Plant Phenology Simulation and Trigger Threshold Based on Total Climatic Production Factors—A Case Study of Stipa krylovii Phenology

Abstract

In response to the new concept of the impact of total climate production factors on plant phenology, this study verifies the feasibility of simulating plant phenology and triggering thresholds based on total climatic production factors by using the phenological and meteorological observation data of S. krylovii plants from 1985 to 2018 at the Xilinhot National Climate Observatory of China Meteorological Administration. The results indicate that the total climate production factors have a significant impact on plant phenological changes and can be effectively utilized for simulating phenology and determining triggering thresholds. The mutation of cumulative climate production potential based on total climate production factors can effectively indicate the green-up date and the wilting date of S. krylovii plants, and their triggering thresholds depend on the parameters of climate resource changes and the characteristics of plant biology, which are (0.085, −5.363) and (0.086, −27.620), respectively. The cumulative climate production potential based on total climate production factors can effectively indicate the heading date of S. krylovii plants, and its triggering thresholds also depend on the parameters of climate resource changes and the characteristics of plant biology, which is (394.632, −38,026.268). Furthermore, the results support the viewpoint that abrupt changes to the climate determine the beginning and ending of plant growth, while the accumulative climate resources determine the other phenological dates. This study provides new ideas for the study of plant phenology.

1. Introduction

Plant phenology is a mutually adaptive growth and development rhythm formed by the long-term adaptation of plants to seasonal changes in environmental conditions such as temperature, precipitation, and light [1], which not only reflects plant growth and development but also indicates climate change [2] and has been widely used to guide agricultural activities and disaster prevention and mitigation [3]. Meanwhile, plant phenology affects water-heat exchange and carbon cycling in ecosystems [4,5], and is an important parameter in land surface process models and plant productivity models [6].

It is found that plant spring phenology tends to be earlier and autumn phenology tends to be later in a warming context [7,8], and this phenomenon is more pronounced in the high latitudes of the Northern Hemisphere [9]. However, there is a clear spatial variability in the response of grassland plant phenology to climate change in Inner Mongolia. In the southern region, there is a trend of earlier green-up and later wilting, while in the central and northern regions, there is a trend of later green-up and earlier wilting [10]. However, the factors causing spatial variabilities in grassland plant phenology are still unclear. The meteorological factors that affect plant phenology include temperature, precipitation, light, air humidity, and carbon dioxide concentration. Among them, temperature is considered to be the most important environmental factor affecting plant phenology [11], and plants can only grow and develop in a certain temperature environment and require a certain cumulative temperature to complete their life cycle [12]. A water deficit limits the utilization of light and heat conditions by plants [13] and is considered a key factor in regulating vegetation activity in arid and semi-arid regions [14]. Light is the source of energy for photosynthesis in plants, and the organic matter produced by photosynthesis is the material basis for plant growth and development [15]. Additionally, the photoperiod has also been shown to be an important indicator of how light affects plant phenology [16]. Meanwhile, there are significant interactions between temperature, precipitation, and light, which collectively influence the changes in plant phenology [17,18]. Phenology models have evolved from statistical to mechanistic models, and have been widely used to simulate phenology and predict the impacts of future climate change in areas where phenology observations are lacking [19]. The original phenological models can be traced back to the thermal time model, which assumes that crop development is proportional to the accumulated thermal time. Considering the effects of dormancy processes and environmental factors during both endodormancy and ecodormancy on plant phenology, a two-phase model has been developed, including a sequential model and parallel model [19]. However, the two-phase model gives an unclear relationship between endodormancy and ecodormancy. Furthermore, there are models that have proposed that plant dormancy includes more than two phases, such as the deepening rest model and the four-phase model [19]. The models mentioned above assume that plant spring phenology is solely induced by air temperature. Some recent studies have also incorporated photoperiod effects into phenological models. The research on autumn phenology models is limited due to the inadequate and inaccurate ground observation data. Currently, the phenological research conducted through different models has been useful and insightful in identifying and demonstrating the causal relationship between various climatic factors and plant phenology [19]. However, the literature lacks a comprehensive model that can determine the combined effects of environmental factors on plant phenology [20]. Therefore, by including the comprehensive response of plant phenology to all environmental factors, the accuracy of the prediction model can be improved.

