The cold truth: low temperature exclusion biases accurate estimation of base temperature but has limited impact on optimum and ceiling temperatures

• Excluding low (<10 °C) data destabilizes base temperature (Tb) estimates. • Optimum (To) and ceiling (Tc) temperatures remain robust across scenarios. • Tb estimates shift by up to ±40 % without sub-optimal temperature data. • Warm-season crops show less bias than cool-season species in Tb estimation. • Standardized low-temperature protocols are critical for reliable germination models. Accurate estimation of cardinal temperatures is essential for modeling crop germination responses to climate variability, yet experimental design strongly influences parameter reliability. This study systematically evaluated how truncated incubation ranges—excluding low (<10 °C) or high (>30 °C) temperatures—affect cardinal temperature estimates. Germination datasets from ten crop species were analyzed under four experimental scenarios using three non-linear response models, with model validity assessed through AIC, BIC, R², and RMSE. Results demonstrated that excluding sub-optimal low temperatures critically compromised the estimation of base temperature (T b ), inflating or deflating values by up to 40 % relative to full-range estimates. In contrast, optimum (T o ) and ceiling (T c ) temperatures were stable across scenarios, with deviations consistently <5 %. The heightened sensitivity of T b reflects both mathematical imprecision when low data points are absent and physiological processes regulating germination onset. Warm-season crops with intrinsically higher T b values exhibited less bias than cool-season species, underscoring species-specific vulnerability to truncated ranges. These findings highlight the necessity of including low-temperature treatments in germination experiments, particularly for crops with suspected low T b values. By contrast, omission of supra-optimal (>30 °C) data introduces minimal error, confirming the robustness of T c estimation. We propose standardized low-temperature protocols and species-specific modeling approaches as critical prerequisites for reliable cardinal temperature determination. Such methodological rigor will improve the predictive power of crop models, refine cultivar screening, and strengthen breeding strategies for climate resilience.