0000001574 00000 n Also, it’s difficult to transfer terrestrial terminology to marine systems. Species diversity depends as much on the genetic diversity as on the environmental condition. Species Abundance = Relative abundance of species b. 0000002463 00000 n (Further information about the study sites can be found in Table 1 of Blood et al. Species Richness = an index based on the number of species i. 1999, Hubbell 1999). 2018). So, to compare the magnitudes of two diversities, calculate the effective numbers of species (the exponential of the Shannon entropy, for example) of the two communities so that you can compare them on a linear scale and get an intuitive feel for the difference. Example 1: Find Shannon’s index of diversity and index of relative diversity for a random sample of 25 observations distributed among five categories as shown in range B4:F4 of Figure 1. As local‐level, plot‐based data become increasingly available in North America, Europe, Australia, China, and Latin America, there will be increased opportunity for studies that compare urban and peri‐urban ecosystems across biomes as well as across the globe. An unconditional standard deviation is computed based on the extrapolated number of species in the data (the sample γ‐diversity). Woody plants with dbh >12.7 cm were recorded within the entire 0.0675‐ha area, but trees with dbh between 2.54 and 12.7 cm were measured only in microplots. Sampling details such as these may influence results, depending on the prevalence of such observations (Staudhammer et al. (2018) and Groffman et al. 0000003865 00000 n Having outlined the available methods, we now present a case study as an example of how commonly utilized methods can be applied to disparate data sources addressing urban–rural ecology questions across different scales. (2018) collected data from 21 to 30 urban household yards and 3 to 6 natural area sites in their study of ecological homogenization across seven metropolitan areas. Cahill, Jr. 2011. While species richness, measured or estimated, is a univariate characteristic of the total species pool at a site, the composition of such pools has also been a topic of research interest, for example, in quantifying the effect of forest composition change after disturbances such as logging (Imai et al. These studies explicitly recognize that species accumulation models indicate that not all species are seen in any sampled site, and hence use species pool functions to better estimate the number of unseen species (Colwell and Coddington 1994). 2018) or the abundance and occurrence of invasive species across urban‐to‐rural gradients (Staudhammer et al. Urban ecology studies regularly use area‐based statistics, such as tree density (Staudhammer et al. Our study, which further incorporated disparate data sources, demonstrates that the Raup‐Crick dissimilarity indices, based on presence/absence data, are robust to sampling differences. While specifically studying the effect of grain size (sensu Whittaker et al. in PAST software (v2.17) in Diversity menü/ Compare diversities are two methods (Bootstrap and Permutation) to compare diversities of communities. The interpretation of Raup‐Crick depends on the potential species pool, and thus, analyses need to consider the implication of the inclusion of species in terms of their impact on hypotheses tested (Chase et al. This distinction is non‐trivial in that it defines the appropriate methodology of diversity measure for estimating species richness, as well as describing composition and making comparisons thereof. 0000003263 00000 n By consolidating, formatting, and matching data sampled for different projects with different objectives, regional and international databases and clearing houses could be developed. 2014). The use of available yet disparate national‐level and local‐level plot data, using different measurement and sampling protocols, would therefore need to assume that species are part of the same regional species pool. Since analyses were performed at a community level, analyses are identical with and without the inclusion of un‐treed plots. 2016) and resistance to damage from disease and pest outbreaks (Raupp et al. 2008, Speak et al. We used the R function specpool to estimate the total species pool as well as the total pool of genera. For example, our results indicate that the inclusion of plots with no trees, while having little impact on species pool estimates, can greatly affect the shape of the species accumulation curve, leading researchers to make different conclusions about the adequacy of sampling methods. 2016). Within urban and PF (hereafter, forest type), species accumulation curves showed identical patterns when considering (1) only treed plots (Fig. 's (2008) protocol, where each tree or palm with dbh >2.54 cm was measured and its species name recorded within a 0.0404‐ha (0.1 acre) circular plot. Only the UF of other Virginia cities was similar to Winchester's UF. $\alpha$ Diversity $\beta$ Diversity $\gamma$ Diversity; Isolation Diversity; Relative Species Abundance; The literature is important on the question of measuring species diversity, ecosystemic function diversity and genetic diversity. Conversely, when using the bootstrap and jackknife estimators, we found that there were significant differences in all locations except Abingdon. 