Woodland Restoration Plot Network: Vegetation Structure and Composition Data, Western Sydney Parklands (Western Sydney Regional Park), Australia, 2012

Abstract: The Woodland Restoration Plot Network Vegetation Structure and Composition Data Package contains vegetation floristics and structure survey data for thirty-six 0.1 hectare sites which were established on retired farmland that includes a mosaic of restored vegetation (native plantings) of varying ages juxtaposed with patches of remnant vegetation and untreated, abandoned pasture. All sites were originally woodland prior to agricultural development about 200 years ago. The Woodland Restoration Plot Network research plots commenced in 1992 and have been revisited every 3-4 years since 2001.

A synopsis of related data packages which have been collected as part of the Woodland Restoration Plot Network’s full program is provided at https://doi.org/10.25911/5c36e5688ff89.

Sampling method: The restoration process was initiated with exclusion of livestock and weed control (J. Christie, Greening Australia, pers. comm.). All sites were slashed and sprayed with glyphosate before planting was undertaken. Twenty-six indigenous trees and shrub species, propagated to tubestock from local seed sources, were planted mechanically in rows after the pasture began to break down. The mix of planted species varied across the landscape, the aim being to match species with soils and topographic positions occupied by their wild populations. All plants were weed matted with a recycled paper disc and surrounded by a protective plastic sleeve. Maintenance sprays of glyphosate were applied in spring and autumn for two-three years after planting to reduce competition from weeds in the vicinity of plants. To reduce the risk of fire, mechanical slashing was carried out amongst the plantings and hazard reduction fires were lit in areas surrounding plantings at approximately annual intervals. Since the late 1990s, cattle have beenreintroduced into the southern areas of plantings, but fencing has been constructed to exclude livestock from some restoration areas and remnants (D. Williams, Greening Australia, pers. comm.). Cattle continue to be excluded from the northern areas of plantings. The study landscape therefore comprises a mosaic of patches including untreated pastures, remnant woodlands with various histories of past disturbance and a chronosequence of restoration plantings with dates of origin from 1992 to 2002. The aims of the study design were to sample management treatments (untreated pasture, the chronosequence of restored vegetation and remnant woodland) across locations within Western Sydney Regional Park and Prospect Reservoir. A fully randomised factorial sampling design was not possible because all management treatments were not represented randomly across the study area. Sample sites representing different management treatments were therefore situated haphazardly within patches, to sample areas with different management histories on similar topography (upper and mid slopes). The first set of vegetation surveys were carried out in 2001 (25 sites), with repeat surveys of new haphazardly selected sites to sample the same patch types in 2004 (52 sites), 2006 (20 sites) and 2012 (36 sites). The total number of sites varied between surveys depending on available resources, but the proportional stratification among patch types and ages remained the same. Invertebrates were sampled at these sites in late 2002 and 2009. Soils were sampled in 2010 by collaborators at UWS. A survey of avifauna commenced in 2013.

Study extent: The study site is located on retired farmland that includes a mosaic of restored vegetation (native plantings) of varying ages juxtaposed with patches of remnant vegetation and untreated, pasture (both grazed and abandoned). During their period of pastoral management (prior to 1990), the sites have been grazed by cattle, fertilised and planted with exotic pasture grasses, particularly Phalaris species. All sites were originally woodland prior to agricultural development about 200 years ago. Restoration projects commenced in the area in 1992 with a stated goal of ‘re-establishment of native vegetation’ (Perkins 1997). The restoration plantings were carried out in a pattern designed to connect remnant patches of woodland, which were also the primary sources of seed for tubestock. To evaluate success against the above goal, we therefore identified the remnants as suitable reference sites to which the restored sites were expected to increase their resemblance in composition and structure over time. Disturbance resulting from past agricultural practices in the area have impacted upon remnant patches to varying degrees, but these were the best available examples of native woodland in the region. Untreated pasture is defined as a control, from which restored sites are expected to become increasinglydissimilar in species composition and vegetation structure with time.

Project funding: Between 2012 and 2018 this project was part of, and funded through the Long Term Ecological Research Network (LTERN) a facility within the Terrestrial Ecosystem Research Network (TERN) and supported by the Australian Government through the National Collaborative Research Infrastructure Strategy.

