| Ongoing changes in global climate are challenging a wide range of species through altering their habitat conditions,particularly on the edge of the geographical distribution,community structure and population dynamic,which resulting in species' range expansions or contractions.Up to now,most studies focused on upward or polo-ward range shifts.However,less attention was put on compare synchronously range dynamics associated with bio-factors across latitudinal,longitudinal and altitudinal gradients for a certain woody species.In the present study,we predicted the potential distribution of Quercus variabilis Blume,one of the most widely distributed species in East Asia,under past(Mid Holocene,6000 years ago),current(1950–2000)and future(2050)climate conditions by the maximum entropy model(Maxent).Then,we investigated the differences in the population demography of Quercus variabilis,along latitudinal,longitudinal and altitudinal gradients covering the full distribution range.Specially,we focused on geographical patterns of their demographical structure(age,density and mortality rate),regeneration(recruitment rate and conversion rate),growth(radial and height increment)and population spatial pattern.Furthermore,environment factors(climate,soil and stand conditions)shaping the distribution range limits across three geographical gradients was ranked and apportioned using the redundancy analysis(RDA)and Linear mixed-effects models.(1)The potential distribution of Q.variabilis under the current climate conditions centered on the south of East Asia,comprising up to 21.88% of the total studied area,comprising 5.69% for the suitable habitats.The climate change scenarios showed that the areas of the suitable habitats have changed slightly,while significantly for the distribution range.With changes in global climate volatility,the distribution centers of Q.variabilis are gradually concentrating in the regions of Qinling and Daba Mountains and Yunnan-Guizhou Plateau,and the areas of suitable habitats are gradually expanding.Minimum air temperature of the coldest month(Bio6),mean air temperature of the coldest quarter(Bio11),and annual precipitation(Bio12)were found to be dominant in influencing the geographic distribution of Q.variabilis,with contribution rates of 48.6%,21.4%,and 14.2%,respectively.(2)Over the horizontal gradients,the population structure of Q.variabilis exhibited an inverse-J shape in the north,middle and west sites,but was not observed in the south and east sites due to the scarcity of individuals or fault age in recruitment populations.The relative seedling density,seedling density and survival from birth to the highest class of regeneration(SR)were all significantly higher in the Middle populations and decline towards to distribution range along the latitude and longitude,particularly in the south and the east sites In altitude,an inverse-J-type population structure were present in the low and middle altitude populations but was not observed in the high altitude due to the scarcity of individuals in regenerated populations,and both the seedling density and sapling density,along with the y0 and SR,decreased with altitude.But no significant differences in the density and total basal area of adults were found across latitudes,longitudes and altitudes.In addition,the sexual and asexual reproduction jointly maintained the population development of Q.variabilis,and presented a collaborative relationship,rather than the trade-offs.(3)The growth of Q.variabilis was significantly influenced by the latitude and longitude,but no significant effect in the altitude.The radial growth of Latewood and Ringwood(radial width,growth rate and basal area increment)were all increased with latitude,but no significant trends was found for the height growth across latitudes.In longitude,the radial growth of Latewood(radial width,growth rate and basal area increment)and height growth were both decreased with increased longitude,but no significant trends was found for the Earlywood and Ringwood.In altitude,the radial growth of Latewood was signficant higher in height altitude than the low and middle,but no significant trends for the Earlywood,Ringwood and height growth.The DBH had positive correlations with the age in all geographical gradients,but the increment rate of DBH was varied among the sites(north>south;west>east;high altitude> middle altitude> low altitude),and the age up to maximum increment were different for all sites.(4)The growth pattern and the radial growth response to the climate of Q.variabilis were different across latitude and longitude,but no significant variance was found across altitude.The north and south sites have the same growth pattern respectively,but the growth pattern was complicated across longitudes.The growth of northernmost sites(BJ)had positive correlations with the precipitation of April and June,but negative with the max temperature of June.The growth response to climate were similar between the north sites of LC and JY and the middle sites(BTM),which all had positive correlation with the temperature of previous September and current Spring,and negative correlation with the precipitation of previous September,but the response degree of growth to the climate was increased from LC to BTM along latitude.From the middle sites to the southernmost,the correlation between the growth(Earlywood and Ringwood)and temperature was decreased,except the Latewood.In longitude,the growth of west sites(TS,MX and SL)response to climate is similar with the middle,which were positive correlated with the precipitation of previous and current September,but no significant correlation was found between the growth of east sites(JZ and NJ)and precipitation.The Earlywood and Ringwood of east sites were negative correlated with temperature,while the Latewood present positive correlation with temperature.In altitude,the growth response to the climate were similar,which all present the positive correlation with the temperature of spring and early Summer,and negative correlation with the temperature of previous September.(5)Over the horizontal gradients,the spatial patterns of adult individuals were all clumped except the northern populations which were randomly distributed;while juveniles were clumped in the northern,middle and western populations,but were randomly distributed in the southern and eastern populations.The aggregation intensity of juveniles across latitude decreased with the order as the central,northern and southern populations,but as the central,southern and northern populations for adult individuals.The aggregation intensity of both juveniles and adults across longitude followed a decreased order as the central,western and eastern populations.Along the altitudinal gradients,both juveniles and adults in populations across altitude were clumped,except the adults of low-altitude population with random distribution.The aggregation intensity of both juvenile and adult were higher in the middle than the other altitudinal populations.Compared with adults,juveniles generally have higher aggregation intensity across various geographical gradients.Our results revealed that the spatial distribution pattern of Q.variabilis were mainly determined by environment variation across geographical gradients and the species' biological property,and supported the central-marginal hypothesis.(6)In large-scale areas,canopy density,precipitation and soil PH were identified as the primary factors affecting regeneration,and the growth were mainly affected by the climate(temperature and precipitation),stand condition(adult density and total BAI)and soil(Ca and Mg).Stand condition and soil were the direct environment factors in determining the geographic variation of regeneration and growth,but the climate could be an indirect drivers of Q.variabilis distribution,affecting the competition and light,especially the range limit of south,east and high altitude,and our result supported the lower-margin-competition and stress-gradient hypothesis.Therefore,populations of Q.variabilis at the range edges would decline in the context of climate warming,particularly in the South,East and high altitude,and these results were partly consistent with the model prediction.To summarize,this study advances our understanding of the population dynamics of woody species along latitudinal,longitudinal and altitude gradients,and suggesting a potential range shift across horizontal and vertical geographical gradients under climate change,although further evidentiary support needs to be obtained in future studies. |
PDF Full Text Request