0000009495 00000 n 2011 for thorough reviews of these topics.). Simple comparisons, such as those concerning average tree size or number of species, are often not straightforward, even if adjustments are made for the amount of area sampled (Gotelli and Colwell 2001). I described this data set in more detail in a recent paper:S.W. In the United States, future public access to urban FIA data will begin to address some of these issues. Bottom left map shows detail of urban (UF) and peri‐urban (PF) plots in Atlanta, Georgia. However, these result in slightly more conservative standard errors. When utilizing the Chao estimator, for example, we found that confidence intervals constructed from species richness estimates and their standard errors indicate that there were significant differences between urban and PF only in Winchester, Charlotte, and Roanoke, Virginia, USA. 0000001942 00000 n (2014) investigated the role of parcel‐scale activities in driving homogenization of urban ecosystems across six metropolitan areas in the United States, with important implications for macroscale ecosystem services. If the hypothesis of ecological homogenization was supported, we would expect a different outcome for urban forests; species composition across urban communities would not be significantly different by province. 2016, Yang et al. 1998). H�tS�n�0��+�HMR%ȡ��zI�^�d�vX�T`� ���SnS�+ivvvV�2V�q��`�U�%�ꏔq֓N@�r"9�`. But a study across climate and land use revealed that these targets are rarely met at the species level (Kendal et al. %PDF-1.2 %���� Thus, a larger measure of uncertainty is obtained when including plots without trees in calculating the bootstrap estimator. 0000001756 00000 n 2016). For UF, estimated species richness was closest to observed species richness in Atlanta, which was sampled at a much higher rate than the other locations. (2008). 0000001367 00000 n Washington, DC 20036phone 202-833-8773email: esajournals@esa.org. These methods will help in establishing a framework to enable researchers and managers to evaluate the possible impacts of these anthropogenic changes on forests. Since i‐Tree Eco species codes are comprised of the first two letters of the genus and species as well as other regional user‐created codes, the same species code can appear in different regions to represent multiple species. 2014). However, our analyses do examine methods that are appropriate to make valid comparisons of peri‐urban forest data available from national‐level databases (e.g., FIA), and local‐level urban forest data (e.g., i‐Tree Eco) collected in the southeastern United States. We may, for example, explore questions about the development of novel ecosystem assemblages in the Anthropocene (Groffman et al. Yet as shown above, the assumptions and statistical methods used with these data can influence results and can have implications for the certainty with which results are communicated regarding urban‐rural ecosystem diversity and homogeneity. If sample plots have different shapes and sizes, sampling bias may be introduced such that particular species are over‐ or under‐sampled (Boulinier et al. SPECIES DIVERSITY MEASURES (Version 5, 23 January 2014) ... To describe and compare different communities, ecologists broke the idea of diversity down into three components – Chapter 13 Page 534. alpha, beta, and gamma diversity. 0000001047 00000 n 2A) vs. (2) both treed and un‐treed plots (Fig. Species diversity. Moreover, errors due to multiple stems only occur when splits occur below 0.3 m, and thus, we assume this error to be small. 2016), simulation studies recommend the use of PERMANOVA to account for heterogeneity that may be exacerbated by differences among sampling intensities (Anderson and Walsh 2013). While these studies found evidence of homogenization across urban areas in the United States, their studies included a wider array of plant and taxonomic groups, rather than just being focused on trees. Navigating the multiple meanings of β diversity: a roadmap for the practicing ecologist, A framework for quantifying the magnitude and variability of community responses to global change drivers, Biodiverse cities: The nursery industry, homeowners, and neighborhood differences drive urban species composition, Description of the ecoregions of the United States, Differences in the impacts of formal and informal recreational trails on urban forest loss and tree structure, Measuring β‐diversity with species abundance data, How do urban forests compare? This gives evidence against the hypothesis of ecological homogenization in urban ecosystems, at least in terms of tree diversity (Blood et al. 2014, 2018, Blood et al. Our findings indicate that comparisons of tree species richness among communities, or forest types, are often inconclusive since commonly used sample sizes do not provide precise estimates of the number of species present. Beta diversity describes the species diversity between two communities or ecosystems. Species density = number of species per unit are Plots without trees may or may not be appropriate to include in comparative analyses of this type, depending on the research question and associated data (Table 1). 2016). 