SALUS MODEL

Dr. Joe T. Ritchie⁽¹⁾

Overview

The SALUS (System Approach to Land Use Sustainability) program is designed to model continuous crop, soil, water and nutrient conditions under different management strategies for multiple years (Fig. 1.). These strategies may have various crop rotations, planting dates, plant populations, irrigation and fertilizer applications, and tillage regimes. The program will simulate plant growth and soil conditions every day (during growing seasons and fallow periods) for any time period when weather sequences are available. For any simulation run, a number of different management strategies can be run simultaneously. By running the different strategies at the same time we can compare this effect on crops and soil under the same weather sequences. This also provides a framework whereby the interaction between different areas under different management practices can be easily compared.

Every day, and for each management strategy being run, all major components of the crop-soil-water model are executed. These components are management practices, water balance, soil organic matter, nitrogen and phosphorous dynamics, heat balance, plant growth and plant development. The water balance considers surface runoff, infiltration, surface evaporation, saturated and unsaturated soil water flow, drainage, root water uptake, soil evaporation and transpiration. The soil organic matter and nutrient model simulates organic matter decomposition, N mineralization and formation of ammonium and nitrate, N immobilization, gaseous N losses and three pools of phosphorous. The development and growth of plants considers the environmental conditions (particularly temperature and light) to calculate the potential rates of growth for the plant. This growth is then reduced based on water and nitrogen limitations.

Module Descriptions and Derivations

The biophysical model is composed of three main structural components: i) a set of crop growth modules; ii) a soil organic matter and nutrient cycling module and; iii) a soil water balance and temperature module.

The crop growth modules are derived from the CERES and IBSNAT family of crop production models that were originally developed for single year, monoculture simulations. The crop growth algorithms from these were extracted and restructured into crop growth modules that are linked to the soil water, nutrient and management submodels. SALUS has been programmed in the "C" language to make the memory allocation more dynamic and program code more platform independent. Current operational crop growth modules include maize, wheat, barley, sorghum and millet. A generic grain legume module (e.g., soybean, dry beans) and alfalfa growth modules are being added to the system. The basic controls on production processes are the same as in the original models. Phasic development is controlled by environmental variables (e.g., degree days, photoperiod) governed by variety-specific genetic coefficients. Carbon assimilation and dry matter production are a function of potential rates (controlled by light interception and parameters defining the variety-specific growth potential) which are then reduced according to water and/or N limitations. The main external inputs required for the crop growth routines are the genetic (variety-specif) coefficients and daily solar radiation as a driving variable.

The soil organic matter (SOM) and nitrogen module is derived from the Century model (Parton et al. 1987), with a number of modifications incorporated. The model simulates organic matter and N mineralization/immobilization from three SOM pools (active, slow and passive) which vary in their turnover rates and characteristic C:N ratios. There are two crop residue/fresh organic matter pools (structural and metabolic), for representing recalcitrant and easily decomposable residues, based on residue lignin and N content. Decomposition and N mineralization rates for the different pools are influenced by soil temperature and moisture, soil texture and tillage intensity (as well as pool C:N ratio for N mineralization). The main external inputs the soil process module needs are soil texture, bulk density, horizon depths, total organic C and N, and initial mineral N content.

Several modifications were made to adapt the model for use with daily time-step crop growth routines. The original Century model operates on a monthly time step and therefore, rate constants were recalibrated to correct for the difference in integration interval (from monthly to daily). A surface active SOM pool associated with the surface residue pools was added to better represent conservation tillage systems and perennial crops. Soil organic matter and litter pools were also added for up to 10 soil layers (vs. only a single top soil layer in Century). The soil moisture control function for decomposition was replaced to make decomposition a function of percent water-filled pore space (Linn and Doran 1984). Separate ammonium and nitrate pools were represented with nitrification rate calculations based on Johnsson et al. (1987). An algorithm was developed which determines the initial fraction of total organic matter C and N in each of the three SOM pools (for model initialization), as a function of soil texture, type of original native vegetation and time under cultivation, based on a steady-state analytical solution of the decomposition equations (Paustian 1992).

The soil phosphorous (P) model incorporates inorganic and organic phosphorous dynamics. Inorganic P is divided into three pools i) labile; ii) active; and iii) stable. The plant absorbs P from the labile pool defined as resin-extractable P. The labile P estimates rapid equilibrium with an active mineral P pool following the addition of P or uptake of P. Stable P is the crystalline, inorganic P form that maintains a slow equilibrium with active P. The organic P is simulated in a similar way to the nitrogen pool where C:P ratios are used to control decomposition rates. Organic phosphorous details are the least tested and least understood compared to other components in SALUS.

