Response Of Vegetation Phenology To Major Environmental Factors In The Forest Distribution Area Of The Greater Khingan Mountains

The response of vegetation phenology to global climate change is one of the main focuses of terrestrial ecosystem research,which can provide a comprehensive understanding of how global climate change affects surface ecosystems in multiple dimensions.Due to significant and complex spatial heterogeneity in surface environments,there is currently no consensus on how vegetation phenology responds to climate change at the regional scale.The Greater Khingan Range,characterized by less human disturbance and dominated by natural secondary forests,is an ideal region for studying vegetation phenology.Using remote sensing data from MODIS09GQ,MODIS13Q1,and MODIS11A1 V6 on the Google Earth Engine（GEE）,the study calculated the 16-day optimized mean,daily normalized difference vegetation index（NDVI）,and monthly average land surface temperature（LST）for the period from 2000 to 2020.The Mann-Kendall abrupt change test was then used to determine the start of growing season（SOS）,end of growing season（EOS）,and length of growing season（LOS）as phenological indicators.Subsequently,the study analyzed the spatiotemporal variation patterns of vegetation phenology and their response to surface temperature factors in the forest distribution area of the Greater Khingan Range and the distribution area of Larix gmelinii based on four major geographical environmental factors:latitude,slope,elevation,and vegetation coverage.The results are as follows:（1）In the forest distribution area of the Greater Khingan Range,vegetation phenology exhibits significant spatiotemporal variation patterns.On the temporal scale,SOS shows an advancing trend（Slope=-0.66）,EOS shows a delaying trend（Slope=0.61）,and LOS shows an extending trend（Slope=1.27）.Along the spatial（latitude）gradient,both SOS and EOS exhibit advancing trends.For every 1°increase in latitude,SOS advances by 0.34d and EOS advances by 0.19d.LOS initially shortens and then lengthens,with the inflection point at 50°N.Before the inflection point,LOS shortens by1d for every 1°increase in latitude,while after the inflection point,LOS lengthens by0.57d for every 1°increase in latitude.The impact of LST on SOS and LOS is significantly stronger than its impact on EOS.The average LST in April has a significant effect on SOS（Sig.=0.000）and LOS（Sig.=0.000）,indicating that higher LST leads to earlier SOS and longer LOS.（2）In the distribution area dominated by Larix gmelinii,the temporal variation of vegetation phenology is more apparent than the spatial variation based on different slopes.On the temporal scale,SOS shows an advancing trend for all slopes,with the eastern slope(0.71 d·a^-1)and western slope(1.02 d·a^-1)showing greater magnitudes than the southern slope(0.50 d·a^-1)and northern slope(0.64 d·a^-1).Except for the eastern slope,which shows an advance,EOS exhibits a slight delaying phenomenon for the other slope aspects.LOS shows an extending trend for all slope aspects.On the spatial scale,the latitude difference between 51°N and 53°N does not have a significant influence on the vegetation phenology of Larix gmelinii on different slopes,indicating that latitude differences controlled by slope need to be studied within a broader latitude range.The impact of LST on SOS and LOS is significantly greater than its impact on EOS for any slope.The average LST in April and the spring average LST have significant effects on SOS and LOS（Sig.=0.000）.The multi-year average LST is significantly negatively correlated with SOS but has an insignificant positive correlation with LOS.（3）In the distribution area dominated by Larix gmelinii,vegetation phenology shows significant variations with different elevations.SOS exhibits an advancing trend（Slope=-1.02）during the study period,EOS shows a slight delaying trend（Slope=0.51）,and LOS shows a significant extending trend（Slope=1.51）.In terms of spatial distribution,SOS significantly delays with increasing elevation,indicating that vegetation enters the growing season later at higher elevations.EOS and LOS show a pattern of initial extension followed by contraction,with the inflection points occurring at higher values,specifically at elevations between 800 m and 1000 m.LST directly affects vegetation phenology,and the average LST in April and the multi-year average LST have significant effects on SOS and LOS,showing a significant negative correlation.LOS in the low elevation zone has a significant positive correlation with the multi-year average LST,indicating that higher temperatures lead to longer LOS.However,the positive correlation is not significant in the high elevation zone.（4）Considering different vegetation cover gradients,the variation of vegetation phenology in the distribution area dominated by Larix gmelinii shows an optimal scale.On the temporal scale,SOS exhibits a significant advancing trend,EOS shows a delaying trend,and LOS shows an extending trend when both SOS and EOS are taken into account.On the spatial scale,all three phenological indicators,SOS,EOS,and LOS,show trends of initial delay or extension followed by advance or contraction.These trends all have an optimal vegetation coverage scale,with the phenological inflection points occurring in the middle range（average vegetation cover of 49.04%）.Different phenological indicators have varying degrees of response to LST,and the average LST in April is significantly negatively correlated with SOS for any vegetation coverage.LOS shows a significant positive correlation with the average LST in April for different vegetation covers.Vegetation coverage is less influenced by LST and is primarily related to other factors such as vegetation type,precipitation,and soil type.