Research On Vegetation Carbon Phenology Dynamics And Modeling Methods Of Pinus Palustris Ecosystems

Forested ecosystems play important role in the carbon and water cycle of global terrestrial ecosystems,in the context of global climate change.Due to the photosynthesis capacity of plant communities in the forest(understory-midstory-canopy)during the day time,forested ecosystems can regulate atmospheric CO₂ effectively.However,with the progress of climate change,there are more uncertainties in the future carbon sequestration capacity of terrestrial forest ecosystems,and these uncertainties are likely to be reflected in the plant phenology of forest ecosystems.Research on the phenological process of forest ecosystems across the landscape,will greatly improve the current knowledge of carbon cycle process of global terrestrial ecosystems,sustainable forest management practices,and better response to the global climate change.Traditional forest ecosystem phenology mainly focuses on the length of vegetation growing season,and the date changes of biological events such as budding,maturity and senescence of plant communities.In recent years,with the application of Eddy-Covariance(EC)flux observation technology,forest ecosystem phenology has been given new perspectives and has derived such as:plant community-scale photosynthesis phenology,net carbon uptake phenology and ecosystem respiration phenology.Further research on the phenological processes of forest ecosystems based on EC-related technologies could provide better understanding of terrestrial surface phenological processes and prediction of the annual carbon dynamics of forest ecosystems in the future,and ultimately achieves a better response to global climate change.There are vast subtropical mixed coniferous and broad-leaved forests in the southeastern United States.Due to the suitable annual average air temperature and sufficient photoperiod,the southeastern forest landscape has a considerable potential for carbon sequestration.Among them,the long-leaf pine(Pinus palustris Mill.)is one of the main native subtropical coniferous forest species of this region.The two-year conifer regeneration cycle provides this species with considerable ecosystem productivity potential.This research is based on the three long-leaf pine ecosystem EC CO₂ flux observation sites in this region with varieties of soil water holding capacity,and uses remote sensing(RS)and Li DAR observation data from 2009 to 2019 with a phenological model,to analysis and discuss the characteristics of the vegetation carbon phenology process of the three subtropical long-leaf pine ecosystems and its response to global climate change,periodic weather events,natural disasters,and forest management.Finally,this research also addresses the uncertainties in the modeling of long-leaf pine ecosystem respiration rate and its relationship with RS-derived leaf greenness phenology.The main content and conclusions of the study are as follows:(1)Using the phenological model to fit the Gross Primary Productivity(GPP)observation data of three mature subtropical long-leaf pine sites from 2009 to 2017,and combining site-specific micrometeorological data.The results show that the gradient of soil water holding capacity will affect the forest phenological process at the ecosystem-scale,and higher soil water content will extend the growing season of the ecosystem.The low-intensity prescribed fire in the early spring has no significant effect on the phenological process of the ecosystem by site,but the it delays the start of the growing season by ten days on average,and as the site-level soil water holding capacity increases,the start of the growing season will be further delayed(eighteen days).Low-intensity prescribed fire is also related to the increased sensitivity of early spring phenology to radiation and air temperature by site.Finally,this study suggests that inter-annual climate change and periodic weather anomalies has a stronger impact of the annual ecosystem phenology process than the low-intensity prescribed fire.When the water availability increases after a short-term drought during summer,the growing season of the ecosystem will be significantly extended.(2)Hurricane can impact the structure and function of the forest by removing the leaf area of the ecosystem,reducing biomass,and causing plant mortality.On October 10^th,2018,Hurricane Michael made landfall in the northern Gulf of Mexico and caused severe damage to the forests along the way.Our results show that hurricane accelerates the senescence rate of ecosystems in the autumn,which leads to a significant shortening of the growing season.During the passage of the hurricane,the long-leaf pine site with a higher canopy suffered more physical damage and therefore significantly extended the phenological recovery period after the hurricane.In the summer of the first year after the hurricane Michael(2019),the ecosystem physiological functions recovered to pre-hurricane conditions.During the restoration of ecosystem phenology after the hurricane,the heterogeneity of site-level forest structure controls the spring phenology process of the ecosystem.Understory biodiversity contributes the ecosystem recovery in early spring,but also delays the start of the peak ecosystem growing season.The post-hurricane summer phenology is synchronized with the vegetation coverage,and is mainly driven by the forest canopy.(3)Comparing the two phenological modeling approach of ecosystem respiration,the growth rate method(GR)and the third derivative method(TD),the former method may be more suitable for the phenological modeling of subtropical forests since some important phenological signals may be lost during the buffer period by the third derivative method.Meanwhile,the study also shows the uncertainty of ecosystem phenology modeling due to climate change.For example,warming in winter will lead to re-activation of ecosystem respiration,and changes in site water availability in summer will lead to significant effect on the ecosystem phenological process,and exhibit multi-peak and asymmetric behavior.Although there is a significant negative correlation between the end date of ecosystem respiration and the dormancy date derived from the 2-band enhanced vegetation index(p<0.01),this could indicate that the early end of the vegetation growing season will increase the ecosystem respiration rate during winter.However,the leaf greenness phenology obtained from remote sensing observations may not fully explain the vegetation carbon dynamics of subtropical evergreen plant community in spring and summer.This research provides a new perspective of forest ecosystem phenology provided by EC technologies combining multi-source measurements and phenology model.The characteristics of the phenological processes of three typical subtropical long-leaf pine ecosystems in the past ten years were evaluated in detail(2009-2019),and their responses to forest disturbances and climate change.The technical approach and model processing methods presented in the research can be used as services and reference for research on other similar sites.Finally,the outcomes of this Ph.D.dissertation have important theoretical and practical significance for an improved understanding of the land surface phenology,scientific forest management and better establishment of the post-disaster emergency management system.