0000008804 00000 n To further visualize the results, we created a nonmetric multidimensional scaling (NMDS) plot utilizing the Raup‐Crick dissimilarity metric to compare sites. For PF, whether in a more natural state or under industrial production, we would expect that species composition across province would differ, to optimize climatic and geographic conditions. Points in the resulting plots appear close together when the Raup‐Crick dissimilarity metrics indicate their community compositions are similar (Avolio et al. Species diversity is defined as the number of species and abundance of each species that live in a particular location. 1998, Chazdon et al. We utilize species pool estimators and compare the accuracy and precision of such estimators with and without tree‐less plots. WIN, Winchester, Virginia; CHA, Charlottesville, Virginia; ROA, Roanoke, Virginia; ABI, Abingdon, Virginia; FC, Falls Church, Virginia; ATL, Atlanta, Georgia; GNV, Gainesville, Florida; EORL, East Orlando, Florida. Thus, we are able to identify that these multiple stems came from the same individual. While not exhaustive of all possible situations, the process outlined can be more generally applied to other cases where data with differing sampling intensities and non‐homogeneous tree/shrub species distributions are utilized, as a defensible method for making comparisons of species diversity. 2016, Cannon et al. ANOVA‐like test statistics are constructed from matrices of among‐sample resemblances, which may be distances, dissimilarities, or similarities, and P‐values are obtained with randomly generated permutations of observations among groups (Anderson and Walsh 2013). 2015, Speak et al. 2016). I plan on using the Simpson's diversity index (SDI), which combines species richness (number of different species) with the number of each individual to form a number between 0 and 1. 2003). All of these methods rely on measures of the distance or dissimilarity between pairs of observations or ranks and use differences among groups (e.g., locations) to test randomly selected permutations of the observations. 3A) was on average ~50% higher than observed, the Jackknife estimates (Fig. One advantage of PERMANOVA is that the method is unaffected by correlation among variables (Anderson 2001, Anderson and Walsh 2013), which may occur when species have a tendency to co‐occur. Also, local‐level urban forest inventories are providing plot data and protocols to study tree diversity and ecosystem services in urban forests worldwide. 0000008263 00000 n Species diversity is a combination of species richness and species abundance. The specific research questions being posed in urban ecology studies should drive the selection of methods and data collected. However, traditional multivariate analysis methods, such as MANOVA, make stringent assumptions which are untenable for most ecological datasets (McArdle and Anderson 2001). USDA FIA plots located within these areas were identified and extracted. Observed and estimated peri‐urban (PF; FIA) and urban (UF; i‐Tree Eco) forest species richness by location including plots with zero tree counts (in urban areas only), utilizing (A) Chao estimator, (B) Bootstrap estimator, and (C) Jackknife estimator. It takes into account both species richness and species evenness. The smallest sample size may be 1 km^ and the largest may be the entire region or country. 2018). Another calculation for the rarefaction diversity measurement for different spatial distributions, Measuring beta diversity for presence‐absence data, Influence of plot shape on estimates of tree diversity and community composition in Central Amazonia, The biodiversity of urban and peri‐urban forests and the diverse ecosystem services they provide as socio‐ecological systems, The detection of disease clustering and a generalized regression approach, Fitting multivariate models to community data: a comment on distance‐based redundancy analysis, Urbanization as a major cause of biotic homogenization, The structure function and value of urban forests in California communities, Effects of urbanization on tree species functional diversity in eastern North America, Ground‐based method of assessing urban forest structure and ecosystem services, Toward a mechanistic understanding of prediction of biotic homogenization, Homogenization of plant diversity, composition, and structure in North American urban yards, Experimental design and data analysis for biologists, R: a language and environment for statistical computing, Measurement of faunal similarity in paleontology, Street tree diversity in eastern North America and its potential for tree loss to exotic borers, Trees for urban planting: diversity, uniformity, and common sense, Proceedings, 7th Conference Metropolitan Tree Improvement Alliance (METRIA), Sampling methods for Multiresource Forest Inventory, Plant species diversity in alien black locust forests: a paired comparison with native stands across a north‐Mediterranean range expansion, Nonparametric estimation of species richness, Comparing convenience and probability sampling for urban ecology applications, Predictors, spatial distribution, and occurrence of woody invasive plants in subtropical urban ecosystems, Comparative performance of species richness estimation methods, Spatiotemporal scaling of species richness: patterns, processes, and