【作者】 郭秋菊；

【导师】 王得祥；

【作者基本信息】 西北农林科技大学 ， 森林经理学， 2013， 博士

【摘要】 长叶松（Pinus palustris）一直以来被认为是美国南方最适宜的树种，曾广泛分布在整个大西洋和墨西哥湾沿岸平原。欧洲移民的大量涌入加速了对长叶松林的采伐和利用，强烈地破坏了长叶松林生态系统平衡，林分面积急剧下降，目前已经不足原来的5%。因此，无论是顾全生态效益，还是着眼于经济效益，恢复和发展长叶松生态系统已经成为美国森林可持续经营的主要任务。异龄林的择伐方式能够为长叶松林下幼苗更新创造较开放的生长环境，计划火烧也可以控制病虫害和减少林下阔叶树的竞争，并将枯枝落叶的营养释放到土壤中，为幼苗更新提供营养，还能减缓地被物对长叶松林下更新的负面影响，利于长叶松生态系统的恢复和可持续发展。本研究以美国三个主要长叶松林原始分布区域（SDE，ERI，BRF）的幼苗更新为研究重点，通过分析和研究主要干扰方式——择伐干扰（包括单株择伐和群团状择伐）和火干扰（火烧季节，火烧频率）对幼苗更新的影响和作用，以期解释影响长叶松林更新的主要影响要素，为促进长叶松天然更新及该系统的可持续发展提供合理的科学依据，推进美国恢复生态学的重要项目——从基因到生态系统推译长叶松生态系统的生态管理和恢复的顺利实施，同时对全球不同森林的生态恢复有着积极的参考价值。主要研究结果如下：⑴通过对三个研究区域不同择伐模式干扰下长叶松幼苗更新密度、更新频度进行调查和分析，显示择伐干扰明显提高了林地长叶松幼苗更新密度和更新频度。不同择伐模式干扰程度下幼苗更新密度表现为未择伐<轻度择伐<重度择伐；单株择伐和群团状择伐模式下的幼苗更新频度较高，都高于80%，而未择伐林地上幼苗更新频度低于50%。不同研究区域的林分断面积与长叶松幼苗更新频度有显著线性相关关系。⑵群团状择伐模式林地的长叶松幼苗高生长量较未择伐林地上幼苗高生长量高，小幼苗（<15cm）高生长量值高0.71cm a^-1，大幼苗（15⁵0cm）高生长量值高10.01cma^-1。群团状择伐模式条件下的长叶松小幼苗成活率都大于40%，比未择伐林地上小幼苗成活率高6.95%；大幼苗成活率也高于70%，比未择伐林地上大幼苗成活率高25.91%。大幼苗的成活率是随着择伐强度的增加及随着林分密度减少而增加，而长叶松小幼苗成活率随着择伐强度的增加而降低。⑶比较不同择伐模式干扰条件下两年生长叶松幼苗基径生长量（GLD）表明：群团状择伐模式条件下两年的幼苗平均基径生长量为1.89mm，单株择伐模式条件下长叶松幼苗两年的平均基径生长量为1.86mm，比未择伐林地上的幼苗基径生长量高出0.38mm。⑷利用回归分析方程对幼苗的根系深度进行预测，结果显示预测的最大根系深度出现在群团状择伐模式下，与实际观测值一致，为58.18cm。不同择伐模式下不同大小幼苗10cm处的主根直径预测值与实际值无显著差异，只有在群团状择伐模式下小幼苗的预测和实际观测值较显著差异值。⑸对林分郁闭度及林下透光率、土壤水分含量与幼苗生长量进行回归分析，结果显示长叶松幼苗基径生长量与林分郁闭度呈负相关响应，而对林下透光率有一定的正相关响应，并且这种响应关系只有在群团状择伐林地上较为明显。所有幼苗的基径生长量与土壤水分含量呈一定的正相关关系，但是相关系数很小，最大值为0.14。⑹比较不同季节的计划火烧对长叶松幼苗不同生长时期的幼苗密度影响，发现冬季火干扰后新生期和禾草期幼苗密度最大，秋季火和夏季火干扰后幼苗密度低于冬季火干扰，且秋夏两个季节干扰后幼苗更新密度无明显差异。比较不同季节火烧后长叶松不同时期幼苗的基径生长量，冬季火干扰后长叶松新生期幼苗和禾草期幼苗基径都达到不同干扰模式中的最大值，而夏季火干扰后则呈现所有火烧模式的最小值。⑺长叶松生态系统中大多数的维管束植物和非维管束植物覆盖率对不同频率的火烧干扰无明显响应变化。对照林地草本植物相对较少，覆盖率仅为2%，而在不同火烧干扰林地上草本植物覆盖率达到20%以上，在两年一次的火烧干扰林地内草本植物覆盖率达到40%。两年一次的计划火烧使得林下植物覆盖率比对照林地高2500%以上。⑻不同火烧频率林地林下维管束植物种类显著高于无火烧林地的种类。林下植物种类在两年一次的计划火烧林地最多，为39种，比每年一次火烧林地的植物多5种，比三年一次火烧林地多8种。无火烧林地内固定出现的林下植物仅为17种。⑼计划火烧干扰显著降低了林下枯枝落叶覆盖率。对照林地内的枯枝落叶较多，在无火干扰条件下，林下枯枝落叶物覆盖度平均为19%。三年一次火烧林地枯枝落叶覆盖度为15%，与对照无显著差异。两年一次火烧林地的枯枝落叶覆盖率为9%，显著低于对照和三年一次林地上的枯枝落叶覆盖率，并且比每年一次火烧林地上的枯枝落叶覆盖率低2%。⑽火烧干扰林地草本植物的生物量明显高于无火烧林地草本植物生物量。两年一次火烧林地的草本生物量为602kg·hm^-2，三年一次火烧林地植物生物量为284kg·hm^-2，每年一次火烧林地草本生物量为217kg·hm^-2，无火烧对照林地的草本生物量很小，仅为15kg·hm^-2。异龄林择伐抚育管理可以维持长叶松林地上的林冠覆被，频繁的计划火烧改变了长叶松生态系统的林地结构和林下植物组成，因此综合使用这两种抚育管理措施可以提高长叶松幼苗成活，加速长叶松幼苗生长，促进长叶松生态系统可持续发展。更多还原

【Abstract】 Longleaf pine （Pinus palustris） is known as the best timber tree in southern UnitesStates. Prior to European settlement, this ecosystem was widely distributed throughout thewhole Atlantic and Gulf Coastal Plain. Since then the ecosystem has suffered a progressiveexploitation, decline and resulted in unbalanced ecosystem, with current levels estimated atless than5%. Therefore, depending on the economic and ecological benefits, restoration anddevelopment of longleaf pine ecosystem have become the main tasks of sustainable forestmanagement in Unites States.Uneven-aged selection would create open environment which are benefit to seedlingregeneration. The prescribed fire