【作者】 彭容豪；

【导师】 陈家宽；

【作者基本信息】 复旦大学 ， 生态学， 2009， 博士

【副题名】上海崇明东滩实例研究

【摘要】 生物入侵作为全球变化的组成部分,对土著生态系统产生深刻影响,其生态系统水平的后果之一便是改变生态系统的碳氮循环。尽管对此已有大量研究,但仍缺乏对其影响机制的全面深刻认识,尤其是目前的研究多集中于群落内部循环,而入侵种的生态系统工程师效应对系统同外部输入/输出的影响往往被忽视。本文以互花米草(Spartina alterniflora)入侵长江口崇明东滩的盐沼湿地芦苇(Phragmites australis)群落为例,研究了入侵种如何影响生态系统氮循环的科学问题,同时以其中潮汐交换和促淤过程的作用揭示了作为生态系统工程师的入侵种影响外部输入这一机制的重要性。互花米草显著升高了生态系统的碳、氮库。2007年4月至2008年4月的野外观测显示互花米草群落的地上生物量(干重1.54±0.05 kg m-2)显著高于芦苇群落(干重0.87±0.05 kg m-2),约为后者的180%；互花米草群落的地上总碳库(514±18g m-2)显著高于芦苇(333±16g m-2),约为后者的154%；互花米草群落的地上总氮库(14.60±0.65g m-2)显著高于芦苇(10.63±0.54g m-2),约为后者的137%；互花米草群落的0-20cm土壤总碳库(3270±54g m-2)显著高于芦苇群落(2998±49g m-2),约为后者109%；互花米草群落的0-20cm土壤总氮库(175.94±4.91g m-2)显著高于芦苇群落(153.49±5.76g m-2),约为后者115%；互花米草群落的0-20cm土壤无机氮库(2.50±0.06g m-2)显著高于芦苇群落(1.97±0.05g m-2),约为后者的127%,其中互花米草群落的0-20cm土壤硝态氮库(1.71±0.04g m-2)显著高于芦苇群落(1.22±0.02g m-2),约为后者的140%,而互花米草群落的0-20cm土壤铵态氮库(0.80±0.03g m-2)与芦苇群落(0.75±0.05g m-2)相比无显著差异。在2005年4月至2006年9月的移栽试验中,互花米草的地上生物量与地上总碳、氮库同样显著高于芦苇群落,但土壤总碳氮库与无机氮库与芦苇群落差异不显著。这说明互花米草增加生态系统碳、氮库的现象不能仅以群落内部循环机制解释。互花米草在潮汐交换中获取大量外源无机氮输入。2007年进行了土柱法野外操纵实验,2008年进行了大潮期间土壤-潮水无机氮交换的直接观测实验,结果均显示在潮汐交换中互花米草群落获取的无机氮(操纵实验中土壤无机氮库外源补充：14.84±0.67mg kg-1 month-1；直接观测实验中土壤无机氮浓度增量：9.43±1.18mg L-1)显著高于芦苇群落(外源补充：2.97±0.24mg kg-1month-1；无机氮浓度增量：4.92±0.51mg L-1)。操纵实验中互花米草群落土壤无机氮净矿化速率(14.15±0.68mg kg-1month-1)与芦苇群落(13.59±0.69mg kg-1month-1)无显著差异,直接观测实验中互花米草群落内潮水无机氮浓度下降量(2.92±0.36mg L-1)显著高于芦苇群落(2.07±0.20mg L-1)。以上结果表明对潮汐外源无机氮输入的截取与利用是互花米草增加生态系统地上总氮库与土壤无机氮库的重要机制。淤积物是互花米草累积碳氮库的重要来源。于2007年至2008年以埋瓶法和埋板法收集淤积物,结果显示,互花米草群落中淤积量(埋瓶法：6.69±0.68kg m-2 month-1;埋板法：38.60±3.33kg m-2yr-1)显著高于芦苇群落(埋瓶法：4.27±0.53kg m-2 month-1;埋板法：21.42±2.16kg m-2 yr-1).同时,埋板实验结果还显示互花米草群落中淤积物总碳库(0.69±0.08kg m-2 yr-1).总氮库(40.05±4.50g m-2 yr-1)和无机氮库(953±85mg m-2 yr-1)均显著高于芦苇群落中淤积物总碳库(0.42±0.05kg m-2 yr-1).总氮库(24.56±2.97g m-2 yr-1)和无机氮库(415±37mg m-2 yr-1).于2007年至2008年进行的空地凋落物添加试验结果显示,互花米草地上直立凋落物的降解速率(K=0.0038 day-1)高于芦苇的地上直立凋落物(K=0.0021 day-1)；添加互花米草凋落物下方土壤总碳库(1.96±0.02kg m-2)显著高于未添加凋落物空白对照土壤的总碳库(1.91±0.02kg m-2,),而添加芦苇凋落物后的土壤总碳库(1.93±0.02kg m-2)其多重比较组别介乎二者之间；添加互花米草凋落物,添加芦苇凋落物和未添加凋落物空白对照土壤的总氮库与无机氮库均无显著差异。以上结果表明互花米草增加土壤总碳库与总氮库的机制存在差异,后者不能仅由初级生产与凋落物降解过程解释,促淤效应在其中起着重要作用。互花米草对氮循环的影响将作用于自身的扩张。于2008年6月测定互花米草与芦苇叶硝酸盐还原酶活性(nitrate reductase actiVity,NRA),结果显示互花米草叶的NRA(对数值1.69±0.11)整体上显著高于芦苇叶(对数值0.87±0.08)。加氮处理使互花米草与芦苇叶的NRA均显著上升,淹水处理使互花米草叶NRA升高,而芦苇叶NRA降低。对野外观测实验与操纵实验的数据进行多元回归和PCA分析,结果显示淤积物性质与植物群落性质有显著的相关性。以上结果提示互花米草入侵改变生态系统氮循环的后果可能影响其自身扩张,形成一定程度的反馈作用。总之,上述结果强调了在湿地生态系统中,生态系统工程师效应对系统与外界输入/输出的控制是植物入侵影响土著生态系统氮循环的重要机制。这一结论不但有助于对外来入侵植物的管理,也加深了对植物调控生态系统氮循环的理解。更多还原

