Modeling seedling emergence of dallisgrass and bahiagrass forage cultivars with an empirical environmental variable

ESPERANZA PABLO; BATLLA DIEGO; GLISON NICOLAS; VIEGA L

Conferencia; International Forage & Turf Breeding Conference; 2019

In temperate and sub-tropical regions, perennial warm-season grasses adoption is wanted for forage production systems, which are adapted to local climate. These species has hindered their adoption because low seedling emergence, which depends on temperature and soil moist that affects seed germination and dormancy. Thus, selection of sowing date is crucial, but in regions with high inter-annual environmental variability, the suggested date may be misleading. The hydrothermal time model (HTT) has been used to predict seedling emergence in warm-season grasses and several other species. This model has strong physiological basis, which associates seed germination and dormancy with mean temperature (Tm) and water potential (Ψ). HTT model can be predictive, but needs lot of information from germination in controlled environments. However, it can be applied in empirical approaches with field data, where the incidence of environmental factors on germination behaviour can be deduced for species without previous information. To know the best environment to obtain higher emergence, five experiments were carried out in 2013 and 2014 in three locations of SE South America (Buenos Aires, Montevideo y Salto). In each, a split-split-plot design with three replications, two Paspalum cultivars (P. dilatatum cv. Chirú and P. notatum cv. INIA Sepé) were sown within four sowing dates (two Falls and two Springs), and within irrigated and non-irrigated main plots. For each minor plot, frequent count of emergence was done until no other seedling was registered; then, final proportion of emergence (pEm) was annotated. In previous analysis, pEm showed a linear trend with mean temperatures lower than 20°C. With this background, we propose the Average Weighted Thermal Time (AWTT), an empirical variable inspired in HTT model, which would linearly adjust with pEm. The AWTT is the average thermal time sum with a weighted coefficient, which takes a value in accordance with daily Tm and Ψ in reference to two temperature and two water potential thresholds. The thresholds and coefficient values in each rank were iterated to reach the best adjustment.First, temperature thresholds and corresponding coefficient values were iterated to adjust with irrigated pEm, where high Ψ was assumed. The best regression for Chirú (r2 = 0.47) was achieved with lower temperature thresholds than INIA Sepé (r2 = 0.46), and low coefficient in high temperatures for both cultivars. Then, the water potential thresholds and coefficient values from low Ψ rank were iterated to adjust with non-irrigated pEm. The best regressions for both genotypes was obtained with the same water potential thresholds (-0.15 and -3.0 MPa, respectively) and adjusted coefficient values, which the highlight was -1 when Tm is high and Ψ is between water potential thresholds. Lower adjustments was obtained for non-irrigated data (r2 = 0.29 and 0.35 for Chirú and INIA Sepé, respectively), but both slopes were significant (P < 0.0001). The AWTT has potential to empirically model the seedling emergence, because the interaction between temperature and water potential is embedded. The low coefficient values adjusted for high temperatures and some moist suggest that dormancy may be higher in such conditions.