Research has shown that plant phenology is closely related to the dynamics of photosynthesis [21]. Plant photosynthesis is the result of the interaction between environmental factors and plant biological properties, reflecting the influence of all climate factors affecting plant production, which is referred to as the total climate production factors. Climate is the most important factor affecting plant growth and development, and it also serves as the foundation for the establishment of plant morphology and physiological and biochemical changes [22,23]. Climatic production potential refers to the highest biological or agricultural yield per unit area of land when other conditions (such as soil, nutrients, carbon dioxide) are optimal and the local climatic resources such as light, heat, and water are fully and rationally utilized [24]. Climate production potential not only reflects the influence of all climate factors (such as temperature, humidity, light) and their combined effects on plant production, but also ensures the uniformity of the influencing factors throughout the process of plant growth and the cyclical changes in their interaction with the environment. At the same time, climate production potential also reflects the combined effects of biological factors (such as leaf area), environmental factors, and their interactions, and can reflect the effects of extreme weather and climate events. Therefore, utilizing climate production potential as a driving factor for plant phenology changes can avoid the limitations of existing models and achieve an accurate simulation of phenology [25].

Chinese temperate grasslands are the third largest in the world [26], sensitive to climate change, and play an important role in the global carbon cycle [27]. The grassland of S. krylovii is one of the representative types of typical grasslands [28], which occupies an important position in livestock production [29] and has been significantly affected by the warm and dry climate [30]. The climatic mechanism of delaying the start and advancing the end of the growing season of S. krylovii and the effects of phenological changes on plant production of S. krylovii in a semi-arid region were revealed [31,32,33]. Using the long-term phenological and corresponding meteorological observation data from 1985 to 2018 in the grassland of S. krylovii, this study proposes new concepts of climate production potential, cumulative climate production potential (reflecting resource accumulation), the first-order derivative of cumulative climate production potential (reflecting the rate of resource change), and the second-order derivative of cumulative climate production potential (reflecting sudden resource change) based on the total climate production factors influencing plant phenological changes [25]. On this basis, this study intends to (1) verify the feasibility of simulating plant phenology and triggering thresholds based on the total climatic production factors, and (2) clarify the relationship between the main phenological periods of S. krylovii and total climatic production factors, as well as their triggering thresholds, to improve the understanding of the response of phenology to the combined effect of meteorological conditions and provide a basis for the development of phenological models.

2. Materials and Methods

2.1. Study Area and Data

The study data are from the Xilinhot National Climate Observatory in Xilinhot, Inner Mongolia, China (44°08′ 03′ N, 116°19′ 43′ E, 990 m a.s.l.). It is located in the middle of the Inner Mongolian Autonomous Region, which is a typical temperate semi-arid continental climate zone. The average annual temperature and average annual precipitation are 2.0 °C and 260 mm, respectively. Winter is cold and dry, summer is warm and humid, and solar radiation is strong. The soil type is chestnut soil. The dominant species of the ecosystem is S. krylovii, and the accompanying species include Leymus chinensis (Trin.) Tzvel.), Allium tenuissimum, Cleistogenes squarrosa (Trin.) Keng, Artemisia frigida Willd., Allium amsopodium Ledeb., Kochia prostrata (Linn.) Schrad., Artemisia scoparia Waldst. Et Kit., Heteropappus altaicus (Willd.) Novopokr. The phenology and meteorological data have been observed at the Xilinhot National Climate Observatory, Inner Mongolia, China Meteorological Administration since 1985. All meteorological data were examined and verified by the National Meteorological Information Center of the China Meteorological Administration. Plant phenology was obtained at the natural pasture observation site from the Xilinhot National Climate Observatory, Inner Mongolia, China Meteorological Administration. The fence area of the observation field is 100 × 100 m, which is divided into 4 plots of 50 × 50 m, and each plot is divided into 4 replicates for observation (Figure 1). The multi-spectral phenological observation camera (made in China) is located in the center of the phenology observation site. Phenological periods, which were recorded by local professional observers in accordance with standard observations [34], were observed every 2 d. And 10 plants with respectable growth and complete life history for 3 consecutive years were selected for observation. In this study, the phenological dataset of dominant species S. krylovii in typical grassland include the green-up date, the heading date, and the wilting date from 1985 to 2018. The green-up date is defined as the date when 50% of the individual plants in the plots have their elasticity restored and turned from yellow to green. The heading date is defined as the date when the aristae of 50% of plants are exposed from the top or side of the leaf sheath. The wilting date is defined as the date when two-thirds of the aboveground part of the first plant in the plot is withered and yellow. Each phenological stage is recorded. The phenological observation records are converted into the daily sequence from January 1st by using the method of Julian day. The meteorological dataset includes daily mean air temperature (°C), precipitation (mm), sunshine duration (h), average pressure (hpa), mean wind speed (m/s) at the height of 10 m, and relative humidity (%) from 1985 to 2018.

