through leaching

Table 4.1 Examples of ecophysiological models used in climate change research

Name

Crop species

Description and references

Example applications

CropSyst

CSM-CERES

CSM-CROPGRO

EPIC

Barley, maize, sorghum, soybean, wheat, and others

Barley, maize, millet, sorghum, and wheat

Common bean, faba bean, peanut, soybean and other legumes, cotton, and others

Maize, millet, rice, sorghum, soybean, wheat, and others

Radiation use efficiency. Daily time step. Allows cropping sequences (Stockle et al. 2003)

Radiation use efficiency. Daily time step. Allows cropping sequences (Jones et al. 2003; Hoogenboom et al. 2004)

Farquhar-type photosynthesis calculated on an hourly basis with a hedge-row model for light interception. Considers growth and maintenance respiration. Allows cropping sequences (Jones et al. 2003; Hoogenboom et al. 2004) Radiation use efficiency. Daily time step. Allows cropping sequences and can model effects of tillage and soil erosion (Williams et al. 1989)

Rainfed wheat in south-eastern Australia

Bambara groundnut, peanut, maize, sorghum and soybean in Cameroon

(Tingem et al. 2008)

Maize and winter wheat in the United

States (Alexandrov and Hoogenboom 2000)

Wheat and maize in the Iberian Peninsula

(Minguez et al. 2007)

Maize production in Africa and Latin

America (Jones and Thornton 2003)

Soybean and peanut in the United Sates

(Alexandrov and Hoogenboom 2000)

Soybean in northeastern Austria

(Alexandrov et al. 2002)

Soybean in southern Quebec, Canada

(Brassard and Singh, 2008)

Maize, soybean and wheat in the Midwestern US, considering different spatial scales for climate and soil (Easterling et al. 2001) Maize, sorghum, millet, rice and cassava in Nigeria (Adejuwon 2006)

4.2.1 Development

Development includes the processes used by a plant to schedule important changes in growth such as the seedling emergence, formation of flowers, the onset of rapid grain growth, or the end of grain growth, which usually is considered to represent physiological maturity. This life history can be interpreted as a series of phases that are demarcated by stages, so the modeling approach is often termed "phasic development" (Ritchie and NeSmith 1991). Each phase is characterized by a duration that is expressed in physiological time, which is mathematically similar to thermal time, growing degree days or heat units but may include influences of photoperiod, vernalization or other processes. The duration represents the minimum time required for the plant to progress from one stage to another under optimum conditions. Each day (or hour), the plant is assumed to progress in time at a developmental rate (Dt), which is estimated from a potential rate (DP) and potential rate adjusting factors such as for temperature (T), photoperiod (P) and water deficits (W):

The rate adjusting factors usually vary from 0 to 1 in order to slow development below the maximum rate, but stresses such as water deficits may be used to accelerate development, resulting in the factor exceeding a value of 1. An alternate approach to phasic development that is especially common in modeling cereals is to use leaf number as the main indicator of developmental progress. While the terminology differs, the underlying physiology is similar (Jamieson et al. 2007).

The details of how phenology is modeled differ greatly with the biology of the crop species and decisions of the model developers concerning how to represent specific responses. For temperature, the decisions involve the selection of the temperature variables and specification of a curve that describes the assumed shape of a given physiological response. Common assumptions are that each crop has a "base temperature", below which it does not grow or develop and an "optimum temperature" that allows the maximum rate of growth or development. Temperatures may be observed or estimated from hourly values, daily averages, or averages adjusted or weighted in various manners. Of course, a crop does not respond to a daily value of the maximum or minimum temperatures; it is exposed to temperature (and all other environmental conditions) on a continuous basis. The models use simplified temperature data and associated equations. Soil temperature is often used to control germination, seedling emergence, and in cereals, early development of the shoot since the crown remains close to the soil surface.

Response curves vary from simple "broken stick" models to non-linear functions such as the beta function. Cardinal temperatures identify transition points in these responses. Besides the base and optimal temperatures described previously, models differ in how effects of supra-optimal temperatures are represented. The simplest approach is to assume that the maximum developmental rate is sustained above the designated optimum. Alternately, the rate may be assumed to decline to a lethal temperature, considered the maximum temperature for development, or the maximal

-Pre-anthesis dev.

