Comparisons of biodiversity metrics reveal dissimilarities in biodiversity hotspot identification

doi:10.21203/rs.3.rs-4522774/v1

Download PDF

Analysis

Comparisons of biodiversity metrics reveal dissimilarities in biodiversity hotspot identification

https://doi.org/10.21203/rs.3.rs-4522774/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

As global climate change impacts ecosystems, establishing conservation priorities is crucial for managing threatened areas with limited resources. Biodiversity hotspots, typically defined by high degrees of endemism, play a key role in conservation. However, traditional hotspots may not capture the full extent of biodiversity, including functional and phylogenetic biodiversity or biodiversity incorporating traditional ecological knowledge (TEK). This study compares biodiversity hotspots identified by 17 diversity indices in the Pacific Northwest, USA, using data from 318 plant species. We consider species richness, phylogenetic diversity, functional diversity, and TEK-based diversity. Using simulated plant communities, we assess whether indices identify the same biodiversity hotspots. We find biodiversity metrics form two groups based on shared hotspot identification, suggesting single metrics may overlook other forms of biodiversity. Interestingly, TEK metrics cluster with some traditional indices, including species richness. This work offers new insights on integrating biodiversity measures for discerning regional biodiversity hotspots and conservation priorities.

Biological sciences/Ecology/Biodiversity

Earth and environmental sciences/Ecology/Conservation biology

biodiversity hotspots

Traditional Ecological Knowledge

functional diversity

phylogenetic diversity

In an era of global climate change, habitat loss, and species extinction, conservation strategies are crucial for minimizing global biodiversity loss. Identifying global areas of primary concern allows for prioritizing biodiversity conservation efforts as time and financial resources are typically limiting factors (Brooks et al. 2006). To this end, biodiversity hotspots are typically suggested as conservation priorities, which include habitats containing high levels of endemism, and are under high human threat (Myers et al. 2000). This type of conservation priority has alone mobilized millions of dollars to fund conservation, marking major increases in conservation investment in regions containing biodiversity hotspots (Brooks et al. 2006). However, this criterion for determining conservation priorities may overlook geographic regions harboring biodiversity that lies outside of the traditional criteria including measures of abundance-based taxonomic diversity, phylogenetic diversity (PD), functional diversity (FD); (Pollock et al. 2017), and socio-ecological values from traditional ecological knowledge (TEK); (Robinson et al. 2021). Incorporating a more comprehensive set of biodiversity dimensions may provide flexibility in decisions regarding how to measure biodiversity and allow managers and policymakers to refine conservation priorities.

Two biodiversity facets, PD, the cumulative evolutionary history of species, and FD, the diversity and distribution of functional traits within a set of species, have enabled transformational insights into fundamental questions in ecology and biogeography (Lamanna et al. 2014). Functional traits are morpho-physio-phenological properties of species that directly impact their fitness and, when aggregated to the community level, provide information about an ecosystem's stability and resilience to disturbances (Funk et al. 2017). Traits are often conserved along evolutionary lines, allowing phylogenies to serve as a practical substitute for understanding functional composition in the absence of specific trait information (Cadotte and Jonathan Davies 2010), but see (Venail et al. 2015). Additionally, phylogenetic diversity reflects evolutionary potential, which may be especially important to consider in times of rapid environmental change, such as the ongoing impact of climate change on ecosystems (Forest et al. 2007). However, functionally or phylogenetically diverse regions may not align with the current conservation priorities focusing on classically defined biodiversity hotspots, potentially leading to regions with outstanding and unique PD and FD being overlooked as global hotspots (Tietje et al. 2023).

Overall, conservation priorities derived from taxonomic and phylogenetic data are more similar to each other than those derived from functional trait data (Pollock et al. 2017). As a result, biodiversity hotspots defined by FD do not necessarily align with regions that include communities with high species richness or evenness (Stuart-Smith et al. 2013). The functional traits of species play a key role in influencing ecosystem properties (de Bello et al. 2010) and FD can explain variation in ecosystem function even when richness does not (Cadotte et al. 2011). Therefore, considering both PD and FD in conservation prioritization is important given that conservation efforts uphold not only species presence and abundance but also ecosystem functions, which are contingent on the traits expressed by species.