2.2. Method

2.2.1. Climate Tendency Rate

The climate tendency rate reflects the changing trend of meteorological elements in a region. X_i is a meteorological element value at the Xilinhot National Climate Observatory from 1985 to 2018, T_i is the corresponding chronological order, and the linear regression equation is as follows:

$$X_i=a+b{T}_i$$

(1)

where a is the regression constant, b is the regression coefficient, and b × 10 is the climate tendency rate. If the climate tendency rate is positive, it indicates that the element is showing an increasing trend; if it is negative, it indicates a decreasing trend.

2.2.2. Climate Production Potential

Climate production potential (LNPP) refers to the climate production per unit leaf area [25], which is obtained by dividing the climate production (YW) by the leaf area correction function f(L), thus making the climate production potential comparable among different phenological periods.

$$LNPP=\raisebox{1ex}{$YW$}\!\left/ \!\raisebox{-1ex}{$f(L)$}\right.$$

(2)

Climate production (YW) is calculated based on the step-by-step correction method recommended by the Food and Agriculture Organization of the United Nations (FAO) [25]. Firstly, photosynthetic production (YQ) is estimated based on plant physiological mechanisms and energy conversion. Secondly, light temperature production (YT) is evaluated by temperature corrections on YQ. Finally, climate production (YW) is obtained by moisture corrections on YT. YW can be calculated based on daily scale meteorological data during plant growth:

$$YW=Q\times f\left(Q\right)\times f\left(t\right)\times f\left(w\right)=YQ\times f\left(t\right)\times f\left(w\right)=YT\times f\left(w\right)$$

(3)

where YW is the climate production per unit area (kg·hm⁻²); Q is the total solar radiation during the plant growth period (MJ·m⁻²); f(Q) is photosynthetic efficiency coefficient (g·kJ⁻¹), f(t) is the temperature correction coefficient, f(w) is the moisture correction coefficient; YQ is the photosynthetic production potential per unit area (kg·hm⁻²); YT is the production potential of light and temperature per unit area (kg·hm⁻²).

YQ is the plant yield determined solely by solar radiation under assumed optimal conditions such as temperature, moisture, soil conditions, CO₂ concentration, and agricultural facilities, calculated as follows:

$$\begin{array}{l}YQ=f\left(Q\right)\times Q\\ =C\times S\times \epsilon \times \phi \times \left(1-\alpha \right)\times \left(1-\beta \right)\times \left(1-\rho \right)\times \left(1-\gamma \right)\\ \times \left(1-\omega \right)\times f\left(L\right)\times E\times {\displaystyle \sum Q_i\times {\left(1-\eta \right)}^{-1}\times {\left(1-\xi \right)}^{-1}}\times q^{-1}\end{array}$$

(4)

where Q is the total solar radiation during the plant growth season (∑Q_i) (MJ·m⁻²) and f(L) is the leaf area correction function. The other parameters listed in the equation are constants related to the biological characteristics of the plant itself (

Table 1. Explanation of each parameter in the equation.

Parameters	Definition	Values	Reference
C	Unit conversion factor (MJ·m−2/KJ·g−1→kg·hm−2)	10,000	[35]
$\mathrm{S}$	Ability of photosynthetic fixation of CO2	0.95	[35]
$\epsilon$	Ratio of photosynthetic active radiation to total radiation	0.5	[36,37]
$\phi$	Quantum efficiency of photosynthesis	0.224	[38]
$\alpha$	Community reflectance	0.1	[35]
$\beta$	Community leakage rate	0.04	[35]
$\rho$	Ratio of radiation interception by non-photosynthetic organs	0.1	[38]
$\gamma$	Light saturation limitation	0.03	[35]
$\omega$	Ratio of respiratory expenditure to photosynthetic products	0.3	[38]
$\eta$	Plant water content (hay)	0.1	[35]
$\xi$	Inorganic ash ratio	0.08	[38]
$s$	Economic coefficient	0.65	[35]
$q$	Unit dry matter heat content (KJ·g−1)	17.77	[38]

).