--Leaf growth

- a- Photosynthesis

- - Vernalization

Fig. 4.3 Examples of temperature response functions assumed in the CSM-Cropsim-CERES wheat model. Curves are for pre-anthesis development, leaf growth, photosynthesis, vernalization, and grain growth. All responses are based on daily mean temperature

-Pre-anthesis dev.

--Leaf growth

- a- Photosynthesis

- - Vernalization

Fig. 4.3 Examples of temperature response functions assumed in the CSM-Cropsim-CERES wheat model. Curves are for pre-anthesis development, leaf growth, photosynthesis, vernalization, and grain growth. All responses are based on daily mean temperature rate may be sustained up to a second optimum above which the rate decreases to the lethal temperature (Fig. 4.3). This is an area that requires further research, especially as it relates to projected increases in temperature.

To non-specialists, the diverse approaches for modeling development may seem unscientific. The causes of the diversity are complex and reflect fundamental difficulties in simulating plant responses to weather conditions in general. Foremost is that, while air or soil temperatures can be measured accurately, a plant in a community experiences a complex, fluctuating temperature environment. Temperature sensing for a given process may reside in a specific tissue, such as the shoot apical meristem for vernalization (Sung and Amasino 2004). Models often assume that the air temperature reported from the nearest weather station approximates an average above-ground crop temperature, but there often are large temperature gradients within a canopy (Desjardins et al. 1978). Another challenge is that temperature responses involve circadian rhythms (the internal biological clocks) of the plant, and results from studies under constant temperatures or from simple constant day/night regimes likely have limited utility for quantifying temperature responses. A further problem is that response variables such as time of floral initiation or onset of flowering are usually scored visually on a sample of plants that vary in their developmental progress. These scores have error due to observation bias and sampling.

4.2.2 Growth

Growth is described through accumulation of dry matter in the main plant organs plus changes in a few additional traits such as leaf area and root length. An assimilate balance for a given time interval may be expressed as

where G is a growth increment per unit time, P is net photosynthesis for the plant or crop, and R and S are losses due to respiration and senescence (death of tissues related to stress or aging).

At the single leaf level, photosynthesis is simulated in response to light intensity, temperature and leaf external CO2 concentration. The Farquhar, von Caemmerer and Berry model for leaf photosynthesis (Farquhar et al. 1980) is often used for species with the C3 photosynthetic pathway, and the basic model is readily extended to account for the concentration of CO2 for the C4 pathway (von Caemmerer, 2000).

Temperature and [CO2] are obtained as external inputs or from other routines of the model. Estimating the light level, or more specifically the photon flux density for photosynthetically active radiation, requires describing how radiation is intercepted by the canopy. Many simple approaches assume that the irradiance (I) declines exponentially with the leaf area index (leaf area per unit land area, L),

where K is a dimensionless extinction coefficient that varies from 0 to 1. A canopy that predominantly contains horizontally oriented leaves has a higher value of K, and I declines more rapidly. Numerous complications are introduced when consideration is given to the diurnal cycle of radiation, effects of canopy shape and leaf angle distribution, diffuse and direct components of radiation, reflection from leaves, and other factors (Hay and Porter 2006). Pursuing these complications, however, may bring little benefit in accuracy where solar radiation data are unavailable and have to be estimated. We note especially that estimation of changes in solar radiation with climate change remain problematic for global circulation models (GCMs).

Plant tissues that are not actively photosynthesizing release CO2 through respiration just like any heterotrophic organism. This is because metabolic activity requires energy, whether it is to maintain existing tissue, construct new tissue, take up nutrients, or transport sugars. Models typically recognize two components to respiration. Growth respiration occurs in the construction of new tissues. Its rate varies primarily with the composition of the tissues being synthesized because the metabolic cost of synthesizing lipid, protein or lignin is much higher than for cellulose or starch (Penning de Vries et al. 1974). Maintenance respiration involves transport of nutrients, protein turnover, maintenance of ion gradients across membranes, and a host of other processes that are difficult to monitor individually. This component is usually assumed to increase with temperature and plant protein content, which is a good indicator of the overall metabolic activity of the plant and is proportional to total plant biomass.