摘要：本林地恢复样地网络植被结构与组成数据集包，包含36个0.1公顷样地的植物区系与植被结构调查数据。这些样地设于弃耕农田之上，该区域镶嵌有不同林龄的恢复植被（乡土植物种植区）、残存植被斑块以及未受干预的废弃牧场。所有样地在约200年前的农业开发前均为原生林地。林地恢复样地网络的研究样地始于1992年，自2001年起每3-4年开展一次回访调查。

林地恢复样地网络完整项目所收集的相关数据集包的概要，可通过https://doi.org/10.25911/5c36e5688ff89获取。

采样方法：恢复工作始于移除家畜与杂草防控（J·克里斯蒂，澳大利亚绿化组织，私人通信）。所有样地在种植前均进行了刈割与草甘膦（glyphosate）喷施处理。在牧场开始退化后，研究人员采用机械方式成行种植了26种乡土乔灌木物种，这些物种均由当地种源繁育为容器苗（tubestock）。种植物种的组合随区域而异，目标是使所选物种匹配其野生种群所占据的土壤与地形位置。所有植株均覆盖再生纸盘以抑制杂草，并以塑料护套包裹防护。种植后的2-3年内，于春秋两季喷施草甘膦进行维护，以降低植株周边杂草的竞争压力。为降低火灾风险，在种植区内开展机械刈割，并约每年在种植区周边区域开展一次防火减灾烧除（hazard reduction fires）作业。自20世纪90年代末起，牛群被重新引入种植区南部，但已搭建围栏以排除部分恢复区域与残存植被区内的家畜（D·威廉姆斯，澳大利亚绿化组织，私人通信）。种植区北部则持续排除牛群。因此，本研究区域由多种斑块镶嵌组成，包括未受干预的牧场、具有不同过往干扰历史的残存林地，以及起源时间为1992年至2002年的植被演替序列（chronosequence）恢复种植区。本研究设计的目标是在悉尼西部区域公园与展望水库的范围内，对不同管理措施（未受干预的牧场、恢复植被演替序列、残存林地）进行采样。由于无法在研究区内随机分布所有管理措施处理，因此无法采用完全随机析因采样设计。因此，代表不同管理措施的样地被随机设置于各斑块内，以在相似地形（上坡与中坡）上采样具有不同管理历史的区域。首轮植被调查于2001年开展（25个样地），后续于2004年（52个样地）、2006年（20个样地）与2012年（36个样地）对新随机选取的样地开展重复调查，以覆盖相同的斑块类型。各次调查的样地总数因可用资源而异，但斑块类型与林龄间的比例分层保持一致。2002年末与2009年，于上述样地开展了无脊椎动物采样。2010年，西悉尼大学（University of Western Sydney, UWS）的合作者开展了土壤采样。鸟类区系调查则于2013年启动。

研究范围：本研究样地设于弃耕农田之上，该区域镶嵌有不同林龄的恢复植被（乡土植物种植区）、残存植被斑块以及未受干预的牧场（包括放牧牧场与废弃牧场）。在1990年前的牧场管理阶段，这些样地曾被牛群放牧、施加肥料，并种植了外来牧草，尤其是鸭茅属（Phalaris）物种。所有样地在约200年前的农业开发前均为原生林地。本区域的恢复项目始于1992年，其既定目标为“重建乡土植被”（Perkins，1997）。恢复种植的布局旨在连接残存林地斑块，而这些残存林地同时也是容器苗繁育的主要种源地。为评估上述目标的达成情况，本研究将残存斑块设为适宜的参照样地，预期恢复样地的物种组成与植被结构将随时间推移愈发接近参照样地。过往农业活动带来的干扰对残存斑块造成了不同程度的影响，但这些斑块仍是本区域现存最佳的原生林地样本。未受干预的牧场被设为对照样地，预期恢复样地的物种组成与植被结构将随时间推移与该对照样地的差异逐渐增大。

项目资助：2012年至2018年间，本项目隶属于陆地生态系统研究网络（Terrestrial Ecosystem Research Network, TERN）下属的长期生态研究网络（Long Term Ecological Research Network, LTERN），并由澳大利亚政府通过国家合作研究基础设施战略提供资助。