【Abstract】 As a component of global change, biological invasions have profound effects on the native ecosystems. One of the ecosystem-level consequences of invasive alien species caused is their impacts on carbon and nitrogen cycling. Although a growing number of studies on these impacts have been conducted over the last two decades, the underlying mechanisms are still poorly understood. Recent studies especially concentrate on the production-decomposition cycling within communities, while the system input/output exchange with environment controlled by invasive ecosystem engineer effects is left unconcerned. The first objective of this study was to explore the effects of Spartina alterniflora (SA) invasion on ecosystem N pools and cycling in the Phragmites australis(PA) communities at Dongtan estuarine wetlands, Chongming island. The second objective was to estimate the importance of invasive ecosystem engineer effects (sediment accretion) as a mechanism of invasive impact on native ecosystem nitrogen cycling.Spartina significantly increased ecosystem C and N pool. Field observation experiment was conducted from Apr 2007 to Apr 2008. The results showed that the average aboveground biomass in SA communities(1.54±0.05 kg dry weight m-2) was significantly higher than that in PA communities(0.87±0.05 kg dry weight m-2), respectively 80% greater.The average aboveground total carbon pool in SA communities (514±18g m-2) was significantly higher than that in PA communities(333±16g m-2), respectively 54% greater.The average aboveground total nitrogen pool in SA communities (14.60±0.65g m-2) was significantly higher than that in PA communities(10.63±0.54g m-2), respectively 37% greater.The average soil total carbon pool(0-20cm) in SA communities (3270±54g m-2) was significantly higher than that in PA communities(2998±49g m-2), respectively 9% greater.The average soil total nitrogen pool(0-20cm) in SA communities (175.94±4.91 g m-2) was significantly higher than that in PA communities(153.49±5.76g m-2), respectively 15% greater.The average soil inorganic nitrogen pool(0-20cm) in SA communities (2.50±0.06g m-2) was significantly higher than that in PA communities(1.97±0.05g m-2), respectively 27% greater. While the average soil NO3-nitrogen pool(0-20cm) in SA communities (1.71±0.04g m-2) was significantly higher than that in PA communities(1.22±0.02g m-2), respectively 40% greater, the average soil NH4-nitrogen pool(0-20cm) in SA communities (0.80±0.03 g m-2) and PA communities (0.75±0.05g m-2) had no significant difference. Controlled transplant experiment was conducted from Apr 2005 to Sep 2006. While the aboveground plant pools showed same patterns as in field observation experiment, all the soil pools had no significant difference between SA and PA communities. These contrast results indicated that production-decomposition cycling within communities was not the only mechanism to explain the soil carbon and nitrogen pool increase caused by SA invasion.Spartina obtained extra inorganic N subsidies during tidal exchange. Field manipulation experiment on soil columns in PVC tubes was conducted in 2007, and field observation experiment on soil-water inorganic nitrogen exchange was conducted during spring tide in 2008. Both results showed that SA communities acquired more inorganic nitrogen than PA communities during tidal exchange. In the manipulation experiment, inorganic nitrogen pool increase in soil columns caused by tidal subsidy was significantly higher in SA communities (14.84±0.67mg kg-1 month±1) than in PA communities (2.97±0.24mg kg-1 month-1), in spite of the original community which the soil column was sampled in, while inorganic nitrogen pool increase caused