Impact of preharvest conditions and post-harvest processing on quality traits of hard wheat (Triticum aestivum L.)

Abstract

Hard winter wheat is a key crop in Kansas, primarily used for leavened bread production. The milling and baking quality of this wheat are pivotal for its marketability. Various factors influencing wheat flour’s baking quality are: grain variety, agricultural and climatic conditions during growth and harvest, and the milling process. This study investigates the effects of preharvest practices and post-harvest processing on hard winter wheat’s baking quality. Specifically, the research aims to: (i) concurrently increase the wheat yield and baking quality through preharvest interventions, (ii) understand the mechanistic aspect of tempering and its impact on wheat kernel behavior, and (iii) assess the role of blending in flour functionality. In the first study, the research investigated the effects of intensive management practices on wheat yield and quality attributes through a factorial experiment conducted across six environments in Kansas. Five wheat genotypes were subjected to four management intensities, including farmer practice, high-input, high-input minus nitrogen, and high-input minus fungicide. Grain yield and 13 parameters representing grain, flour, dough, and baking qualities were evaluated. Results indicated that high -input practices significantly increased yield compared to farmer practice. Application of foliar fungicides improved milling characteristics, while nitrogen-intensive management enhanced the baking quality. Furthermore, improved management positively impacted yield, thousand kernel weight, test weight, grain protein, water absorption, and dough stability. The next stage of research investigated the impact of tempering, a critical step in wheat milling, on wheat kernel properties. A finite element model was developed to analyze moisture distribution within single wheat kernels, finding an effective diffusivity in the range of 1.63 – 3.81 × 10⁻¹¹ m²/s for hard red winter (HRW) kernels. The model indicated that the kernel center reached tempering moisture levels in 12h, contrary to the conventional 24h. To enhance predictability, future models could incorporate heterogeneous diffusion characteristics of endosperm, pericarp and germ. Additionally, the study evaluated the influence of tempering on the compressibility of HRW and hard red spring (HRS) kernels through a full factorial experiment. Uniaxial compression tests revealed significant impacts of tempering time on kernel viscoelastic properties. Elastic modulus ranged from 89.2 to 155.8 MPa for HRW and from 115.0 to 192.2 MPa for HRS kernels. Elastic work decreased from 8.8 to 5.4 N.mm, while plastic work increased to 1.9 N.mm with increasing moisture content. Notably, changes in compressibility followed first order kinetics across all moisture levels, with rate constants ranging from 0.002 to 0.0057 h⁻¹ for HRW. Further, the study analyzed the effects of tempering on the chemical and rheological properties of flour from blends composed of different proportions of hard wheat. Blends included B1-25:75 (HRS: HRW), B2-50:50, and B3-75:25. Results indicated significant influence of blending, tempering, and milling streams on protein, damaged starch, and particle characteristics. Protein content varied notably among break flour streams, while damaged starch content varied in the reduction streams across all blends. Higher proportions of HRS led to decreased pasting temperature of the dough. Principal component analysis revealed protein content as a key determinant in flour particle characteristics, water absorption (WA), and pasting properties, especially in blends with higher HRS proportion. Additionally, an artificial neural network (ANN) model was developed to predict the rheological (farinograph) properties of wheat dough based on flour properties. Multi-layer perceptron type feed-forward ANN models with increasing complexity were developed for each response variable, i.e., ANN-WA, ANN-DDT, and ANN-DS, with optimal models exhibiting the lowest mean squared error (MSE) and highest correlation coefficient (r). ANN- WA and ANN- DDT models demonstrated superior predictive performance when compared to ANN-DS. Feature importance analysis highlighted the significance of flour protein content in the model’s prediction. This study demonstrates the potential of data-driven ANN models in predicting the rheological characteristics of dough. Thus, the comprehensive investigation provides valuable insights into various aspects of hard winter wheat production and processing, aiming to enhance both yield and quality attributes. Furthermore, these findings offer valuable insights into the interplay between tempering conditions and kernel properties, facilitating the optimization of tempering processes for enhanced efficiency. The analysis of flour blends and development of predictive models for dough rheological properties further contribute to our understanding of wheat processing and its impact on product quality.