Senescence is the process of controlled death of tissues. Leaf death is the most readily observed form, but stems, roots and fruits also senesce. Leaf senescence is largely associated with either shading or aging of early-formed leaves as the canopy develops or with mobilization of nitrogen during grain filling. Other drivers of senescence include water deficits, heat stress, flooding, and cold or frost damage. Typically, a moderate stress slows growth but if a threshold is exceeded, senescence occurs.

A widely used alternative to simulating dry matter growth through component processes is to evaluate G on a daily basis by assuming that net daily growth is the product of light intercepted by the canopy (I) and an integrative conversion factor called radiation use efficiency (RUE):

As with other estimators of photosynthesis, RUE can be modeled using a potential or reference value that varies with genotype, temperature, atmospheric CO2 or specific environmental stresses.

A comparison of the CSM-CERES models for maize, rice and sorghum illustrates the complexity underlying seemingly simple approaches (Fig. 4.4). Firstly, the temperature responses are based on a weighted average, TAVGD, calculated from the daily maximum (TMAX) and minimum (TMIN) as

The averaging implies that daytime temperatures have a greater effect on the processes underlying G than night temperatures. In comparing the respective rate modifiers for maize, rice and sorghum (Fig. 4.4), the responses for maize and rice are similar, not withstanding that C4 species such as maize and sorghum are generally considered more heat tolerant than C3 species like rice. The sorghum response agrees with the expectation that this species is more heat tolerant than maize. The curves, however, only partially define the response of G to temperature because the

Fig. 4.4 Examples of temperature response functions assumed in the CSM-CERES models for radiation use efficiency of maize, rice and sorghum. All responses are based on a weighted average temperature calculated as 0.75 of the daily maximum and 0.25 of the minimum

models assume that when temperature, nitrogen, water, or other stresses affect G, only the most severe stress is effective.

The [CO2] response for the CSM-CERES models is applied regardless of impacts of other environmental factors. The reference values of RUE used by the models are assumed to have been estimated for recent historic conditions with [CO2] of 350 ppm, so values above this level increase growth rate (Fig. 4.5). The expected greater responsiveness of species with C3 photosynthesis is in accordance with the basic expectations. However, the use of only two curves reflects the scarcity of reliable data on field-level responses to [CO2] rather than a consensus that there are no differences among the species (see Chapter 7).

Leaf area growth is modeled in order to estimate light interception. A common approach is to estimate an increment in leaf area from new leaf mass using the leaf area to mass ratio, also known as the specific leaf area (SLA). A reference value of SLA may be input as a cultivar specific parameter, and the actual SLA applied for a growth increment is varied with crop physiological age, temperature, solar radiation or other factors.

Simulating water or nutrient uptake requires information on the distribution of roots in the soil, including root length. Once a root mass increment is determined, root length growth varies with the tendency of the crop to be deep or shallow-rooted, the length to mass ratio of new roots, the current root length distribution, and soil physical constraints. Downward growth of roots is mainly temperature driven. Jones et al. (1991) reviewed these processes in more detail.

Fig. 4.5 Response to atmospheric [CO2] assumed in the CSM-CERES models for radiation use efficiency of C3 crops (barley, oats, rice and wheat) and C4 crops (maize and sorghum)

Atmospheric CO2 (ppm)

Fig. 4.5 Response to atmospheric [CO2] assumed in the CSM-CERES models for radiation use efficiency of C3 crops (barley, oats, rice and wheat) and C4 crops (maize and sorghum)

Renewable Energy 101

Renewable Energy 101

Renewable energy is energy that is generated from sunlight, rain, tides, geothermal heat and wind. These sources are naturally and constantly replenished, which is why they are deemed as renewable. The usage of renewable energy sources is very important when considering the sustainability of the existing energy usage of the world. While there is currently an abundance of non-renewable energy sources, such as nuclear fuels, these energy sources are depleting. In addition to being a non-renewable supply, the non-renewable energy sources release emissions into the air, which has an adverse effect on the environment.

Get My Free Ebook


Post a comment