内蒙古典型草原和草甸草原相对花粉产量重估

相对花粉产量（RPP）是利用地层花粉数据定量重建古植被的重要参数，但目前估算的不同地区同一植物类型的RPP多存在差异。为了解植被盖度计算方法对花粉产量的影响，本项研究利用卫星影像数字化重新计算了内蒙古锡林郭勒典型草原和乌拉盖草甸草原采样点周边100 m以外的植被数据，利用ERV模型重新估算了主要花粉类型的RPP，并运用景观重建算法（LRA）进行了检验。锡林郭勒典型草原ERV模型分析显示，ERV子模型1结合1/d距离加权法与植被-花粉数据拟合度最高，得到的相关花粉源范围（RSAP）为1470 m；以蒿属为参照种（RPP=1±0），各科属RPP依次为藜科（0.53±0.03）、十字花科（0.17±0.01）、唐松草属（0.14±0.02）、菊科（0.09±0.01）、莎草科（0.08±0.006）、委陵菜属（0.07±0.01）、禾本科（0.050±0.003）。乌拉盖草甸草原ERV模型分析显示，ERV子模型1结合Prentice距离加权法与植被-花粉数据拟合度最高，得到的RSAP为770 m；以蒿属为参照种（RPP=1±0），各科属RPP依次为藜科（5.85±0.58）、蓼科（5.27±0.37）、石竹科（5.15±0.47）、莎草科（4.16±0.13）、菊科（1.63±0.09）、百合科（0.49±0.08）、唇形科（0.38±0.06）、禾本科（0.200±0.004）。重估的RSAP和RPP均与首次估算时存在差异，表明植被盖度计算方法确实会影响RSAP和RPP。景观重建算法对RPP检验结果表明，卫星影像数字化估算的RPP重建的植被盖度与实际植被的相关性更高。

Re-estimated relative pollen productivity of typical steppe and meadow steppe in Inner Mongolia

Relative pollen productivity(RPP) is an important parameter for quantitative reconstruction of past vegetation using stratigraphic pollen data. However, the relative pollen productivity of the same plant type is different in different regions. In order to understand the effect of vegetation calculation on pollen productivity, by digitalizing satellite images, we have recalculated vegetation data beyond 100 m in the typical steppe of Xilingol area(42°18'~43°55'N, 115°34'~117°14'E), Inner Mongolia, and in the meadow steppe of Wulagai area (45°05'~46°38'N, 118°27'~119°19'E), by using the extended R-value(ERV) model, the relative pollen productivity of major pollen types was re-estimated, and the Landscape Reconstruction Algorithm(LRA) is used to test. Analysis of the ERV model of typical steppe in Xilingol shows that, sub-model 1 combined with 1/d distance weighting method had the highest goodness of fit with the vegetation-pollen data, the relevant source areas of pollen(RSAP) was obtained at the distance of 1470 m from each sampling point. With Artemisia as the reference taxon(RPP=1±0), the RPP of other plants were Chenopodiaceae(0.53±0.03), Brassicaceae(0.17±0.01), Thalictrum (0.14±0.02), Asteraceae(0.09±0.01), Cyperaceae(0.08±0.006), Potentilla (0.07±0.01), Poaceae(0.050±0.003). Analysis of the ERV model of the meadow steppe in Wulagai shows that, sub-model 1 combined with Prentice model had the highest goodness of fit with the vegetation-pollen data, the RSAP was obtained at the distance of 770 m from each sampling point. With Artemisia as the reference taxon(RPP=1±0), The RPP of other plants were Chenopodiaceae(5.85±0.58), Polygonaceae(5.27±0.37), Caryophyllaceae(5.15±0.47), Cyperaceae(4.16±0.13), Asteraceae(1.63±0.09), Liliaceae(0.49±0.08), Lamiaceae(0.38±0.06), Poaceae(0.200±0.004). Re-estimated RSAP and RPP both differ from the first estimate, suggesting that vegetation coverage calculations do affect RSAP and RPP. The results of LRA test on RPP showed that the digital estimation of satellite images was more related with the actual vegetation coverage.