Technical note: Improving modeling of coagulation, curd firming, and syneresis of sheep milk

The importance of milk coagulation properties for milk processing, cheese yield, and quality is widely recognized. The use of traditional coagulation traits presents several limitations for testing bovine milk and even more for sheep milk, due to its rapid coagulation and curd firming, and early syneresis of coagulum. The aim of this technical note is to test and improve model fitting for assessing coagulation, curd firming, and syneresis of sheep milk. Using milk samples from 87 Sarda ewes, we performed in duplicate lactodynamographic testing. On each of the 174 analyzed milk aliquots, using 180 observations from each aliquot (one every 15 s for 45 min after rennet addition), we compared 4 different curd firming models as a function of time (CFt, mm) using a nonlinear procedure. The most accurate and informative results were observed using a modified 4-parameter model, structured as follows: CFt = CFp x((1-e-kCF(t-RCTeq)))xe(kSRx(t-RCTeq)), where t is time, RCTeq (min) is the gelation time, CFp (mm) is the potential asymptotical CF at an infinite time, k(CF) (%/rnin) is the curd firming rate constant, and k(sR) (%/min) is the curd syneresis rate constant. To avoid nonconvergence and computational problems due to interrelations among the equation parameters, CFp was preliminarily defined as a function of maximum observed curd firmness (CFmax, mm) recorded during the analysis. For this model, all the modeling equations of individual sheep milk aliquots were converging, with a negligible standard error of the estimates (coefficient of determination >0.99 for all individual sample equations). Repeatability of the modeled parameters was acceptable, also in the presence of curd syneresis during the lactodynamographic analysis.