by net mineralization had no significant difference between SA (14.15±0.68mg kg-1 month-1)and PA (13.59±0.69mg kg-1 month-1)communities. In the observation experiment, soil inorganic nitrogen content increase was significantly higher in SA communities (9.43±1.18mg L-1) than in PA communities (4.92±0.51mg L-1), and inorganic nitrogen content decrease in tidal water was also significantly higher in SA communities (2.92±0.36mg L-1) than in PA communities (2.07±0.20mg L-1). All the results revealed that difference in tidal subsidy acquirement was a principal mechanism to explain the increase of aboveground total nitrogen pool and soil inorganic nitrogen pool after SA invasion.The C and N pool accumulated by Spartina were mainly composed of sediments. Bottles and PVC plates were buried in both communities to collect sediments from 2007 to 2008. Sediment load in both methods was significantly higher in SA communities (in bottle:6.69±0.68kg m-2 month-1; on plate:38.60±3.33kg m-2 yr-1) than in PA communities (in bottle:4.27±0.53kg m-2 month-1; on plate:21.42±2.16kg m-2 yr-1). Besides, the total carbon pool, total nitrogen pool and inorganic nitrogen pool of sediments on plate were also significantly higher in SA communities (TC:0.69±0.08kg m-2 yr-1; TN:40.05±4.50g m-2 yr-1; inorg-N:953±85mg m-2 yr-1) than in PA communities (TC:0.42±0.05kg m-2 yr-1; TN:24.56±2.97g m-2 yr-1; inorg-N:415±37mg m-2 yr-1). Standing dead litter of SA and PA was added to bare ground from 2007 to 2008. SA litter had significant higher aerial decomposition rate (K=0.0038 day-1) than PA litter (K=0.0021 day-1). In multiple comparison test, Soil total carbon pool in quadrates with SA litter addition (1.96±0.02kg m-2) was significantly higher than that in the control groups (bare ground without litter addition, 1.91±0.02kg m-2), soil total carbon pool in quadrates with P A litter addition (1.93±0.02kg m-2) had no significant difference with both. Soil total nitrogen and inorganic nitrogen pool showed no significant difference on the whole. Thus, the conclusion was that sediment accretion effect was an important mechanism for SA to raise soil total nitrogen and inorganic nitrogen pool as an invasive ecosystem engineer, while the rise of soil total carbon pool could be mainly explained by increase in NPP and decomposition rate.The impact of Spartina on N cycling also affected the expansion of this invasive plant itself. Nitrate reductase activities (NRA) in SA and PA leaves were measured in Jun 2008. The NRA in SA leaves (1.69±0.11, In value) were significantly higher than that in PA leaves (0.87±0.08, In value) on the whole. Nitrogen fertilization could significantly raise NRA in both leaves. Submerging treatment significantly increased NRA in SA leaves, and decreased NRA in PA leaves. When using multiple regressions and PCA method to analyze datasets in field observation and manipulation experiments, sediment properties were found to have significant correlation with plant community characteristics. These results suggested that the consequence of impact on nitrogen cycling in native ecosystem invaded by SA might affect the expansion of SA itself. Such interactions were likely to create potential feedbacks.All the results in this study emphasized the importance of the control on ecosystem exchange with environment by invasive engineers as an underlying mechanism to explain the impact on nitrogen cycling in native ecosystems caused by invasive plants. This conclusion provided basic information for invasive plants management, as well as an insight into the mechanisms that individual plant species (including non-invasive plants and native plants) changed ecosystem-level biogeochemical cycling.更多还原