Prediction of forest aboveground net primary production from high‐resolution vertical leaf‐area profiles

Abstract

Temperature and precipitation explain about half the variation in aboveground net primary production (ANPP) among tropical forest sites, but determinants of remaining variation are poorly understood. Here, we test the hypothesis that the amount of leaf area, and its vertical arrangement, predicts ANPP when other variables are held constant. Using measurements from airborne lidar in a lowland Neotropical rain forest, we quantify vertical leaf‐area profiles and develop models of ANPP driven by leaf area and other measurements of forest structure. Vertical leaf‐area profiles predict 38% of the variation among plots. This number is 4.5 times greater than models using total leaf area (disregarding vertical arrangement) and 2.1 times greater than models using canopy height alone. Furthermore, ANPP predictions from vertical leaf‐area profiles were less biased than alternate metrics. Variation in ANPP not attributable to temperature or precipitation can be predicted by the vertical distribution of leaf area in this system.

Defining a spectrum of integrative trait‐based vegetation canopy structural types

“…Our results indicate that CSTs can provide additional explanatory power beyond that of broad eco‐climatic domains and forest functional types in predicting ecosystem processes and functions. The importance of CSTs was especially apparent when assessed within domains, suggesting that variation in canopy physical structure could be an particularly important predictor of functioning at the landscape scale within regions (Cushman & Kellner ). Although individual canopy traits have previously been shown to be highly influential on ecosystem functions (Reich ; Atkins et al ; Jucker et al ), a focus on multivariate suites of canopy traits could help further elucidate fundamental ecological mechanisms underpinning ecosystem structure–function relationships and isolate the distinct role of physical structure (e.g.…”

Defining a spectrum of integrative trait‐based vegetation canopy structural types

“…Our results indicate that CSTs can provide additional explanatory power beyond that of broad eco‐climatic domains and forest functional types in predicting ecosystem processes and functions. The importance of CSTs was especially apparent when assessed within domains, suggesting that variation in canopy physical structure could be an particularly important predictor of functioning at the landscape scale within regions (Cushman & Kellner ). Although individual canopy traits have previously been shown to be highly influential on ecosystem functions (Reich ; Atkins et al ; Jucker et al ), a focus on multivariate suites of canopy traits could help further elucidate fundamental ecological mechanisms underpinning ecosystem structure–function relationships and isolate the distinct role of physical structure (e.g.…”

From canopy complementarity to asymmetric competition: The negative relationship between structural diversity and productivity during succession

“…The decision to eliminate a pathway was based on overall model fit and the pvalue for the path (Grace, 2006). Model fit was assessed using (Cushman & Kellner, 2019). The combined loading weight of a particular vertical layer indicates its effect on productivity (positive or negative).…”

“…In the final step, we chose a satisfactory model (Chi-square test of p > 0.05) with minimum Akaike information criterion by deleting non-significant pathways. Overall model fit was carried out with the 'piecewiseSEM' package(Lefcheck, 2016) in r. Non-normal variables were natural logarithm-transformed and standardized to unit variance.Our final aim, question (4), was to examine the relationships between vertical leaf area distribution and coarse woody productivity of each forest type to infer spatial complementarity or asymmetric competition between overstorey and understorey using a partial least squares regression model(Cushman & Kellner, 2019). Partial least squares regression model is a kind of simultaneous principal components analysis and regression.…”

Prediction of Biomass in Dry Tropical Forests: An Approach on the Importance of Total Height in the Development of Local and Pan-tropical Models

“……”

Defining a spectrum of integrative trait‐based vegetation canopy structural types

“…Our results indicate that CSTs can provide additional explanatory power beyond that of broad eco‐climatic domains and forest functional types in predicting ecosystem processes and functions. The importance of CSTs was especially apparent when assessed within domains, suggesting that variation in canopy physical structure could be an particularly important predictor of functioning at the landscape scale within regions (Cushman & Kellner ). Although individual canopy traits have previously been shown to be highly influential on ecosystem functions (Reich ; Atkins et al ; Jucker et al ), a focus on multivariate suites of canopy traits could help further elucidate fundamental ecological mechanisms underpinning ecosystem structure–function relationships and isolate the distinct role of physical structure (e.g.…”

From canopy complementarity to asymmetric competition: The negative relationship between structural diversity and productivity during succession

“…The decision to eliminate a pathway was based on overall model fit and the pvalue for the path (Grace, 2006). Model fit was assessed using (Cushman & Kellner, 2019). The combined loading weight of a particular vertical layer indicates its effect on productivity (positive or negative).…”

“…In the final step, we chose a satisfactory model (Chi-square test of p > 0.05) with minimum Akaike information criterion by deleting non-significant pathways. Overall model fit was carried out with the 'piecewiseSEM' package(Lefcheck, 2016) in r. Non-normal variables were natural logarithm-transformed and standardized to unit variance.Our final aim, question (4), was to examine the relationships between vertical leaf area distribution and coarse woody productivity of each forest type to infer spatial complementarity or asymmetric competition between overstorey and understorey using a partial least squares regression model(Cushman & Kellner, 2019). Partial least squares regression model is a kind of simultaneous principal components analysis and regression.…”

Prediction of Biomass in Dry Tropical Forests: An Approach on the Importance of Total Height in the Development of Local and Pan-tropical Models

“……”

Defining a spectrum of integrative trait‐based vegetation canopy structural types

“…Our results indicate that CSTs can provide additional explanatory power beyond that of broad eco‐climatic domains and forest functional types in predicting ecosystem processes and functions. The importance of CSTs was especially apparent when assessed within domains, suggesting that variation in canopy physical structure could be an particularly important predictor of functioning at the landscape scale within regions (Cushman & Kellner ). Although individual canopy traits have previously been shown to be highly influential on ecosystem functions (Reich ; Atkins et al ; Jucker et al ), a focus on multivariate suites of canopy traits could help further elucidate fundamental ecological mechanisms underpinning ecosystem structure–function relationships and isolate the distinct role of physical structure (e.g.…”

From canopy complementarity to asymmetric competition: The negative relationship between structural diversity and productivity during succession

“…The decision to eliminate a pathway was based on overall model fit and the pvalue for the path (Grace, 2006). Model fit was assessed using (Cushman & Kellner, 2019). The combined loading weight of a particular vertical layer indicates its effect on productivity (positive or negative).…”

“…In the final step, we chose a satisfactory model (Chi-square test of p > 0.05) with minimum Akaike information criterion by deleting non-significant pathways. Overall model fit was carried out with the 'piecewiseSEM' package(Lefcheck, 2016) in r. Non-normal variables were natural logarithm-transformed and standardized to unit variance.Our final aim, question (4), was to examine the relationships between vertical leaf area distribution and coarse woody productivity of each forest type to infer spatial complementarity or asymmetric competition between overstorey and understorey using a partial least squares regression model(Cushman & Kellner, 2019). Partial least squares regression model is a kind of simultaneous principal components analysis and regression.…”

Prediction of Biomass in Dry Tropical Forests: An Approach on the Importance of Total Height in the Development of Local and Pan-tropical Models

“……”