An overlooked measure of biodiversity is the knowledge of ecological communities and interactions held by local Indigenous communities. Traditional Indigenous territories contain a disproportionate amount of biodiversity: while these territories comprise only 22% of the world's land surface, they hold 80% of the planet's biodiversity (Sobrevila 2010). Successful conservation of natural resources requires collaborative efforts on a large scale, transcending ownership boundaries and involving those whose activities significantly impact the ecosystem (Charnley et al. 2007). In addition to the loss of land tenure due to colonization (Redvers et al. 2020), Indigenous knowledge has been historically disregarded when planning areas for conservation; however, in recent decades, there has been an increased awareness of the need to engage with Indigenous communities for conservation efforts, management, and policy decisions (Wheeler et al. 2020). Indigenous perspectives often see nature as an intricate, interconnected web of relationships rather than as a set of isolated, unrelated objects (Salmón 2000) and Indigenous land stewardship principles are often rooted in a deep understanding of local ecosystems (Tengö et al. 2014), a concept referred to as traditional ecological knowledge (TEK) (Velázquez-Rosas et al. 2018). Incorporating TEK into conservation policies, and decision-making has several advantages over utilizing information gathered only from Western science practices, including access to location-specific knowledge, increased knowledge of environmental connections, and local capacity building (Drew, 2005). There are a variety of metrics that come from traditional knowledge that can be invaluable to measuring biodiversity, including (but not limited to): folk taxonomy and systematics, traditional phenological knowledge, cultural keystone indicator species, socio-economic importance/uses, and linguistic diversity (Berlin 1973; Drew 2005). Name diversity is a measure of Indigenous naming systems that refers to the names and number of names associated with an organism. A single species may be associated with multiple names when there are different names for various plant tissues, life stages, and morphs. Similarly, the number of identified uses for a single plant can be quite large as different plant tissues (such as the leaf, fruit, or roots) may be used for a variety of practices, such as medicine, food, clothing dye, or building (Ens et al. 2023). Despite the value of TEK, to our understanding, species name and number of uses are not often considered when designating biodiversity hotspots or in combination with other metrics (although see, Gorenflo et al. 2012).

Incorporating various perspectives relating to biodiversity has can guide efforts in effective resource management when prioritizing habitats for conservation (Wilder et al. 2016). Hence, comparing TEK-based biodiversity measures with Western science-based biodiversity measures, such as those based on taxonomy, phylogeny, and functional traits, could offer new perspectives on biodiversity hotspot identification and establishing conservation priorities. In this study, we address two key questions: (i) Do various biodiversity metrics differ in the hotspots they identify? (ii) Do the hotspots identified from TEK-based metrics align with the hotspots identified by classically utilized Western science-based biodiversity metrics? Hotspots determined from taxonomic or phylogenetic information might be expected to be more similar to each other than they are to functional hotspots (Pollock et al. 2017). Additionally, we expect that hotspots identified from TEK-based measures will differ from the hotspots identified by other biodiversity indices (Tobes et al. 2022).

To explore these questions, we focused on a region renowned for its biodiversity and large occupation by Indigenous people, the Pacific Northwest (PNW), USA, for which a sizable repository of traditional ecological knowledge has been made publicly available with consent from local Indigenous communities (Turner 2014). Our analysis encompassed a range of biodiversity measures, including species richness, phylogenetic diversity, functional diversity, and TEK-based metrics of diversity.

Properties of simulated patches

We measured 17 different biodiversity indices across 100,000 simulated plant communities (100 simulations of 1000 plant communities) for our final dataset comprising 318 plant species. Within a single simulation, the distributions of biodiversity index values varied substantially across patches (Figure 1). Some biodiversity indices have unimodal, Gaussian-like distributions while others had a more dispersed distribution. Biodiversity metrics exhibiting a highly peaked distribution resulted in hotspot cutoffs (the top 5%) that were close to their means, which varied substantially among indices. This was the case for several metrics using functional or phylogenetic species data but not for species richness, functional species richness, or the TEK biodiversity indices.

Variation in identified hotspots

Across the 100 simulations run, biodiversity indices resulted in a consistent number of hotspots identified for some biodiversity indices while others had high variability (Figure 2; Table S2). The mean number of identified hotspots from various biodiversity indices ranged from 48.08 to 50 with a standard deviation ranging from 0 to 0.25 (shown for each biodiversity index in Table S2). The largest variation in the number of hotspots in the top 5% was seen in the functional biodiversity metrics (Figure 2).

