Monitoring biochemical limitations to photosynthesis in N and P-limited radiata pine using plant functional traits quantified from hyperspectral imagery

Abstract

The prediction of carbon uptake by forests across fertility gradients requires accurate characterisation of how biochemical limitations to photosynthesis respond to variation in key elements such as nitrogen (N) and phosphorus (P). Over the last decade, proxies for chlorophyll and photosynthetic activity have been extracted from hyperspectral imagery and used to predict important photosynthetic variables such as the maximal rate of carboxylation (Vcmax) and electron transport (Jmax). However, little research has investigated the generality of these relationships within the nitrogen (N) and phosphorus (P) limiting phases, which are characterised by mass based foliage ratios of N:P ≤ 10 for N limitations and N:P > 10 for P limitations. Using measurements obtained from one year old Pinus radiata D. Don grown under a factorial range of N and P treatments this research examined relationships between photosynthetic capacity (Vcmax, Jmax) and measured N, P and chlorophyll (Chla+b). Using functional traits quantified from hyperspectral imagery we then examined the strength and generality of relationships between photosynthetic variables and Photochemical Reflectance Index (PRI), Sun-Induced Chlorophyll Fluorescence (SIF) and chlorophyll a + b derived by radiative transfer model inversion. There were significant (P < .001) and strong relationships between photosynthetic variables and both N (R2 = 0.82 for Vcmax; R2 = 0.87 for Jmax) and Chla+b (R2 = 0.85 for Vcmax; R2 = 0.86 for Jmax) within the N limiting phase that were weak (R2 < 0.02) and insignificant within the P limiting phase. Similarly, there were significant (P < .05) positive relationships between P and photosynthetic variables (R2 = 0.50 for Vcmax; R2 = 0.58 for Jmax) within the P limiting phase that were insignificant and weak (R2 < 0.33) within the N limiting phase. Predictions of photosynthetic variables using Chla+b estimated by model inversion were significant (P < .001), positive and strong (R2 = 0.64 for Vcmax; R2 = 0.63 for Jmax) within the N limiting phase but insignificant and weak (R2 < 0.05) within the P limiting phase. In contrast, both SIF and PRI exhibited moderate to strong positive correlations with photosynthetic variables within both the N and P limiting phases. These results suggest that quantified SIF and PRI from hyperspectral images may have greater generality in predicting biochemical limitations to photosynthesis than proxies for N and chlorophyll a + b, particularly under high foliage N content, when P is limiting.