Carbon Allocation to Leaves and Its Controlling Factors and Impacts on Gross Primary Productivity in Forest Ecosystems of Northeast China
[ad_1]
3.2. Analysis of Driving Factors of Carbon Allocation to Leaves
As shown in Figure 2e, leaf development is completed in July, and hence, we comprehensively analyzed the importance of TEM, PRE, SR, and SOS to carbon allocation to leaves (ΔLAI) from April to June (Figure 6). There were significant differences in the main influencing factors in different forest types. Within the GUP, all influencing factors had the greatest impact on carbon allocation to leaves in DBF, followed by MF and DNF. For DBF, the impacts of SOS and SR were higher than those of other factors. In MF, the impact of SOS was significant, and those of other factors were low. In DNF, except for the low importance of SR, the other factors were similar (Figure 6a). In April, SOS had the highest impact on all forestlands. For DBF and MF, the main influencing factors of ΔLAI were SOS and TEM, followed by SR, whereas PRE was the lowest. All factors had a greater impact on DBF than on MF. For DNF, except for SOS, the importance of the other factors was similar. It is worth noting that the impact of PRE on DNF was greater than that of other forestlands, whereas the impact of TEM was lower than that of other forestlands, and the impacts of SOS and SR were intermediate (Figure 6b). In May, the pattern remained almost the same as that observed in April. The impact of SOS remained high only in DBF, and its impact on other forestlands was significantly lower than that in April. The impact of TEM was lower than that in April. In DNF, except for SOS, the influence of the other three factors increased compared to April, and PRE exceeded that of TEM. This was similar to that in April in the MF (Figure 6c). SOS has the most significant lag effect on ΔLAI in June. Compared to May, the impact of all factors was significantly reduced in June. Only SOS in MF had a greater impact in June than in May, and its importance was higher than those of other climatic factors in the same time period. This phenomenon was most evident in MF, followed by DBF, with DNF having the lowest value. Other factors also confirmed this pattern (Figure 6d).
By constructing structural equations, we further explored the direct and indirect effects of the quantified TEM, PRE, SR, and SOS on carbon allocation to leaves (ΔLAI) (Figure 7). Owing to multicollinearity in some indicators, this study only selected representative indicators to construct the overall fit of the SEM to the standard. Obvious differences in the driving factors of ΔLAI of different forest types in different time periods were observed. The indirect effect coefficient of the meteorological factors on ΔLAI through SOS was higher than that of the direct effect coefficient.
In GUP, all factors have similar driving mechanisms for ΔLAI in DBF and DNF, with positive path coefficients for TEM and SOS, and negative effects on PRE and SR. However, this differs in DBF, wherein TEM has the highest positive path coefficient (0.359), and in DNF, wherein SOS has the highest coefficient (0.315). Among the negative effect coefficients, both DBF and DNF have the highest SR, with values of 0.302 and 0.476, respectively. In MF, except for SR, all are positive path coefficients, with SOS being the maximum value (0.296).
In April, SOS had high path coefficients for all forest types, and in descending order as follows: DBF (−1.191) > DNF (−1.186) > MF (−0.815). Specifically, the factors with the highest positive and negative path coefficients in DBF are SR (0.866) and TEM (−1.213), respectively. In DNF and MF, SOS had the highest negative path coefficient, while PRE (0.206) and SR (0.704) had the highest positive path coefficient in the two woodlands, respectively.
In May, all direct path coefficients for ΔLAI are negative except SOS. The relationship between SOS and ΔLAI shows a pattern opposite to that in April and is the only positive direct driver, and in descending order as follows: DBF (0.672) > MF (0.542) > DNF (0.331). The highest negative path coefficients are TEM (−0.33) and SR (−0.44) in DBF and MF respectively. In DNF, SR has the highest negative path coefficient (−0.739).
In June, the path coefficients of SOS in all forest lands were positive, with DNF being the highest (0.782), followed by MF (0.586), and DBF (0.552). In DBF, the positive effect of SR exceeds that of SOS and becomes the highest positive driving force with a path coefficient of 0.71, and TEM has the highest negative effect coefficient (−0.845). The highest positive coefficient of ΔLAI in DNF and MF is SOS. The difference is that SR is the only negative path coefficient in DNF (−0.676), while TEM has the highest negative path coefficient in MF (−0.76).
3.3. Effects of Carbon Allocation to Leaves on Gross Primary Productivity
Carbon allocation to leaves (ΔLAI) at different growth stages (April to June) has varying degrees of impact on the GPP (Figure 8). Overall, increased carbon allocation to leaves during the GUP of the forest was most beneficial for GPP accumulation in September. There is a lag effect between ΔLAI and GPP in different months of the forest. The ΔLAI in April had the strongest positive correlation with GPP in May (Figure 8b, R = 0.801), indicating a lag time of 1 month. From June to October, this positive correlation gradually weakened, and even became negative. The ΔLAI in May showed a strong positive correlation with the GPP in all months except May. The promotion effect of the increase in ΔLAI on the GPP in June was significantly lower than that in May, and it had the strongest lag effect, with a lag time of two months, on the GPP in August (Figure 8m, R = 0.218).
The correlation of ΔLAI and GPP in the GUP for different forest types was similar; however, it varied for different months. For MF, the increase in ΔLAI in April has a positive effect on the GPP from April to June, but it has an inhibitory effect on the GPP in July and subsequent months. The ΔLAI in May is positively correlated with the GPP in July and subsequent months. The ΔLAI in June continues to have a positive lagging effect on the GPP. For DBF, although the impact of ΔLAI on the GPP in April and May was always positive, this effect weakened in June. For DNF, the increase in ΔLAI in April promoted GPP in the months before October. ΔLAI and GPP in May and June showed similar positive correlations.
[ad_2]
