% , " .% , #。四型表现为

应用方差分析、经验正交函数、合成分析,研究了天目山柳杉(Cryptomeria fortunei)树轮中δ13C随方位的分布特征.应用相关分析探讨了这种分布特征的年际差异与环境因素的关系.结果表明:树轮δ13C值有显著方位差异和年际变化,不同方位的差异可以用4个基本角分布型表示,其中一型表示各方位δ13C比多年平均值偏大,其中北(N)、西(W)、西南(SW)、南(S)较平均值偏大约0.5,而其余4个方位偏大不到0.2;二型各方位也较多年平均值偏大,但各方位差异不大,除N方位较小外,其余均在0.2～0.4之间.三型为各方位δ13C较平均状况偏小,其中西北(NW)、东北(NE)方位约偏小0.4,其余方位偏小0.2～0.3.四型表现为NE、东(E)、东南(SE)与平均状况一致,西北(NW)方位较平均状况偏大约0.1,N方位偏小0.2,其余3个方位则偏小约0.3.而其年际变化既可用这4个角分布型的更替表示,也可以用第一、二主成分表示.δ13C的第一主成分与前一年8～12月降水量负相关,与当年1～3月降水正相关.第二主成分与前一年8月至当年1月的6个月平均气温正相关.不同的角分布型与不同的气候状态相联系.一型出现的有利气候条件是:R8～12少、R1～3多、T3～7低并且T8～1高;二型出现的有利气候条件是:R8～12少、R1～3多、T3～7低并且T8～1低;三型出现的有利气候条件是:R8～12多、R1～3少、T3～7高并且T8～1高;四型出现的有利气候条件是:R8～12多、R1～3少、T3～7高并且T8～1低.可见角分布型的差异可能包含着环境变化的信息,因此同时利用树轮整体平均δ13C值的序列和角分布型序列来重建历史气候,可以得到更多的历史气候变化信息.

Studies on circumferential δ 13 C variability in tree ring have been reported.Despite a variety of species and localities,the range of δ 13 C around the circumference of tree ring is typically 0.1%～ 0.4% .However,studies on climatic implications for the difference among radii in tree rings have not yet been reported.In this paper,four trees sampled from different localities(three from Tianmushan Mount.and one from Lushan Mount.)were analysed.More than 7 radii with forty five degree interval from each tree were analyzed.The results showed that δ 13 C differences among radii reached 0.17%,but the directions for the lowest and the highest were different for each tree.The difference among radii in δ 13 C variance to the circumferential difference was in the same order of magnitude.By analyzing the δ 13 C climate relationship,the δ 13 C differences among radii may be used to study more climatic implications.

对采自江西红壤地区的六株马尾松的圆盘年轮样及它们当年的根部土壤,进行A1,P,Fe,Co,Cu,Zn,Ba,Pb等元素含量分析.实测结果表明年轮元素含量的分布平均值C(Z)与当年根部土壤中元素含量的分布平均值C'(Z)之间存在统计相关性,并满足对数线性相关模式:lgC'(Z)＝a(Z)+b(Z) lgC(Z)其中Z是元素原子序数,用以表征不同化学元素;统计回归系数a、b和元素、树种、气候因子以及土壤类型有关.因此,年轮元素含量时间(年份)序列C(Z,ti)可反映根部土壤中元素含量C'(Z,ti)的逐年变化,即再现环境的历史演变.

湖泊沉积研究中常把总有机质含量（TOM）或总有机碳含量（TOC）作为判识湖泊环境的一个重要指标.事实上,湖泊沉积物中的有机碳来源广泛,简单地可分出内源和陆源两部分.

TOTAL organic matter(TOM) or total organic carbon(TOC) is often treated as an importantindicator of lake environment. As a matter of fact, TOC is of multi-origin, and can simply bedivided into two parts: autochthonous and allochthonous organic carbon. The autochthonousorganic carbon is mainly controlled by the production of the plankton of lake itse1f; and the al-lochthonous one by climate and environment in the catchment area, e. g. more precipitationand runoff and densely developed vegetation will give rise to more allochthonous organic com-

The chemical element contents in tree rings are correlated with those in the soils near the tree roots,The results in the present study showed that the correlation between them could be described using the following logarithmic linear correlation model:lgC'(Z)=a(Z)+b(Z)lgC(Z).Therefor,by determining the chrono-sequence C(Z,t),where Z is the atomic number and t the year,of elemental contents in the annual growth rings of trees,we could get the chrono-sequence C'(Z,t) of elemental contents in the soil,thus inferring the dynaminc variations of relevant elemental contents in the soil.

There exists a logarithmic linear correlation,i.e.,lnC'(Z,t)=a(Z)+b(Z)ln(Z,t) where Z is the atomic number of element and t the year when tree ring grow,between the chemical element contents in tree rings C(Z,t)and those in soils near the tree roots C'(Z,t).By determining the elemental contents of the annual growth rings of trees,we could establish the chronosequences of elemental contents in the tree rings.thus calculating that of the soil,that is,reproducing the dynamic changes of contents of elements in the soil C'(Z,t),The background values of elements in the soil under site conditions of the tree could be estimated from the miniumu C(Z,to) in the chrono-sequences of elemental contents in the tree rings.