Daily solar radiation ($Q_i$) can be expressed as [39]

$$Q_i=Q_0(a+b\frac{n}{N})$$

(5)

where $Q_0$ is astronomical radiation (MJ·m^–2); n and N are the actual daily sunshine hours (h) and the maximum possible sunshine hours (h), respectively; and a and b are empirical coefficients, taken as 0.29 and 0.557, respectively [40].

$$Q_0=\left(T\times \frac{I_0}{\pi {\rho }^2}\right)\times \left(\omega \sin \phi \sin \delta +cos\phi cos\delta \sin \omega \right)$$

(6)

$$N=24\times \frac{\omega }{\pi }$$

(7)

where T is the time in days (=24 × 60 min), I₀ is the solar constant (0.082 MJ·m⁻²·min⁻¹), ω is the time angle (rad), φ is the geographic latitude (rad), δ is the solar declination (rad), ρ is the distance between the Sun and the Earth, and ρ² is the correction coefficient of eccentricity of Earth’s orbit.

$$\frac{1}{{\rho }^2}=1.00011+0.342221\cos \theta +0.00128\sin \theta +0.000719\cos 2\theta +0.00077\sin 2\theta$$

(8)

$$\omega ={\cos }^{-1}[-\tan \phi \tan \delta ]$$

(9)

$$\begin{array}{l}\delta =0.006918-0.399912\cos \theta +0.070257\sin \theta -0.006758\cos 2\theta \\ +0.000907\sin 2\theta -0.002697\cos 3\theta +0.000148\sin 3\theta \end{array}$$

(10)

$$\theta =\raisebox{1ex}{$2\pi (d_n-1)$}\!\left/ \!\raisebox{-1ex}{$365$}\right.$$

(11)

where θ is the solar angle (rad) and d_n is the chronological order, with January 1st as 1, January 2nd as 2, and so on.

Light temperature production (YT) is the plant yield determined solely by solar radiation and air temperature under assumed optimal conditions such as moisture, soil conditions, CO₂ concentration, and agricultural facilities, calculated as follows:

$$YT=YQ\times f\left(t\right)$$

(12)

$$f\left(t\right)=\{\begin{array}{l}0, t<t_{\min }, t>t_{\max }\\ \frac{t-t_{\min }}{t_s-t_{\min }}, t\le t<t_s\\ \frac{t_{\max }-t}{t_{\max }-t_s}, t_s\le t<t_{\max }\end{array}$$

(13)

where $f(t)$ is the temperature correction coefficient, 0 ≤ f(t) ≤ 1; t is the daily average temperature (°C); and t_min, t_s, t_max are the lower-limit temperature, optimal temperature, and upper-limit temperature (°C) of the growth period, where 0 °C, 20 °C, and 35 °C are taken in this study, respectively [35].

Climate production (YW) is the plant yield determined solely by solar radiation, air temperature, and moisture under assumed optimal conditions such as soil conditions, CO₂ concentration, and agricultural facilities, calculated as follows:

$$YW=YT\times f\left(w\right)$$

(14)

$$f\left(w\right)=\{\begin{array}{l}1, VPD<VP{D}_{\min }\\ \frac{VP{D}_{\max }-VPD}{VP{D}_{\max }-VP{D}_{\min }}, VP{D}_{\min }\le VPD\le VP{D}_{\max }\\ 0, VPD\le VP{D}_{\max }\end{array}$$

(15)

$$VIP=0.6108\exp [\frac{17.27t}{t+237.3}](1-RH)$$

(16)

where VPD is the water vapor pressure deficit (Pa); t the daily average temperature (°C); RH is the average relative humidity (%); and VPD_max and VPD_min are constants determined by vegetation type, where VPD_max and VPD_min are taken as 3100 Pa and 650 Pa, respectively [41].

To achieve the prediction of plant phenology, it is necessary to calculate LNPP for each day starting from January 1st and calculate the cumulative LNPP to reflect the climate resource status, i.e., the cumulative climate production potential. The cumulative climate production potential is a function of time and conforms to the logistic curve. The first derivative and second derivative can be obtained according to the cumulative climate production potential, which, respectively, reflects the change speed and acceleration (abrupt change) of climate resources. The mutation (second derivative) of the cumulative climate production potential can be used to indicate the beginning and the ending of plant growth, while the cumulative climate production potential or its change speed (first derivative) can be used to indicate the other phenological dates between the beginning and end of plant growth [25].