provide some benefits that are reducing the competitionfrom understory vegetation, releasing nutrients immobilized to the infertile soil, andsubsequently more rapid uptake by seedling, burning harmful insects and pathogens andreducing litter cover in developing longleaf pine regeneration, and finnaly stimulatinglongleaf pine ecosystem restoration and sustainable developmentThe study focused on seedling regeneration in three original main longleaf pine growtharea（sSDE，ERI，BRF）. The effects and influences of the selection （single-tree and group-tree）and fire （burn season and frequency） disturbances on seedling regeneration were explored andanalyzed, and hope these results will provide a sound scientific basis for longleaf pine naturalregeneration and sustainable development, and promote U.S. important porject-from genes toecosystems, translating longleaf pine ecosystem management and restoration, and also can beprefered as important references for different ecological forest system restoration. The mainfindings are as follows:⑴The regeneration density and frequency under different selection system in threestudy areas were recorded and analyzed, and the selection management improved theregeneration density and frequency. The regeneration density in stands was sequenced as: noselection<single-tree selection<group-tree selection. The regeneration frequency was higherthan80%in single-tree selection and group-tree selection system, less than50%in noselection stands. The regeneration frequency had the significant linear relation with the basal area.⑵The seedling height growth was higher in group-tree selection stands than its in noselection stands, and the small seedling was0.71cm a^-1and the large seedling was10.01cma^-1, respectively. The small seedling survival rate was higher than40%, and6.95%higherthan its in the no selection stands; the large seedling survival rate higher than70%, and25.91%higher than its in the no selection stands. The large seedling survival rate increasedwith the selection density and stand density decrease, and the small seedling survival ratedecreased with the selection density increase.⑶Comparing the seedling groundline diameter（GLD） under different selection system,the average GLD was1.89mm for two years in group-tree selection stands, and1.86mm insingle-tree selection stands, which was0.38mm higher than in the no selection stands.⑷The regression analysis was used to predict the root depth of longleaf pine seedlings.The results presented the maximum root depth （58.18cm） should appear in the group-treeselection stands, and was consistent with the observed value. The diameter at10cm depth ofthe taproot from different class seedlings under different selection system was notsignificantly different between the predicted and observed values, only the diameter at10cmdepth of the small seedlings taproot under group-tree selection had significant differencebetween the predicted and the observed values.