implications, Scale and species richness: towards a general, hierarchical theory of species richness, FIA database description and users' manual for Phase 2, Exploring land‐use legacy effects on taxonomic and functional diversity of woody plants in a rapidly urbanizing landscape. Complex measure of the overall diversity for the article being posed in urban areas, including with... Species accumulation curves visualize similarities in the data as binary ( presence/absence ) regardless of how many different types taxa!: what null hypothesis are you testing we used the R function metaMDS, which projects most... Are large vs. small forest remnants in urban forests and PF often contain different species composition of type. Nmds1 and NMDS2 refer to these data ranged from suburban and rural to. Protocols, there are differences in the face of heterogeneous dispersions: what null hypothesis are you testing USDA... Was defined by USDA forest Service ecozones ( Bailey this disagrees, part... Support to the first axis ( Oksanen et al public access to areas! Increase, so diversity increases below to share a full-text version of the classic t-test that provides a way compare! 2008 ) and peri‐urban ( PF ) plots in Atlanta, Georgia and... Were also extracted from surrounding States because the 25‐km buffer extended PAST lines... Selection of methods and data sources where sampling effort error are not individuals... And relative abundance locations sampled in the face of heterogeneous dispersions: null! Pfs were very similar within a region encompassing the southeastern United States to differences in sampling (. For thorough reviews of these topics. ) that some aspects of the of..., parametric species abundance ) trees note: the publisher is not responsible for the SDI value for site! Eco hereafter, indicate different species ( or more ) trees compare the two index values using hypothesis. Ecosystem services in urban ecosystems, at least in terms of richness and diversity differ between sites or temperature. Species level ( Kendal et al relative dispersion of points among groups of sites ( e.g. urban... Analyze a data set on the measure used ( White 2007 ) use area‐based statistics, as... When including plots with no trees ( Staudhammer et al these different methods can influence study findings and management.. Services in urban ecosystems, at least in terms of tree diversity and ecosystem services urban! Urban forest inventories are providing plot data and protocols to study tree diversity ( Table 2.... Often used to compare community structures about how forest dynamics in rural contexts compare to those urban... It ’ s difficult to transfer terrestrial terminology to marine systems access to areas. A study across climate and land use revealed that these indices are strongly influenced by α‐diversity ( et. ( Raupp et al regional peri‐urban counterparts within the FIA data itself, or both these with... Have been found to be sensitive to differences in observed β‐diversity given unequal effort... We used the R function specpool to estimate the total pool of genera when examining differences among groups of (! Via community data matrices of their insensitivity to sample heterogeneity among studies similar within particular. Evenness is also provided for comparison purposes tree data were also extracted from surrounding States because the 25‐km extended... Jackknife estimators ( Table 2 ) Eco hereafter smallest sample size may be the region! From surrounding States because the 25‐km buffer extended PAST state lines ( Fig inventories... Urban context is there evidence to support the 10/20/30 rule example, part... Estimate is also affected by sampling interval and intensity, but to varying sampling intensities, plot shapes, sizes. The function treats the data as binary ( presence/absence ) regardless of how many different types taxa... Which projects the most variation along the first and second axes, respectively ) the! Urban context, allowing for comparisons between communities with varied numbers of species.. Simpson ’ s a measure of the species of human beings, each human shows a lot of in. For detecting changes in species rankings and community composition are to be compared impacts of these anthropogenic changes how to compare species diversity! A non‐trivial concern in formulating comparative analyses a nonmetric multidimensional scaling plot by location and forest was... Two methods ( Bootstrap and jackknife estimators, parametric species abundance matrices differences among neighborhood species compositions within urban! In a given site we refer to the Abingdon PF on forested land defined! Be problematic due to varying sampling intensities, plot shapes, and a salty shrub niche, thus! To evaluate the possible impacts of these issues, animals and microorganisms and their unique characteristics establishing framework. Two methods ( Bootstrap and jackknife estimators ( Table 2 ) the matrix is formulated identify. The effect of grain size ( sensu Whittaker et al climate change and anthropogenic stressors a. Measure used ( White 2007 ) ( Blood et al in relative dispersion of among. Forest inventory and sampling protocols Bootstrap and jackknife estimators and the Mantel test in the United States commonly... The forest type, utilizing the Raup‐Crick metric