Estimation of six leaf traits of East Asian forest tree species by leaf spectroscopy and partial least square regression

Nakaji, Tatsuro; Oguma, Hiroyuki; Nakamura, Masahiro; Kachina, Panida; Asanok, Lamthai; Marod, Dokrak; Aiba, Masahiro; Kurokawa, Hiroko; Kosugi, Yukio; Kassim, Abdul Rahman; Hiura, Tsutom

doi:10.1016/j.rse.2019.111381

Cited by 22 publications

(20 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The best performing PLSR models based on mixed NACs (Table 2, Figure 7) exhibited slightly lower R 2 than other recent studies: LMA R 2 = 0.79 LMA (present study) compared to 0.90 in [97]; phenolics R 2 = 0.73 (present study) compared to 0.83 in [98]; chlorophyll R 2 = 0.71 (present study) compared to 0.89 and 0.94 in [92,100], respectively; and water R 2 = 0.60 (present study) compared to 0.89 in [97]. In some cases, a lower model performance in the present study probably resulted from the needle-like leaf form of Norway spruce, which may cause difficulties in measuring optical properties at the needle level, as discussed above.…”

Section: Needle Traits Plsr Modeling-nacs Effectcontrasting

confidence: 70%

“…We selected PLSR as the target regression method for modeling Norway spruce needle biophysical traits for following reasons: as a nonparametric approach, PLSR takes advantage of model training with the full spectrum of information, is designed to deal with spectral data multicollinearity, and includes a dimensionality reduction step [67,96]. Among the multitude of approaches in biophysical parameter retrieval, PLSR has recently been successfully applied in plenty of vegetation studies both on leaf level spectral data, e.g., [92,[97][98][99], canopy-level imaging spectroscopy, e.g., [72], and other studies mentioned in the review by the authors of [96].…”

Section: Needle Traits Plsr Modeling-nacs Effectmentioning

confidence: 99%

“…In some cases, a lower model performance in the present study probably resulted from the needle-like leaf form of Norway spruce, which may cause difficulties in measuring optical properties at the needle level, as discussed above. The cited studies for model performance comparison [97,98] were performed on broad leaved species-annual crops and poplar cuttings, and Asian broadleaved forest tree species, respectively.…”

Section: Needle Traits Plsr Modeling-nacs Effectmentioning

confidence: 99%

See 2 more Smart Citations

Foliage Biophysical Trait Prediction from Laboratory Spectra in Norway Spruce Is More Affected by Needle Age Than by Site Soil Conditions

et al. 2021

View full text Add to dashboard Cite

Scaling leaf-level optical signals to the canopy level is essential for airborne and satellite-based forest monitoring. In evergreen trees, biophysical and optical traits may change as foliage ages. This study aims to evaluate the effect of age in Norway spruce needle on biophysical trait-prediction based on laboratory leaf-level spectra. Mature Norway spruce trees were sampled at forest stands in ten headwater catchments with different soil properties. Foliage biophysical traits (pigments, phenolics, lignin, cellulose, leaf mass per area, water, and nitrogen content) were assessed for three needle-age classes. Complementary samples for needle reflectance and transmittance were measured using an integrating sphere. Partial least square regression (PLSR) models were constructed for predicting needle biophysical traits from reflectance—separating needle age classes and assessing all age classes together. The ten study sites differed in soil properties rather than in needle biophysical traits. Optical properties consistently varied among age classes; however, variation related to the soil conditions was less pronounced. The predictive power of PLSR models was needle-age dependent for all studied traits. The following traits were predicted with moderate accuracy: needle pigments, phenolics, leaf mass per area and water content. PLSR models always performed better if all needle age classes were included (rather than individual age classes separately). This also applied to needle-age independent traits (water and lignin). Thus, we recommend including not only current but also older needle traits as a ground truth for evergreen conifers with long needle lifespan.

show abstract

Section: Needle Traits Plsr Modeling-nacs Effectcontrasting

confidence: 70%

Section: Needle Traits Plsr Modeling-nacs Effectmentioning

confidence: 99%

Section: Needle Traits Plsr Modeling-nacs Effectmentioning

confidence: 99%

See 1 more Smart Citation

Foliage Biophysical Trait Prediction from Laboratory Spectra in Norway Spruce Is More Affected by Needle Age Than by Site Soil Conditions

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Our pressed-leaf models often performed as well as fresh- and ground-leaf models published here and elsewhere. For example, our models for LMA had an RMSE (0.0115 kg m -2 ) lower than many fresh-leaf models from the literature, including Serbin et al (2019; 0.015 kg m -2 ), Nakaji et al (2019; 0.015 kg m -2 ), and Streher et al (2020; 0.051 kg m -2 ). The ground-leaf models in Serbin et al (2014) had a validation RMSE of 1.4 and 2.4 for cellulose and lignin percentages, comparable to 1.52 and 2.30 for our pressed-leaf models (Table 1).…”

Section: Discussionmentioning

confidence: 51%

Reflectance spectroscopy allows rapid, accurate, and non-destructive estimates of functional traits from pressed leaves