⑸Regression analyses gave some relations between understory light environment andsoil moisture and seedling growth. A negative trend was between seedling growth and canopyclosure, but a positive trend with light fraction, and all these relations were significant ingroup-tree selection stands. A positive relationship existed between mean percent soilmoisture and seedling growth for all seedlings but the relationships were weak, and thehighest r2was0.14.⑹The new seedling recruitments and grass-stage seedlings were most abundant bywinter prescribed fire burn, and than the autumn and summer burn. No difference was foundbetween the autumn and summer burn for seedling denstiy. The comparison of the GLD fromtwo stage seedlings presented the GLD of new seedling recruitments and grass-stageseedlings were got the highest value after winter prescribed fire, and the lowest value aftersummer prescribed fire.⑺Most categories of vascular and non-vascular plants did not significantly respond tothe fire treatments in longleaf pine ecosystem. Grasses in the unburned no burn plots persistat very low levels of less than2%total cover, while grasses in all of the fire-treated plotsexceed a total cover of20%, with the biennial fire treatment approaching40%foliar cover.Biennial burning increased the overall foliar cover of grasses by2500%over that present in the unburned no burn areas.⑻The understories of all fire-treated plots contained significantly higher numbers ofvascular plant species than did the unburned no burns. Species richness for the biennial firetreatment was highest, with39species present, followed by the annual fire treatment, whichsupported34species, and the triennial fire treatment, which resulted in31species. Only17plant species were typically found in the understories of the unburned no burn plots.⑼Fire treatment resulted in a significant decline in litter cover on the forest floor. Noburn plots had the highest litter cover, averaging19%. Litter cover of15%on trienniallyburned plots was not significantly less than no burns. However, litter cover of9%onbiennially burned plots and11%on annually burned plots was significantly less than ontriennially burned and no burn plots.⑽Standing biomass of herbaceous understory plants was significantly higher on all firetreated plots than on the no burn areas. The largest standing biomass of602kg·hm^-2wasmeasured on the biennial fire treatment, and total284kg·hm^-2for the triennial fire treatment,217kg·hm^-2on the annual fire treatment, and the smallest15kg·hm^-2occurred on the no burnstands.Uneven-aged selection would assure “continuous forest cover in stands. Prescribed firewould improve the stand strucutre and understory composition in longleaf pine ecosystem.The combination of these two managements will promote regeneration survival and growth,and speed up the sustainable development of longleaf pine ecosystem.更多还原