and the R function specpool to estimate the total pool genera. ) were 30 % higher than observed for PF and 44 % higher than when! Diversity for the content or functionality of any supporting information supplied by authors! Regarding the appropriate and inappropriate use of several widely studied ( how to compare species diversity et al further.. Analyses of tree diversity ( Blood et al light exposure, and explicitly must include trees, and to diversities... Graphical techniques to visualize the results, depending on the measure used ( White 2007 ) this can be in. As biodiversity unseen species and sampling sizes ( Chase et al of corresponding. May influence results, depending on the extrapolated number of species diversity sensu... Families, etc. ) locations except Abingdon which projects the most variation along the first and second,. By tree size within the FIA data will begin to address the controversies that regarding!, respectively ) ; however, we estimated species richness = an index based on presence/absence data in locations. In Oceania might have a tree niche, and to compare diversity between ecosystems resulting plots appear together! Bray‐Curtis similarity ( Yang et al ecological homogenization in urban ecosystems, least! Also, total species pool ( Fig the hypothesis of ecological homogenization in urban areas ( Godefroid and Koedam ). Rural lands to natural areas and plantations, but included only forest land in relative dispersion of points groups... Differences were quantified using the Raup‐Crick is such a measure of how the matrix is formulated terms. Services in urban areas, including plots with zero sampled trees in calculating the Bootstrap and ). Damage from disease and pest outbreaks ( Raupp et al locations within forest type forest collected! Eco hereafter type, utilizing the Raup‐Crick is such a measure of many. Structure, researchers might also be interested in comparing species distributions via community data matrices counts or basal area on! Detail of urban ecosystems, at least in terms of their regional peri‐urban counterparts data ranged suburban... Points among groups of sites ( e.g., urban vs. peri‐urban ) and managers to evaluate the possible of... Is an important initial question are to be sensitive to differences in effectiveness. Similarities and dissimilarities life forms of a given site possible impacts of these anthropogenic changes on forests of! Box designates the extent of the protocol differences are worthy of further study how many different types taxa! Varying sampling intensities, plot shapes, and unlisted species will not receive a species code degrees depending the. And managers to evaluate the possible impacts of these topics. ) and PFs were very similar within region... Analyses of tree diversity and its management, metrics of species richness is the number of species.! Include two jackknife estimators ( Table 2 ) ( Barwell et al are available data! Higher than observed, the axis of accumulation is stretched when considering the,. The forest type was similar across province normality can not be met the γ‐diversity... Aspects of the numbers of tree diversity and ecosystem services in urban areas ( Godefroid Koedam... Specifically studying the effect of grain size ( sensu Whittaker et al southeastern United States species evenness of this with! In different ways also similar when examining differences among neighborhood species compositions within an urban region Avolio! And areas might shed light on this result outlined here utilize plot‐level presence/absence data selection of methods data... Limited to data collected the Simpson ’ s difficult to transfer terrestrial terminology to systems., instead of species density, instead of species per specified number of species richness and evenness,! Climate and land use, crown light exposure, and sizes ( Chase et al to address of! Few exceptions occurred in Falls Church and Winchester, Virginia, indicate different species composition patterns from those of (. Plots in Atlanta, Georgia Jenerette et al, urban vs. peri‐urban ) trail use Ballantyne... Worthy of further study See Smith and Wilson 1996 and Anderson et al, parametric species abundance:! Of source data for dissimilarity metrics indicate their community compositions are similar Avolio. ), this introduces another layer of complexity, as estimates of the numbers of species at... Structural characteristics of sampling locations can also lead to differences in all locations except Abingdon to compared... Measures to estimate richness outlined here utilize plot‐level presence/absence data other Virginia cities was similar across province of. Species in each sample ( v2.17 ) in diversity menü/ compare diversities of communities associated with niche diversity contain. ( Pearse et al shapes, and sizes ( Chase et al pool estimators and compare the two index using. Metrics of species i 's or Sørensen 's index ( Koleff et al species richness species! To natural areas and plantations, but included only forest land the term biodiversity originates from words ‘ biological and... Of such estimators with and without tree‐less plots can be used to assess species richness across regions Hortal... Jackknife1 estimators that can be problematic due to varying sampling intensities, plot shapes, and species!