Investigation Of The Spectral Characteristics And The Inversion Of Spartina Alterniflora Chlorophyll Content Under Interspecies Competition

As one essential component of terrestrial ecosystem,wetland plays a critical role in energy exchange and circulation of materials such as carbon and nitrogen.Due to the invasion of alien species,the biodiversity of wetland is facing severe challenges,and thus its ecological security is highly focused.Based on the spectral characteristics of wetland vegetation,the remote sensing technology has become an important means for large-scale,high-efficient and low-cost for monitoring of wetland,which enables to better render ecological security of wetland.However,there are still some deficiencies in previous studies.On the one hand,many studies focused primarily on the species identification and qualitative classification simply based on pure image-pixels,merely shading insights into hybrid habitat of different species,such as the species identification from mixed pixels and their spectrum characteristics.On the other hand,few studies conducted the quantitative inversion of plant chlorophyll(Cab)contents in wetland,particularly in the hybrid region with multiple species,despite they could largely affect the plant physiological and ecological processes.As the most competitive region,the mixed growth area is very critical in wetland ecology,whereas few research studies were conducted by making use of remote sensing and spectroscopy techniques,especially the spectral characteristics and its potential for the quantitative inversion of biochemical contents.To address the abovementioned problems,this study aims to investigate the spectral characteristics of one specific invasive species-Spartina Alterniflora at leaf and canopy levels in Chongming Island,Shanghai,China,which is located in a salt marsh wetland ecological system of Yangtze River Estuary.Toward such a goal,field control experiment and radiative transfer model simulation were conducted,and then the measured spectral responses at different levels were related to chlorophyll contents under different competitive conditions,aiming to establish a quantitative relationship between them for further Cab inversion.The primary research findings of this study include:1.Through a two-year long field control experiment,the leaf and canopy spectral reflectance and biochemical components like chlorophyll content of Spartina Alterniflora(S.alterniflora)and Phragmite Australis(P.australis)were measured under prescribed competitive ratios in Chongming Dongtan wetland,providing the first-hand in-situ data for the investigation of interspecific competition mechanism between S.alterniflora and P.australis.2.Based on the measured leaf reflectance of S.alterniflora and chlorophyll content from the field experiment,the inversion of leaf chlorophyll content under different competitive conditions was examined by making use of a set of 79 indices that were sensitive to chlorophyll content.The results indicate that multi-spectral indices based BP neural network performed better than the single-index based method for the estimation of chlorophyll content,with the RMSE reduced by 10.38%.Meanwhile,the multi-model inversion method yielded higher accuracy than the single model inversion method due to the incorporation of competitive ratios which were not included in the latter,and the RMSE was reduced by 6.4% to 16.3% under different competitive ratios.In general,the method incorporating multi-spectral indices under different competition ratios should be used for the inversion of leaf Cab of S.alterniflora since it could yield higher accuracy.Additionally,the inversion results correlate well with the actual measured values,which overcomes the overestimation issue shown by the single index method.Similar conclusion can be also derived from the PROSPECT model simulations.3.We analyzed the spectral response variations of these two species under different growing phase and mixing ratios,and conducted the inversion of canopy Cab content using the PROSAIL model simulated datasets.The results indicate that these two species can be largely distinguished in late April and early May,owing to their phonological characteristics and competition for natural resources such as sun lights.For the estimation of canopy Cab content,the method incorporating multi-spectral indices and competition ratios could yield higher accuracy compared with the same method without including competition ratios,and the RMSE was reduced by 53.58% on average.4.By making use of remotely sensed high-resolution image from GF-1 in specific seasons,combining with a prior knowledge of habitat for the species,the transitional zone of S.alterniflora and P.australis was successfully extracted,offering practicable outlines for “pure species + mixed transitional zone” classification in wetland.