In the regions where crops were mostly produced by smallholder farmers, the analysis of yield gap is difficult due to diverse cultivars, crop managements and yield levels. In order to find an effective method that can reasonably verify the yield gap and the limiting cultivation factors in narrowing yield gaps in areas that are dominanted by smallholder farmers, we worked out a method consisting five progressive procedures as follows: questionnaire investigation of farmer cultivation regime, identification of yield levels and yield gaps, generalization of key cultivation measurements, reconstruction of representative maize populations, and process-based analysis of yield gap. A case study was carried out in Jiangsu Province, China, in which maize is mostly produced by smallholder farmers. A questionnaire investigation of 1 023 smallholder farmers was carried out firstly, then the frequency distribution of maize yield was simulated by an normal distribution function, and then the covering range and average value of the basic yield, farmer yield and high-yield farmer yield levels were calculated out from the equation. Hereby, the yield gaps 1, 2 and 3 were calculated along with the record highest yield from literature and experts, which were 2 564, 2 346 and 2 073 kg ha^–1, respectively. Moreover, with the covering range of each yield level, the suveyed farmers belonging to each yield level were grouped together and then their major cultivation measures were traced and generalized. With the generalized cultivation measures, representative maize populations of the four yield levels were reconstructed, and thereby clarifing lots of characters of the populations or single plant of each population with process-based analysis of the reconstructed populations. In this case, the main factors causing the yield gap were plant density, fertilizer application rate, logging caused by hurricane, and damages caused by pests. The case study primarily indicated that this five-step method is feasible and effective in yield gap study, especially in smallholder farmers dominant regions.

Two cotton (Gossypium hirsutum L.) cultivars, Kemian 1 (cool temperature-tolerant) and Sumian 15 (cool temperature-sensitive) were used to study the effects of cool temperature on carbohydrates, yield, and fiber quality in cotton bolls located at different fruiting positions (FP).  Cool temperatures were created using late planting and low light.  The experiment was conducted in 2010 and 2011 using two planting dates (OPD, the optimized planting date, 25 April; LPD, the late planting date, 10 June) and two shading levels of crop relative light rate (CRLR, 100 and 60%).  Compared with fruiting position 1 (FP1), cotton yield and yield components (fiber quality, leaf sucrose and starch content, and fiber cellulose) were all decreased on FP3 under all treatments.  Compared with OPD-CRLR 100%, other treatments (OPD-CRLR 60%, LPD-CRLR 100%, and LPD-CRLR 60%) had significantly decreased lint yield at both FPs of both cultivars, but especially at FP3 and in Sumian 15; this decrease was mainly caused by a large decline in boll number.  All fiber quality indices decreased under late planting and shading except fiber length at FP1 with OPD-CRLR 60%, and a greater reduction was observed at FP3 and in Sumian 15.  Sucrose content of the subtending leaf and fiber increased under LPD compared to OPD, whereas it decreased under CRLR 60% compared to CRLR 100%, which led to decreased fiber cellulose content.  Therefore, shading primarily decreased the “source” sucrose content in the subtending leaf whereas late planting diminished translocation of sucrose towards cotton fiber.  Notably, as planting date was delayed and light was decreased, more carbohydrates were distributed to leaf and bolls at FP1 than those at FP3, resulting in higher yield and better fiber quality at FP1, and a higher proportion of bolls and carbohydrates allocated at FP3 of Kemian 1 compared to that of Sumian 15.  In conclusion, cotton yield and fiber quality were reduced less at FP1 compared to those at FP3 under low temperature and low light conditions.  Thus, reduced cotton yield and fiber quality loss can be minimized by selecting low temperature tolerant cultivars under both low temperature and light conditions.

A two-year field experiment was conducted to illustrate the effects of sowing date on cottonseed properties at different fruiting-branch positions (FBPs).  Two cotton cultivars (Kemian 1 and Sumian 15) were sowed on 25 April, 25 May, and 10 June in 2010 and 2011, respectively.  The boll maturation period increased with the delaying of sowing date.  Normal sowing treatment (25 April) had higher seed weight, embryo weight, embryo oil content and protein content than late sowing treatments (25 May and 10 June).  The flowering date, seed weight, embryo weight, embryo oil and protein contents, and the dynamic changes of embryo oil and protein contents were altered by different FBPs.  A significant interaction of sowing date×FBP was observed on embryo weight, embryo oil content, embryo protein content and the dynamic changes of embryo oil and protein contents, but was not observed on seed weight.  Seed weight, embryo weight, embryo oil and protein content had significant positive correlations with the mean daily temperature (MDT), mean daily maximum temperature (MDT_max), mean daily minimum temperature (MDT_min), and mean daily solar radiation (MDSR), indicating that temperature and light resources were the main reasons for different sowing dates affecting the cottonseed properties at different FBPs.  Moreover, the difference in MDT was the main difference in climatic factors among different sowing dates.

Nitrogen (N) fertilizer experiments were conducted to investigate the optimal subtending leaf N concentration for fiber strength, and its relationship with activities of key enzymes (sucrose synthase and β-1,3-glucanase) and contents of key constituents (sucrose and β-1,3-glucan) involved in fiber strength development in the lower, middle and upper fruiting branches of two cotton cultivars (Kemian 1 and NuCOTN 33B). For each sampling day, we simulated changes in fiber strength, activity of sucrose synthase and β-1,3-glucanase and levels of sucrose and β-1,3-glucan in response to leaf N concentration using quadratic eqs.; the optimal subtending leaf N concentrations were deduced from the eqs. For the same fruiting branch, changes in the optimal leaf N concentration based on fiber development (DPA) could be simulated by power functions. From these functions, the average optimal subtending leaf N concentrations during fiber development for the cultivar, Kemian 1, were 2.84% in the lower fruiting branches, 3.15% in the middle fruiting branches and 3.04% in the upper fruiting branches. For the cultivar, NuCOTN 33B, the optimum concentrations were 3.04, 3.28 and 3.18% in the lower, middle and upper fruiting branches, respectively. This quantification may be used as a monitoring index for evaluating fiber strength and its related key enzymes and constituents during fiber formation at the lower, middle and upper fruiting branches.

Crop performance is determined by the combined effects of the genotype of the crop and the environmental conditions of the production system. This study was undertaken to develop a dynamic model for simulating environmental (temperature and solar radiation) and N supply effects on fiber fineness, maturity and micronaire. Three different experiments involving genotypes, sowing dates, and N fertilization rates were conducted to support model development and model evaluation. The growth and development duration of fiber fineness, maturity, and micronaire were scaled by using physiological development time of secondary wall synthesis (PDTSWSP), which was determined based on the constant ratio of SWSP/ BMP. PTP (product of relative thermal effectiveness (RTE) and photosynthetically active radiation (PAR), MJ m-2) and subtending leaf N content per unit area (NA, g m-2) and critical subtending leaf N content per unit area (CNA, g m-2) of cotton boll were calculated or simulated to evaluate effects of temperature and radiation, and N supply. Besides, the interactions among temperature, radiation and N supply were also explained by piecewise function. The overall performance of the model was calibrated and validated with independent data sets from three field experiments with two sowing dates, three or five flowering dates and three or four N fertilization rates for three subsequent years (2005, 2007, and 2009) at three ecological locations. The average RMSE and RE for fiber fineness, maturity, and micronaire predictions were 372 m g-1 and 5.0%, 0.11 m g-1 and 11.4%, 0.3 m g-1 and 12.3%, respectively, indicating a good fit between the simulated and observed data. It appears that the model can give a reliable prediction for fiber fineness, maturity and micronaire formation under various growing conditions.