Kothari

Beauchamp‐Rioux

Laliberté

et al. 2021

Preprint

View full text Add to dashboard Cite

More than ever, ecologists seek to employ herbarium collections as tools to estimate plant functional traits from the past or from places that are hard to sample. However, many functional trait measurements are destructive, which may preclude their use on valuable herbarium specimens. Reflectance spectroscopy is increasingly used to estimate traits rapidly from fresh or dried, ground leaves, as well as to differentiate and identify taxa. Here, we extend this body of work to pressed, intact leaves such as those in herbarium collections. Using a dataset with 619 plant samples belonging to 70 woody and herbaceous species, we used partial least-squares regression to build validated models linking pressed-leaf reflectance spectra to a broad suite of traits, including leaf mass per area (LMA), leaf dry matter content (LDMC), carbon (Cmass) and nitrogen (Nmass) concentrations, and carbon fractions such as cellulose and lignin. We compared the accuracy of these trait estimates to those from fresh- or ground-leaf spectra of the same samples. Our pressed-leaf models predicted these traits with moderate-to-high accuracy (R2 = 0.586-0.924; %RMSE = 5.7-11.7%). For estimating chemical traits, pressed-leaf models performed better than fresh-leaf models but slightly worse than ground-leaf models. Pressed-leaf models did not perform as well as fresh-leaf models for estimating LMA and LDMC, but outperformed ground-leaf models for LMA. Accuracy was no worse for pressed leaves that underwent discoloration in storage. Finally, on a subset of common species in the dataset, we used partial least-squares discriminant analysis to classify specimens to species with near-perfect accuracy from pressed-(>97%), and ground-leaf (>96%) spectra and slightly lower accuracy from fresh-leaf spectra (>89%). The success of trait estimation and species classification from pressed-leaf spectra may owe to the fact that they combine advantages of fresh and ground leaves: like fresh leaves, they retain some of the spectral expression of internal leaf structure; like ground leaves, they reveal minor absorption features of chemical constituents that would otherwise be masked by water. These results show that applying spectroscopy to pressed leaves is a promising way to estimate leaf functional traits non-destructively. Our study has far-reaching implications for capturing the wide range of functional, phenotypic, and taxonomic information in the world’s preserved plant collections.

show abstract

“…Despite the myriad opportunities for remote sensing scientists to explore R&R issues, very few formal efforts have been documented. One reason is likely because remote sensing scientists often work with large datasets and perform complex spectral and spatial manipulations [12][13][14][15][16], which makes R&R difficult if processing code is not made available. Until recently, many scientific publications did not require code to be submitted as part of the manuscript review process, although this is changing.…”

Section: Introductionmentioning

confidence: 99%

Nutrient Prediction for Tef (Eragrostis tef) Plant and Grain with Hyperspectral Data and Partial Least Squares Regression: Replicating Methods and Results across Environments

2020

View full text Add to dashboard Cite

Achieving reproducibility and replication (R&R) of scientific results is tantamount for science to progress, and it is also necessary for ensuring the self-correcting mechanism of the scientific method. Topics of R&R have sailed to the forefront of research agenda in many fields recently but have received less attention in remote sensing in general and specifically for studies utilizing hyperspectral data. Given the extremely local environments in which many hyperspectral studies are conducted (e.g., agricultural field plots), purposeful attention to the repeatability of findings across study locales can help ensure methods are generalizable. This study undertakes an investigation of the nutrient content of tef (Eragrostis tef), an understudied plant that is growing in importance worldwide due to both food and forage benefits, but does so within the context of the replicability of methods and findings across two study sites situated in different international and environmental contexts. The aims are to (1) determine whether calcium, magnesium, and protein of both the plant and grain can be predicted using hyperspectral data with partial least squares (PLS) regression with waveband selection, and (2) compare the replicability of models across differing environments. Results suggest the method can produce high nutrient prediction accuracy for both the plant and grain in individual environments, but selection of wavebands for nutrient prediction was not comparable across study areas. The findings suggest that the method must be calibrated in each location, thereby reducing the potential to extrapolate methods to different areas. Our findings highlight the need for greater attention to methods and results replication in remote sensing, specifically hyperspectral analyses, in order for scientific findings to be repeatable beyond the plot level.

show abstract

Estimation of six leaf traits of East Asian forest tree species by leaf spectroscopy and partial least square regression

Cited by 22 publications

References 70 publications

Foliage Biophysical Trait Prediction from Laboratory Spectra in Norway Spruce Is More Affected by Needle Age Than by Site Soil Conditions

Foliage Biophysical Trait Prediction from Laboratory Spectra in Norway Spruce Is More Affected by Needle Age Than by Site Soil Conditions

Reflectance spectroscopy allows rapid, accurate, and non-destructive estimates of functional traits from pressed leaves

Nutrient Prediction for Tef (Eragrostis tef) Plant and Grain with Hyperspectral Data and Partial Least Squares Regression: Replicating Methods and Results across Environments

Contact Info

Product

Resources

About