JIA-2019-11

2524 ZHOU Tian-mei et al. Journal of Integrative Agriculture 2019, 18(11): 2521–2533 TEP in a certain growth stage is calculated using the daily TEP (DTEP). DTEP is calculated as the product of average relative RTE and PAR each hour. DTEP and TEP are shown in eqs. (2) and (3), respectively (Ni et al . 2009). (RTE i ×PAR i ×3 600/10 6 ) 24 i =1 DTEP= ∑ (2) (DTEP i ) TEP= n i =1 ∑ (3) where RTE i is RTE in hour i ; PAR i is the photosynthetically active radiation (PAR) in hour i (J m –2 d –1 ); DTEP is the daily accumulated product of thermal effectiveness and PAR (TEP) (MJ m –2 d –1 ); and TEP is the accumulated DTEP (MJ m –2 ). Calculation of seedling indexes Many seedling characteristics affect subsequent field performance, and the identification of a unique character that will always correspond to successful field performance is a fascinating but unrealistic thought. By comparative analysis of existing research results, we screened five common and reliable composite indexes (eqs. (4)–(8)) to develop a simulation model of seedling quality in tomato and cabbage based on TEP (Lu et al . 1984; Zhang et al . 1992; Mattsson 1996; Bai et al . 2014). Dry mass of shoot Dry mass of root R/T= (4) Days of seedling Dry mass of seedling G value= (5) Below are three healthy indexes: Seedling height Stem diameter HI1= ×Dry mass of seedling (6) Seedling height Stem diameter HI2= ×Fresh mass of seedling (7) )× Dry mass of shoot Dry mass of root Seedling height Stem diameter HI3=( + Dry mass of seedling (8) 2.5. Model development and validation TEP (eqs. (1)–(3)) and seedling quality indexes (eqs. (4)– (8)) were calculated based on the data collected and the corresponding equation. The seedling indexes at the first time points (on September 18 and 28, 2014 for tomato, and on April 23, 2015 for cabbage (Table 1)) and relevant TEP were then inserted into MATLAB (MathWorks Company, USA) to draw the corresponding curves and develop the models. The data collected from the second sowing (on September 28, 2014 and April 23, 2015 for tomato, and on May 12, 2015 for cabbage; Table 1) were used to validate these models, and the “Curve Fitting Toolbox” in MATLAB was then used to get the fitting function and test the goodness of fit. We analyzed the correlation accuracy between simulated and observed values using RMSE (root mean squared error). The RMSE was calculated by using the eq. (9). By analyzing the 1:1 line and correlation coefficient ( R 2 ) of simulated and observed values, we could identify the fitting degree and reliability of the model. (OBS i −SIM i ) n i =1 RMSE= n ∑ (9) where OBS i is the measured data, SIM i represents the predicted data, and n refers to the number of samples. The lower the RMSE value, the smaller the deviation between the simulated and measured values, and the more accurately the model can predict (Li et al . 2005). 3. Results 3.1. Environmental data and growth characteristics of tomato and cabbage seedlings Seeds of tomato and cabbage were sowed on at four different planting dates for model development and validation (Table 1). The average temperatures during four different plant growth periods slightly varied with 24.59°C for 2014- 09-18, 22.78°C for 2014-09-28, 24.89°C for 2015-04-23 and 26.24°C for 2015-05-12. This suggested that the average temperatures of our greenhouse were generally good for seedling growth of these two species (Table 2). However, elevated temperatures sometimes reached above 30°C, which exceeded the Tm for cabbage or tomato. Extremely elevated temperatures above 40°C were also observed in a narrow time window of approximately 2-h spanning from 10:00 to 13:00 on sunny days (Fig. 1-A). This may be caused by direct exposure of the data logger to sunlight. Compared to high temperatures, low temperatures did not exceed the Tb for either crop (Fig. 1-A). The maximum illumination for each period were 952.15, 897.66, 1 331.9 and 1 168.13 (J m –2 h –1 ), while average daily illumination ranged from 10.25 to 207.98 (J m –2 h –1 ) (Fig. 1-B). The DTEP ranged from 0.75 to 5.52 (MJ m –2 d –1 ) Table 2 Three fundamental temperature points for growth and development in tomato and cabbage seedlings 1) Crop Time Tb (°C) Tob – Tou (°C) Tm (°C) Tomato Day 5 20–25 35 Night 5 10–15 35 Cabbage Day –5 15–25 30 Night –5 8–12 30 1) Tb is the lower development temperature threshold; Tob is the lower optimum temperature threshold; Tou is the upper optimum temperature threshold; Tm is the higher development temperature threshold (Zhang 2004a, b; Cai et al. 2008; Fang 2008; Yu et al. 2009, 2013; Li 2011; Liu 2011).