Ecological restoration aims to revitalize ecosystem integrity and functionality following severe damage or degradation. Often, however, efforts are hampered by an incomplete or flawed concept of historical 'reference' used when choosing or constructing a target ecosystem or landscape to restore 'to'. This problem may stem from a culturally-skewed interpretation of history or from misunderstanding or underestimation of the role that humans have played in a given ecosystem's historical development and dynamics. While strongly confirming the importance of the reference concept in restoration ecology, we argue for the need to refine it, and to broaden the ways it can be conceived, developed, and applied. Firstly, the historical reference system informing a given restoration project should be grounded in both latent and active 'ecological memories', encoded and stored across relevant geographical and temporal scales. Further, the generally neglected geomorphic component of referencebuilding should also be addressed, as well as the contributions of human cultures to current ecosystem and landscape condition. Thirdly, ecosystems are historically contingent and multi-layered. Pre-versus post-disturbance compar-isons are insufficient. Instead, restoration scenarios should be seen as tapestries of multiple and succes-sive states. In sum, a well-conceived reference model helps promote and ensure the recovery and subsequent maintenance of historical continuity, i.e., the reestablishment of an impaired ecosystem to its historic ecological trajectory. We use case studies from Spain and Peru to illustrate how this approach can provide better goalposts and benchmarks, and therefore better guide the planning, implementation, and evaluation of effective restoration projects.

Roadside reclamation involves standard revegetation practices that often fail under the adverse conditions imposed by subordination to the infrastructure construction schedule. We experimentally tested for seed and microsite limitations on roadslopes by assessing the effects of seed addition and habitat suitability upon plant cover and species richness. The relative contributions of topsoil seed bank, seed rain, and hydroseeding with standard or native seed mixtures were analyzed in relation to soil texture, fertility, and stability. In order to increase applicability, this research was fitted into the actual construction design and schedule of a highway in central Spain, which resulted in topsoil of varying quality, steep roadcuts and embankments (34 • ), and out-of-season hydroseedings. During the first 2 years following roadslope construction, there was an uneven but sustained increase in plant cover and species richness. Topsoil spread on embankments led to greater plant cover in a shorter time and to lower sedimentation rates at slope bases. The topsoil seed bank was extremely poor. Hydroseeding invariably failed, regardless of seed mixture and roadslope type. The seed rain provided seven times more seeds than hydroseedings, and was correlated with the distance to vegetation patches. Recruitment, however, was limited by microsite suitability, as the initial soil content in nitrate, total nitrogen, and organic matter explained up to 80% of variation in plant cover. In conclusion, when revegetation was performed outside the optimal season due to schedule constraints, measures aimed at overcoming microsite limitation were more cost-effective and enhanced roadside carrying capacity for local species.

Over the last few decades, road construction has increased dramatically, and new surfaces have appeared in most landscapes. Standard roadside reclamation practices often fail, because vegetation establishment appears to be limited by microsite availability. We considered soil properties as a key factor driving vegetation establishment on roadslopes over time. We address the following questions: (i) Are soil features conditioned by type of roadslope, position thereupon or applied hydroseeding? (ii) Is there any evidence of soil development at the roadside four years after road construction? (iii) Do mutual interactions exist between soil features and vegetation cover? We designed an experimental set‐up on a highway in Central Spain (Madrid). We selected 15 roadslopes (nine roadcuts and six embankments) with three hydroseeding treatments (commercial, alternative and untreated). Four years after the road construction, we considered three roadslope positions (top, middle and bottom) to take into account the geomorphological gradient. We monitored soil features and vegetation cover over 4 years after the road construction. Soil chemical differences were found between roadslope types, mainly resulted from topsoil spreading on embankments and the weathering of the newly exposed materials on roadcuts. Applied amendments do not affect soil fertility or vegetation cover. In the course of time, vegetation establishment and geomorphological gradients operate differentially on roadcuts and embankments. Accordingly, cycling back of organic compounds or geomorphological processes differs between roadslopes types. Restoration efforts should be directed to guarantee key ecological processes and support soil formation. Copyright © 2011 John Wiley & Sons, Ltd.