The soil water balance module is based on that used in the CERES models but incorporates a major revision in calculating infiltration and runoff. In SALUS, a time-to-ponding (TP) concept is used to replace the previous runoff and infiltration calculations which were based on SCS runoff curve numbers. The runoff curve number approach was found to be inadequate in representing variation in infiltration characteristics associated with differences in tillage and residue management. Briefly, time-to-ponding curves relate rainfall intensity to infiltration rate and define the point at which cumulative rainfall intensity exceeds the infiltration capacity of the soil (White et al. 1989, Chou 1990), at which time water ponding in micro-depressions in the soil surface occurs. After ponding begins, infiltration is equal to the amount predicted by the TP curve as long as rainfall rate exceeds the infiltration capacity. When rainfall rate becomes less than the infiltration capacity, rainfall plus surface ponded water are infiltrated until the amount ponded is depleted. The main management-influenced parameter controlling the TP curve is the saturated hydraulic conductivity at the soil surface, which is varied as a function of tillage, soil compaction and surface residue amounts (Dadoun 1993). The TP approach requires additional information regarding rainfall intensity. With the assumption that only daily rainfall is known, a relatively simple disaggregation function (Nicks and Lan 1989) is used to derive rainfall intensities. The disaggregation function is defined for relatively large regions based on data from meteorological stations in the region which have records of storm intensities.

The SALUS model does not explicitly include submodels to predict pest and disease outbreaks or the occurrence of extreme weather events (e.g., hail). We recognize that these factors can have a major impact on crop production and yield and the sustainability of a particular management system. However, the multitude of potential pest species and disease-causing organisms of major crop species in the U.S. precludes the inclusion of pest dynamics submodels explicitly within a general cropping systems model such as SALUS. Similarly, it is unrealistic to attempt to predict the occurrence of extreme weather events within the model structure. Instead, SALUS uses the concept of linkage points (Boote et al. 1983; Teng 1988), which enables it to interface with external models of pest dynamics or other information about the type and extent of pest incidence or other damage factors (e.g., integrated pest management models). A linkage point is a known, quantifiable effect that a pest has on a crop; thus, pests can be classified according to their potential effects on a crop (i.e., the type of linkage point). A list of linkage points is shown in Table 1.

Table 1. Linkage points through which pest species effect crop growth processes.

Linkage Point	Examples
Reduction in plant number	Damping-off fungi
Photosynthetic rate reduction	Some viral pathogens
Leaf senescence acceleration	Some leaf-spotting fungi
Shading	Weeds
Assimilate removal	Aphids, leafhoppers
Tissue consumption	Defoliating insects
Turgor reduction	Wilt pathogens
Metabolic diversion	Root lesion nematodes
Resource competition	Weeds
Translocation disruption	Neck blast syndrome in rice
Tissue disruption	Rust pathogens, root lesion nematode

SALUS incorporates these linkage points by making the relevant variables (e.g., plant number, leaf or root tissue biomass, potential photosynthesis rates, senescence rates, etc.) accessible to interface with a specific pest model. For example, information on densities and tissue consumption rates of defoliating insects can be linked to SALUS to reduce leaf biomass, which in turn affects photosynthesis, growth and allocation, N uptake, and other processes according to the internal dynamics of SALUS. Alternatively the model can be used to assess the impacts of varying levels of pest incidence -- based on previous observations or expert opinion -- to generate a range of responses for use in risk analysis. By employing the linkage point approach the model remains flexible and can incorporate a wide variety of pest effects (or other damage factors such as hail) for a specific application.

User Interface

SALUS provides a user interface consisting of two components: i) input and output file management and; ii) an interactive control system. Input and output files use the DSSAT (version 3) formats (IBSNAT 1994). These standard files allow the user to create data files with existing programs from DSSAT and the output files can be used with the graphics and analysis programs written by DSSAT. The input files contain information to initialize the model (e.g., physical and chemical properties, soil water contents and N concentrations) and supply driving variables (i.e., time series of weather and management practices). The output files contain summaries of important model variables such as crop yield, crop developmental stages, N uptake, nitrate leaching, water drainage, runoff, soil organic C and N levels, irrigation water used and fertilizer applied. All model state variables and important process rates are stored continuously and can be output in graphical form.

The interactive portion of the program allows the user to have flexible control over the use of input files and the amount and type of data that is in the output. Users can select management strategies they wish to use and compare them directly in simultaneous model runs. Users can also interactively change the management practices to examine the effects of a particular management variable. The interactive portion of the program is written in a machine-independent windowing system. The entire program has been written in the Fortran 90 programming language to allow efficient computation and easy portability of the program to different computer platforms. The use of Fortran 90 also provides dynamic memory allocation, allowing for increased flexibility in the number of simultaneous simulations that can be run.

1. Professor, Homer Nowlin Chair; Michigan State University; Department of Crop and Soil Sciences; A570 Plant and Soil Sciences Building; East Lansing, MI 48824