2.2.3. Logistic Curve

The accumulation of daily climate production potential can reflect the degree to which plants can utilize climate resources. Thus, the relationship between plant phenology and climate production potential can be analyzed. To quantitatively describe the accumulation of climate production potential over time, the logistic curve is used to simulate it. The trend equation of accumulative climate production potential based on the logistic curves is calculated as follows:

$$\frac{dY}{dt}=rN(\frac{K-Y}{Y})$$

(17)

The integral equation can be expressed as

$$Y=\frac{K}{1+a{e}^{-rt}}$$

(18)

where dY/dt is the instantaneous growth rate, $K$ is the theoretical upper limit of climate production potential, $t$ is time (d), Y is the cumulative value of climate production potential from January 1st, a is the coefficient, and r is the intrinsic growth rate and is the maximum instantaneous growth rate under specific conditions.

2.3. Data Processing

Microsoft Excel 2019 (Redmond, WA, USA) is used to collate and summarize the data, Matlab 2018a software (Mathworks, Natick, MA, USA) is used to fit the logistic curve and calculate the change rate (the first derivative of the cumulative climate production potential) and mutation (the second derivative of the cumulative climate production potential) of the cumulative climate production potential, and SPSS 21.0 (IBM, Armonk, NY, USA) is used to analyze the change trend of the main phenological dates of S. krylovii and their relationships with the cumulative climate production potential, its change rate, and mutation [21]. Origin 2018a is used to plot the figures.

3. Results

3.1. Climate Change Trends

From 1985 to 2018, the annual average temperature at the Xilinhot National Climate Observatory in Inner Mongolia showed an increasing trend, while the annual precipitation and annual sunshine duration showed a decreasing trend (Figure 2). From 1985 to 2018, the annual average temperature in the region ranged from 1.3 to 4.8 °C, with an average of 3.3 °C, showing a significant upward trend (0.40 °C/10a, p < 0.01). During this period, there was a slight decrease in the average temperature from December to January of the following year, while there was an increasing trend in the average temperature from February to November. Specifically, the average temperature changes in March, July, August, and September were significant, with rates of 1.17 °C/10a (p < 0.01), 0.89 °C/10a (p < 0.05), 0.53 °C/10a (p < 0.05), and 0.55 °C/10a (p < 0.05), respectively. The annual precipitation showed a large interannual variation, ranging from 121.1 to 511.7 mm. The annual average precipitation was 272.4 mm, showing an insignificant decreasing trend (−8.24 mm/10a, p > 0.05). The main decrease in precipitation occurred in July and August (−9.73 mm/10a (p > 0.05) and −10.76 mm/10a (p > 0.05), respectively). The average annual sunshine duration ranged from 2720.8 to 3215.8 h, with an average of 2991.9 h, showing an insignificant downward trend (−32.10 h/10a, p > 0.05). Except for January, February, July, and August, the average monthly sunshine duration in the Xilinhot region showed a decreasing trend, especially in May and June with significant changes (−9.06 h/10a (p < 0.05) and −10.76 h/10a (p < 0.05), respectively).

3.2. Phenological Change Trends of S. krylovii Plant

From 1985 to 2018, the green-up of S. krylovii plants in the Xilinhot region of Inner Mongolia mainly occurred in mid- and late April. The earliest green-up date was on 2 April 1993, and the latest green-up was on 7 May 2009. The green-up date showed a significant delay trend, with an average delay of 5.4 d/10a (p < 0.05) (Figure 3). There were significant interannual differences in the heading dates, occurring from early July to the middle of September, with a difference of nearly three months between the earliest and latest heading dates. The heading date showed an insignificant trend of advancing, with an average of 5.4 d/10a (p > 0.05). After September, S. krylovii plants gradually entered the wilting period, mainly concentrated in late September and early October, with the latest wilting date in the middle of October (1987, 1998, and 2000). In contrast, the wilting date showed a slight advance trend, with an average advance of 1.2 d/10a (p > 0.05). Overall, the significant delay in the green-up date and the advance in the wilting date have led to a significant reduction in the length of the growth season of S. krylovii plants in the Xlinhot region, with an average reduction of 6.3 d/10a (p < 0.05).