Comparing biodiversity hotspots

We calculated precision to compare the hotspots that were identified for each biodiversity index. Pairwise precision values were largely symmetric (Figure S3). For example, comparing the ‘Indigenous names’ and ‘Recorded Indigenous uses’, irrespective of which is considered the baseline index, precision values are roughly the same (0.69 and 0.7; Figure S3A). When the analysis was performed using true positive rate (also known as recall or sensitivity) as the similarity classifier, we found that pairwise precision values were again largely symmetric (Figure S3B).

When assessing all indices across all patches and simulations, we found that the 17 biodiversity indices used here can be classified into two groups based on the similarity of the patches identified as hotspots (Figure 3). Group 1 includes both TEK metrics, species richness, functional species richness and five phylogenetic biodiversity metrics. The second group includes the remaining functional and phylogenetic biodiversity metrics. Many of the indices between the two groups never identified the same hotspots, whether measured through precision or sensitivity. In fact, indices in Groups 1 or 2 never overlapped in their identified biodiversity hotspots by more than 11% (Figure S3).

While many of the functional diversity metrics cluster into group 2, the precision among pairs of these metrics is relatively low with only Rao’s entropy and functional species dispersion having a precision value exceeding 0.4. These four functional diversity metrics share even fewer identified hotspots with functional species richness, which clusters with group 1.

The phylogenetic diversity metrics display the most variation, with Shannon, Menhinick, Faith, Margalef, and phylogenetic species richness in group 1, and McIntosh, phylogenetic species variability, Simpson, and Gini and Simpson in group 2. Several of these metrics clustered strongly with one another (precision values exceeding 0.70), indicating that they often identify the same hotspots. For example, Simpson and Gini and Simpson had a comparison index of 1, meaning that in every simulation, these two indices identify the same hotspots. Notably, the highest amount of overlap in the similarity of identified hotspots for phylogenetic metrics between groups 1 and 2 is 11% (Simpson or Gini and Simpson and Shannon’s diversity index).

The two TEK metrics, number of Indigenous names and number of recorded Indigenous uses cluster closely together within group 1; and have high similarity to functional species richness, species richness and phylogenetic species richness. Notably, the TEK-based metrics never align with the hotspots identified from phylogenetic species variability, or with the following functional metrics: functional species evenness, functional species divergence, Rao’s entropy (Q), and functional species dispersion.

We investigated how various biodiversity indices identify hotspots using plant species found in the Pacific Northwest region of North America. This region is ecologically diverse and has been inhabited by many Indigenous communities for millennia, with a rich and publicly available data repository on ethnobotanically important plant species. We clustered the biodiversity indices into groups based on the similarity in the hotspots they identify, shedding light on the patterns of how these biodiversity indices determine regional biodiversity hotspots. Generally, we find that the TEK metrics (number of Indigenous uses and number of Indigenous names) are clustered with some commonly used biodiversity indices including species richness, phylogenetic species richness, and functional species richness. Most functional trait metrics cluster together, except functional species richness; interestingly, phylogenetic biodiversity metrics exhibit substantial differences in identified hotspots, suggesting unique contributions to conservation prioritization.

Functional and phylogenetic hotspot comparisons

Many of the indices used to measure functional diversity did not identify the same hotspots as species richness, one of the most widely used measures of biodiversity. In fact, functional species richness is the only functional diversity metric that has a comparison index with species richness exceeding zero. Other work on biodiversity indices has reported similar patterns, where measures of species richness and functional richness more closely align with each other in measurements of global biodiversity, whereas other functional diversity metrics show distinct patterns of biodiversity (Safi et al. 2011; Stuart-Smith et al. 2013).

Unsurprisingly, several biodiversity indices cluster tightly together due to similarities in their calculation. For example, Simpson’s diversity index, and Gini and Simpson’s diversity index, which are complements of one another (Gini 1912; Simpson 1949), identified the exact same hotspots in every simulation.

All measures of functional diversity cluster together with the exception of functional species richness, which clusters with group 2. This suggests that, for the purposes of biodiversity hotspot identification, one functional biodiversity index is as good as another. For example, conservation practitioners might choose to use Rao’s entropy (Q) over functional species dispersion based on available resources and time, since both will identify similar hotspots. Importantly, even though these metrics cluster together, they still may not necessarily identify the same biodiversity hotspots. For instance, while McIntosh’s diversity index is nestled among functional diversity metrics, the highest comparison index it has is with functional species divergence, at 0.16- indicating that only 16% of the hotspots these metrics identify are the same.