Question: Does community assembly on roadsides differ between two contrasting habitats along a stress–productivity gradient? Is establishment success determined by the regional pool, environmental filters or historical events? Location: Highway roadcuts and embankments in central Spain (40°29′N, 03°34′W). Methods: Species composition was recorded annually in 45 plots distributed on steep slopes with newly exposed surfaces (roadcuts) or newly built topsoiled substrates (embankments), for 4 years following hydroseeding with standard or native seed mixture. Frequency of appearance and local colonization and extinction rates of individual species were clustered in establishment success groups. We examined the correspondence between descriptive plant traits and species performance on both roadslope types. Results: Roadslope species richness showed a sustained increase over time, although at consistently lower levels and rates on the more productive embankments. Sixty per cent of the colonization events involved species from the surrounding vegetation. Hydroseeded species persisted through time, but did not modify community composition or dynamics. A higher establishment success rate was found in wind‐pollinated species with large seeds and in exotic species. Those species growing on embankments showed an equal or higher establishment success rate on roadcuts and, conversely, species from roadcuts exhibited an equal or lower success rate on embankments. Conclusions: At the initial stages of community assemblage on unproductive newly created areas, species richness was shaped by the regional species pool. Communities on less harsh topsoiled embankments were subjected, however, to the filtering effect of competitive exclusion. In a reclamation context, efforts to increase site productivity may have detrimental consequences for in situ conservation of local diversity.

Questions How do roadsides interact with patches of natural vegetation in shaping perennial plant communities in fragmented agricultural areas? Are the observed differences due to the type of roadside (i.e. roadcuts, embankments or road verges) or are there other factors driving community structure and composition? Location Stretch of motorway A3 and its surrounding area, central Spain. Methods We analysed the variation in perennial plant species composition and diversity among 92 plots (400 m2). The plots were located in five different environmental scenarios, three of them in a fragmented landscape (patches of natural vegetation, embankments and roadcuts) and two in an unfragmented landscape (natural vegetation and road verges). In each plot, the cover of each perennial plant species and eight soil variables were assessed. We used phi coefficient of correlation to determine the scenario preferences of each species, Kruskal–Wallis tests to compare the soil variables between landscape scenarios and eight descriptive variables of the community, and RDA and partial RDA to evaluate the relative importance of the type of environmental scenario on the floristic community. Results We identified 130 species, with only 16 species never appearing on roadsides. Perennial total cover, species richness, inverse Simpson's index and number of protected species showed no significant differences between the five scenarios considered. In contrast, the number of nutrient‐demanding species and restricted‐range diversity had lower values in natural vegetation plots. Soil variables and the type of scenario together explained 28.5% of the species composition variation. Of this percentage, 6.8% was explained by soil variables, 12.1% by the type of scenario and 10.0% of the variation was shared between the two data sets. Conclusions Our results show that almost all perennial species occurring in natural vegetation patches were also able to reach and settle in roadsides. However, soil conditions and other specific roadside variables generate different plant communities. In spite of the differences found between the perennial plant community of roadsides and their surrounding area, roadsides are excellent reservoirs of biodiversity.

No abstract

Plantings are commonly used in roadside reclamation for ornamental purposes and for increasing slope stability and road safety. However, the role of these plantings in restoring ecological processes, such as seed dispersal, has received little attention. We carried out a study to assess the potential role of plantings on roadside embankments to attract frugivorous birds and to enhance seed dispersal mediated by birds from the surrounding landscape. We examined: (1) bird species richness and abundance; (2) patterns of avian spatial distribution within embankments and (3) seed dispersal mediated by birds. Bird richness and abundance did not differ between embankments with and without plantings. However, birds were not distributed randomly within embankments, with levels of species richness and abundance for facultative frugivorous between 4.8–8 times higher in areas closer to plantings. An analysis of bird droppings showed that birds only dispersed seeds of the planted species since no seeds of woody plants from matrices were detected. These results suggest that plantings acted as selective bird attractors, providing food and perches for frugivorous species. Nevertheless, the scarcity of seed-dispersing birds in the surrounding agricultural landscape prevented plantings from enhancing seed dispersal and connectivity to adjacent habitat.

No abstract