3.3. Relationship between the Main Phenology and Climate Production Potential

The cumulative climate production potential of S. krylovii plants followed a logistic curve over time (

Table 2. Results of logistic curve fitting the accumulative climate productivity potential.

Year	Logistic Curve	Determination Coefficients (R2)
1985	72,262.866/(1 + 548.623*e−0.033t)	0.998
1986	74,654.37/(1 + 824.629*e−0.035t)	0.999
1987	74,151.926/(1 + 514.25*e−0.033t)	0.998
1988	69,566.324/(1 + 531.187*e−0.032t)	0.998
1989	75,977.088/(1 + 359.302*e−0.031t)	0.998
1990	77,925.061/(1 + 433.181*e−0.031t)	0.999
1991	74,321.319/(1 + 684.216*e−0.034t)	0.999
1992	75,164.166/(1 + 639.736*e−0.033t)	0.999
1993	78,389.704/(1 + 657.65*e−0.033t)	0.998
1994	74,826.421/(1 + 330.43*e−0.031t)	0.998
1995	72,703.58/(1 + 693.954*e−0.034t)	0.998
1996	74,278.724/(1 + 581.662*e−0.033t)	0.998
1997	66,595.984/(1 + 357.171*e−0.031t)	0.998
1998	80,381.273/(1 + 247.249*e−0.029t)	0.998
1999	71,049.169/(1 + 321.777*e−0.031t)	0.997
2000	68,208.123/(1 + 300.816*e−0.03t)	0.996
2001	70,380.664/(1 + 346.159*e−0.03t)	0.998
2002	71,069.31/(1 + 413.602*e−0.032t)	0.998
2003	72,848.292/(1 + 451.606*e−0.032t)	0.998
2004	74,846.167/(1 + 351.334*e−0.03t)	0.998
2005	73,515.12/(1 + 417.551*e−0.031t)	0.998
2006	71,360.61/(1 + 445.391*e−0.031t)	0.998
2007	68,455.821/(1 + 335.983*e−0.03t)	0.997
2008	71,035.352/(1 + 323.441*e−0.03t)	0.998
2009	73,293.543/(1 + 348.71*e−0.031t)	0.998
2010	67,186.357/(1 + 505.09*e−0.032t)	0.996
2011	70,177.565/(1 + 495.677*e−0.032t)	0.999
2012	74,063.771/(1 + 401.423*e−0.031t)	0.998
2013	71,732.187/(1 + 610.204*e−0.033t)	0.998
2014	76,911.24/(1 + 297.44*e−0.03t)	0.999
2015	77,103.121/(1 + 423.355*e−0.031t)	0.999
2016	65,493.882/(1 + 299.812*e−0.03t)	0.997
2017	68,765.746/(1 + 260.438*e−0.029t)	0.996
2018	73,684.145/(1 + 237.541*e−0.029t)	0.996

). The determination coefficients of the cumulative climate production potential of S. krylovii plants from 1985 to 2018 were all greater than 0.996.

3.3.1. Relationship between the Green-Up Date and Climate Production Potential

The relationship between the green-up date of S. krylovii plants and the cumulative climate production potential showed a significant positive correlation (Figure 4). The correlation coefficients (R²) were 0.667, 0.822, and 0.862 for the cumulative climate production potential and its first and second derivative, respectively. Among them, the relationship between the green-up date and the second derivative of cumulative climate production potential was best, indicating that the drastic changes in meteorological conditions were important reasons for triggering the green-up date of S. krylovii plants. Based on this relationship, it can be concluded that the triggering threshold of the green-up date depended on both resource change parameters and biological characteristic parameters, namely the slope (0.085) and intercept (−5.363) of the second derivative of cumulative climate production potential with diurnal variation.

3.3.2. Relationship between the Heading Date and Climate Production Potential

The heading date of S. krylovii plants in the Xilinhot region showed a significant positive correlation with the cumulative climate production potential (Figure 5). However, there was a significant negative correlation between the heading date and the first and second derivative of the cumulative climate production potential. The correlation between the heading date and the cumulative climate production potential (0.840) was significantly better than the first and second derivative of the cumulative climate production potential (0.754 and 0.607, respectively), reflecting that the heading date of S. krylovii plants mainly depended on the cumulative degree of climate resources. Based on the relationship between the heading date and the cumulative climate production potential, it can be concluded that the triggering threshold of the heading date depended on both resource change parameters and biological characteristic parameters, namely the slope (394.632) and intercept (−38,026.268) of the cumulative climate production potential with diurnal changes.