Some of our results are supported by previous studies which found that phylogenetic diversity metrics and functional diversity metrics do not often correlate (Pollock et al. 2017). Species richness and phylogenetic metrics of biodiversity have shown moderate levels of correlation in previous studies, which is corroborated by the variability in the hotspots identified between species richness and phylogenetic diversity in this study (Devictor et al. 2010; Tietje et al. 2023). This highlights the importance of understanding phylogenetic diversity and its implications for conservation management, as choosing particular metrics for measuring phylogenetic diversity may identify different types of biomes that are important for the conservation of global biodiversity (Tietje et al. 2023).

TEK hotspot comparisons

Initially, we expected that the TEK metrics would align most closely with the functional diversity metrics. We hypothesized that species with more phenotypic variation and distinct uses would garner more individual name synonyms (i.e. when different parts of the plant have different uses and different names). Therefore a community of plants with a large number of uses may be expected to display greater functional diversity (Armstrong et al. 2021). However, we found that TEK metrics had minimal or no overlap with functional metrics when utilizing either precision or sensitivity, and instead were most similar to species richness and measures of phylogenetic diversity. Functional species richness, the total volume of functional space filled by a community (Villéger et al. 2008), was the only functional metric that identified any of the same hotspots as the TEK metrics possibly because the total number of names and uses for the plants in a community might also reliably estimate the range of functions of those plants.

Knowledge shared by Indigenous peoples has resulted in several biological insights and conservation contributions, such as the doubling of known fish aggregation and spawning sites (Mourão et al. 2006), and the designation of the Gladden Spit as a marine reserve for aggregations of mutton snappers and whale sharks (Drew 2005). In the Pacific Northwest, research has found increased taxonomic and functional plant diversity in regions that have historically been utilized by Indigenous peoples compared to neighboring managed or lesser utilized regions, with long lasting effects on plant diversity of over 150 years (Armstrong et al. 2021). Moreover, in this region there are integrative groups working to describe local ecological knowledge, with calls to study how this knowledge can and is being actively implemented in conservation management (Charnley et al. 2007).

Conclusions

The general applicability of these results is limited by assumptions necessitated by our modeling framework, which can be extended in future work to address these limitations. First, our hotspot definition led to variation in the number of hotspots identified both across simulations and across biodiversity metrics. As a consequence of calling the top 5% most diverse patches for every community a “biodiversity hotspot”, there are instances of tie-breaking involved which can lead to fewer than 50 (5% of 1000 communities) hotspots identified. Second, the geographic scale on which biodiversity metrics are applied is important and measuring biodiversity at a local, regional, or global scale is known to generate markedly different results (Crawley and Harral 2001), In the study reported here, we did not incorporate a parameter for scale, which is critical for applications of conservation and restoration biodiversity (Daily et al. 2003). Future work can consider the effect of scale, using more local (finer) or regional (coarser) data.

Finally, we chose to work with a dataset from the Pacific Northwest because this region had one of the most well-developed and ethically obtained datasets for our chosen TEK metrics (Charnley et al. 2007). Future work could consider a more holistic definition of traditional ecological knowledge to include “stories, songs, folklore, proverbs, cultural values, beliefs, rituals, community laws, local language, and agricultural practices, including the development of plant species and animal breeds” (Center for Biological Diversity 2021). Globally, scientists and managers should work to build, improve, and sustain collaborations with local and Indigenous communities to produce similar regional datasets and libraries, taking special care to preserve the intellectual property rights of communities (Ens et al. 2015).

Identifying biodiversity hotspots is key to the efficient allocation of resources to preserve ecosystems across the planet. This study shows that the choice of a biodiversity metric for identifying hotspots is not neutral: different metrics identify very different hotspots. However, this work also highlights key similarities among metrics that can be used to decide how to measure biodiversity cost-effectively. Biodiversity metrics utilizing TEK can recapitulate the results of more familiar metrics like phylogenetic species richness and species richness while incorporating very different types of data. Biodiversity researchers should continue to make use of a wide variety of metrics while taking care to recognize that ecological communities look differently through different lenses of biodiversity (Hanspach et al. 2020). This work offers novel perspectives on incorporating diverse biodiversity metrics, including Indigenous knowledge, to develop effective conservation strategies in ecologically diverse and culturally important ecoregions.

