Simulation of perennial ryegrass quality traits using PaSim in a breeding context

Abstract

Forage quality is important for an efficient uptake and digestion by ruminants. Factors that may limit the animal's ability to reach production goals (high nutritional value of milk and meat) may include the forage's energy and protein content. Moreover, a balanced energy/protein feed can contribute to reduce greenhouse gas emissions from forage protein during digestion. Together with dry matter yield, several quality traits are used as selection criteria in breeding programs of perennial ryegrass (Lolium perenne L.). Though the different components of forage quality have a genetic basis, quality traits are strongly influenced by environmental conditions and their expression varies over the growing season. Consequently, the selection progress is often slow for quality traits because the appreciation of a candidate variety differs largely across years and locations.Modelling approaches can assist here. We refer to the vegetation module of the grassland model PaSim, which simulates different quality traits in response to growing conditions. We investigated whether PaSim can explain the variation in quality traits of candidate varieties of perennial ryegrass throughout the year, and whether model parameter values can be set to characterize each candidate variety.For that, we used a wide set of observations from 65 candidate varieties of the ILVO perennial ryegrass breeding program started in 2012. Observed data from five cutting events in 2013 include: dry matter yield, crude protein content, water soluble carbohydrate content, neutral detergent fiber content (NDF, i.e. cell wall content) and digestible neutral detergent fiber content (NDFD, i.e. cell wall digestibility). Dry matter yield data were also measured in 2014. Model parameter values were estimated for each candidate variety, based on observations and using the parameter optimization function of Package-FME–R Project.This allowed us assigning a set of parameter values to each candidate variety, which can now be evaluated in virtual experiments for a better appreciation of the candidate varieties.