3.3.3. Relationship between the Wilting Date and Climate Production Potential

There was a significant positive correlation between the wilting date of S. krylovii plants and the cumulative climate production potential and its second derivative, but a significant negative correlation with the first derivative of cumulative climate production potential (Figure 6). The correlation coefficients between the wilting date and the cumulative climate production potential and its first and second derivative were 0.315, 0.917, and 0.915, respectively. The wilting date of S. krylovii plants showed a strong correlation with the first and second derivative of cumulative climate production potential. However, there was a significant negative correlation between the wilting date and the first derivative of the cumulative climate production potential, indicating a decrease in the utilization of climate resources rather than a complete halt. Considering the variability between the green-up phenology and the wilting phenology [25], the abrupt changes in cumulative climate production potential can better reflect the changes in plant withering and yellowing periods. Based on the relationship between the wilting date of S. krylovii plants and the second derivative of the cumulative climate production potential, it can be concluded that the triggering threshold of the wilting date also depended on both resource change parameters and biological characteristic parameters, namely the slope (0.086) and intercept (−27.620) of the second derivative of cumulative climate production potential with diurnal variation.

4. Discussion and Conclusions

Meteorological factors are important factors that affect plant growth and development, as well as the basis for plant morphogenesis and physiological and biochemical changes. The meteorological factors that affect the green-up date of grassland plants include temperature, sunshine duration [42], accumulated temperature [19,43], and moisture [44,45]. When the air temperature is below the threshold, the autumn phenology occurs [46,47]; when the photoperiod is shortened to the threshold that limits plant growth and development, it induces plant leaf senescence and enters a dormant state. An in situ simulation experiment of S. krylovii grassland also indicates that it is environmental factors, rather than plant productivity, that drive leaf senescence [48]. In this study, the pronounced relationship between the green-up date and wilting date of S. krylovii plants and the second derivative of cumulative climate production potential was found to be best and significant. This finding reconfirmed the strong relationship existing between meteorological conditions and the green-up to wilting cycles. However, as a novelty, it also highlighted that the plant phenology simulation and trigger threshold could be determined by the total climatic production factors. The heading date of S. krylovii plants had the best relationship with the cumulative climate production potential, indirectly proving the impact of rising summer temperature as a driver of climate production potential on the phenological periods between the beginning and ending phenology, reflecting the cumulative climate resource effects.

In this study, the long-term phenology and corresponding meteorological observation data of S. krylovii grassland from 1985 to 2018 were used to verify the indication of climate production potential based on total climate production factors on plant phenology, and the relationship between the main phenological stages of S. krylovii plants and climate production potential was clarified. The climate production potential not only reflects the comprehensive effect of environmental factors on plant production and their consistency throughout the process of plant growth, but also reflects the effects of biological factors, environmental factors, and their interactions, as well as extreme weather and climate events, which will be helpful for achieving accurate simulations of plant phenology. The main conclusions are as follows:

(1) The feasibility of simulating plant phenology and triggering thresholds based on total climatic production factors is verified. The climate production potential, which is based on these factors, serves as a good indicator for predicting the phenology of S. krylovii plants. The abrupt change in cumulative climate production potential reflects the drastic change in climate conditions, effectively indicating the green-up and wilting periods of S. krylovii plants. On the other hand, the cumulative climate production potential reflects the rate of change in climate resources and, to some extent, reflects resource utilization, which can effectively indicate the heading period of S. krylovii plants.

(2) The relationship between the main phenological periods of S. krylovii and total climatic production factors and their triggering thresholds is clarified. The mutation of cumulative climate production potential based on total climate production factors can effectively indicate the green-up date and the wilting date of S. krylovii plants, and their triggering thresholds depend on the parameters of climate resource changes and plant biology, which are (0.085, −5.363) and (0.086, −27.620), respectively. The cumulative climate production potential based on total climate production factors can effectively indicate the heading date of S. krylovii plants, and its triggering thresholds also depend on the parameters of climate resource changes and plant biological parameters, which is (394.632, −38,026.268), respectively.