Optical traits perform equally well as directly‐measured functional traits in explaining the impact of an invasive plant on litter decomposition

Abstract

Functional traits can help elucidate and predict the impact of invasive plant species on ecosystem functioning. Yet, this approach requires comprehensive and labour‐intensive trait collection campaigns, covering intraspecific trait variation of both the invader and native species in the invaded community. One potential way to overcome these logistic constraints is using hyperspectral remote sensing technology to efficiently quantify functional trait values. Although such spectrally derived or ‘optical’ traits are known to closely link to directly measured functional traits, little research has explored how well these optical traits perform in assessing invader‐induced ecosystem impact. Here, we explored the trait‐mediated impact of the invasive Rosa rugosa on litter decomposition and evaluated whether optical traits perform equally well as directly measured traits in predicting litter decomposition variation. We collected data on species‐specific functional traits, leaf hyperspectral reflectance and standardized ‘tea bag index’ litter decomposition across 25 invaded and 25 uninvaded coastal grassland plots. The selected traits were all potentially related to litter decomposition and covered the leaf economics spectrum, additional leaf structural components and competitive ability. Optical traits were quantified through a combination of a physical radiative transfer model inversion and vegetation indices calculations. Invasion significantly increased the stabilization factor, i.e. the amount of resulting recalcitrant litter. Invader impact on litter decomposition could be entirely explained by changes it induced in the functional traits of the native community, rather than by the invader's traits itself. More specifically, the invader pushed the invaded community towards traits associated with high litter quality. Optical traits performed equally well as directly measured traits in explaining the invasion impact on the stabilization factor (R2 = 41.9% vs. 38.5%). Furthermore, the interpretation of the results based on optical traits resulted in a similar functional understanding of the invader impact. Synthesis. Our results indicate the potential of hyperspectral data to explain changes in ecosystem functioning. The combination of radiative transfer models and vegetation indices allowed to extract all relevant trait information from the hyperspectral data. This framework thus presents a practical shortcut to assess relevant leaf traits, requiring only a limited amount of field trait measurements.

Unveiling the transferability of PLSR models for leaf trait estimation: lessons from a comprehensive analysis with a novel global dataset

“…We synthesized 37 datasets from published literatures and the EcoSIS Spectral Library (https://ecosis.org/) (Hosgood et al ., 1994; Jacquemound et al ., 2003; Singh, 2013; Serbin et al ., 2014, 2017, 2018, 2019a,b,c,d, 2021; Serbin & Townsend, 2014; Couture, 2015, 2016; Couture et al ., 2015; Meerdink, 2016; Serbin, 2016; Serbin & Rogers, 2016; Yang et al ., 2016; Kattenborn et al ., 2017; Wu et al ., 2017, 2019a,b,c; Ely et al ., 2018, 2019; Kamoske et al ., 2018; Chlus, 2019; Ge et al ., 2019; Nakaji et al ., 2019; Wang, 2019a,b, 2022; Grzybowski et al ., 2020; Helsen et al ., 2020a,b; Burnett et al ., 2021a,b,c,d; Villa et al ., 2021a,b; Chen et al ., 2022; Chlus & Townsend, 2022a,b; Kothari et al ., 2022a,b,c), compiling a new dataset that covers all major biomes and climate zones over the world (Fig. 1a, Supporting Information Fig.…”

Invasion at the Edge: The Case of Rosa rugosa (Rosaceae) in Italy

“…Concerning R. rugosa, Zhang et al [50] showed that the invasive population in Germany outperformed the native population in China in terms of reproductive traits (i.e., seed production) and growth traits (i.e., shrub size), but some introduced populations appeared to be more successful invaders than others. The study by Helsen et al [51] also pointed out that R. rugosa had limited trait overlap with the invaded plant community where it is invasive: on average, it contributed to the functional space of the community introducing highly conservative leaf economic traits and strong light-competition related traits. A more recent study by Helsen et al [52] provided a comparison based on five functional traits related to the leaf economics spectrum (specific leaf area (SLA), leaf dry matter content (LDMC), leaf nitrogen content (N), phosphorous content (P), potassium content (K)) and two traits related to light-competition ability (plant height and leaf area) between individuals of R. rugosa collected in Belgium (invaded range) and those surveyed in Japan (native range).…”

Predicting leaf traits of temperate broadleaf deciduous trees from hyperspectral reflectance: can a general model be applied across a growing season?

“…Leaf spectroscopy is widely used to assess foliar properties in situ, with many applications in agriculture, forestry and ecology (Asner and Martin 2009;Asner et al 2017;Chavana-Bryant et al 2017;Martin et al 2018;Wu et al 2017). Leaf spectra have been used to characterise a number of leaf physiochemical traits Asner et al 2014;Chavana-Bryant et al 2017;Nunes et al 2017;Serbin et al 2019;Wu et al 2017;Wu et al 2019), to predict ploidy levels (Blonder et al 2019), to estimate leaf age (Chavana-Bryant et al 2017;Wu et al 2017), and even to predict the impact of an invasive plant on litter decomposition (Helsen et al 2020). Importantly, leaf spectra measured in the field provide key insights into hyperspectral imagery collected from aircraft, providing underpinning data that contribute to maps of leaf traits and the functional diversity at regional scales (Asner et al 2016;Asner et al 2017;Dahlin et al 2013;Kamoske et al 2021;Knyazikhin et al 2013;Schimel et al 2013;Singh et al 2015;Wang et al 2020), and to improved understanding of the impacts of climate change (Doughty et al 2018), pests and diseases (Fallon et al 2020), and logging (Nunes et al 2021;Swinfield et al 2020).…”