No abstract

MotivationThe BioTIME database contains raw data on species identities and abundances in ecological assemblages through time. These data enable users to calculate temporal trends in biodiversity within and amongst assemblages using a broad range of metrics. BioTIME is being developed as a community‐led open‐source database of biodiversity time series. Our goal is to accelerate and facilitate quantitative analysis of temporal patterns of biodiversity in the Anthropocene.Main types of variables includedThe database contains 8,777,413 species abundance records, from assemblages consistently sampled for a minimum of 2 years, which need not necessarily be consecutive. In addition, the database contains metadata relating to sampling methodology and contextual information about each record.Spatial location and grainBioTIME is a global database of 547,161 unique sampling locations spanning the marine, freshwater and terrestrial realms. Grain size varies across datasets from 0.0000000158 km2 (158 cm2) to 100 km2 (1,000,000,000,000 cm2).Time period and grainBioTIME records span from 1874 to 2016. The minimal temporal grain across all datasets in BioTIME is a year.Major taxa and level of measurementBioTIME includes data from 44,440 species across the plant and animal kingdoms, ranging from plants, plankton and terrestrial invertebrates to small and large vertebrates.Software format.csv and .SQL.

Monitoring is a key step for achieving restoration success. Despite increasing advances for selecting ecological indicators, monitoring sampling designs are not always available. We investigated how tree richness and the most used forest structure indicators vary spatially in restoration sites, aiming to provide evidence-based guidance for future monitoring protocols. We collected data from eight forest restoration sites covering overall 1,000 ha in four Brazilian Atlantic Forest regions. Canopy cover, tree density, vegetation height, and species richness were assessed in 18.2 ha of plots ranging from 60 to 300 m 2 in size, in restoration sites aged 1 to 5 years old. Using resampling techniques, we calculated the sampling error for the indicators and compared them with original sampling results, and then estimated the number of plots needed to reach a 20% sampling error. The ecological indicators assessed showed high variability among restoration sites. Canopy cover and height required less plots to reach the targeted sampling error than density of trees. The number of species does not stabilize even when more than 90% of the total number of plots was resampled, indicating high spatial variation. The use of the sampling error approach for defining how much to monitor, associated to appropriate sampling methods, could increase the reliability of monitoring. In addition, they will reduce operational costs, thus providing a key contribution to the effectiveness of largescale restoration programs expected to be implemented globally in the coming years. Thus, we recommend the incorporation of this in the forest restoration monitoring protocols being discussed worldwide.