Data curation

For this study, we curated data on the number of Indigenous plant names and uses, and plant functional trait values for native plant species in the PNW. This geographic focal area was chosen because of the availability of multiple TEK sources and sufficient coverage of trait data in existing databases in this region (see further); although we acknowledge that TEK in other regions is well-documented and publicly shared with consent of local people, i.e., (Jackson and Jain 2006) in Central Asia, the Rizoma database of native plants in Chile (Cordero et al. 2022), the Ewé ethnobotanical database in Brazil (do Nascimento Fernandes de Souza and Hawkins 2020), and (Ens et al. 2020) in Australia).

Traditional Ecological Knowledge data

We obtained the number of distinct names for plant species from Indigenous communities and language groups from the Names of Native Plant Species in Indigenous Languages of Northwestern North America, Appendix 2B in the book Ancient Pathways, Ancestral Knowledge. Ethnobotany and Ecological Wisdom of Indigenous Peoples of Northwestern North America (Turner 2014). Over 300 native plant species, as well as subspecies and varieties, are listed in this appendix, and for each entry, the species names in the Indigenous languages and major dialects are listed. This list is by no means exhaustive but is the most comprehensive list of native species (plant or animal) with their documented names in local Indigenous languages that we were able to identify and access. The number of traditional uses associated with each plant species was obtained from the Native American Ethnobotany Database (Moerman 1998) which stores plant-derived uses (foods, drugs, dyes, and fibers) of Native American Peoples. This database has been curated over the past 25 years and includes uses from close to 300 Native American groups and over 4,000 plant species across North America. Due to the limited availability of published TEK databases, we chose to focus on plants in the PNW region to utilize the above-mentioned resources. This data collection produced an initial list of 330 plant species.

Accounting for botanical name synonyms and TEK data properties

The dataset of 330 plant species included 28 redundant entries due to synonyms, subspecies, and varieties. Species in this dataset were assigned formal binomial nomenclature using the taxize package in R and a global names resolver using The Leipzig Catalogue of Vascular Plants as the data source (Chamberlain et al. 2020). Each entry in our dataset was assigned the best match from gnr_resolve in the same R package using the previously listed databases. To facilitate linking TEK data to information about traits (see further), we reduced subspecies and varieties to their basic species names in case no synonym was found with this approach. The synonymization process led to unambiguous results for all but four species in the dataset. These included Cornus unalaschkensis which had 6 recorded uses while its synonym Cornus canadensis had 40 uses. Additionally, Platanthera stricta had 1 recorded use while its synonym Platanthera dilatata had 2 uses. This difference in the number of recorded uses may be due to the two plants being considered distinct species at the time of recording. Therefore, we summed the number of unique uses for these pairs of species and assigned that number to the synonym. After accounting for all synonyms, our species list was refined to 318 plant species.

The number of Indigenous names and uses had a left-skewed distribution and were highly correlated with one another (r= 0.44, p < 2.2e-16; Figure S2). Most plant species had fewer than 50 names and 100 uses, but several had over 100 names and 200 uses (Figure S1). The highest number of names was associated with Rubus spectabilis, which had a recorded 170 unique names, and several species had more than 150 recorded names. The highest number of Indigenous uses was reported for Thuja plicata, at 368 recorded Indigenous uses, with several species again having more than 150 recorded uses.

Functional trait data

We used the trait dataset compiled by (Díaz et al. 2022), which includes six vascular plant traits that define the primary axes of variation in plant form and function, and added complementary traits from the TRY database (Kattge et al. 2020). For species that were missing from the Díaz dataset, we extracted functional trait information from the TRY database (Kattge et al. 2020). Functional trait data included leaf dry matter content (LDMC; g/g), plant height (m), leaf nitrogen content per dry mass (N_mass; mg/g), leaf area (LA; mm²), plant woodiness (categorical), diaspore mass (SM; mg), stem specific density (SSD; mg/mm³), specific leaf area with petiole excluded (SLA; g.m²), leaf nitrogen content per leaf area (N_area; g/m^-2), and rooting depth (m). Due to variation between observations in the use of descriptive terminology for plant woodiness, we used a binary level of classification, labeling all species as either ‘woody’ or ‘not woody’. Additionally, all leaf area data for leaflet and leaf were combined into a single leaf area trait. Once this dataset was compiled for all species, we removed entries that were either not single observations or measures of central tendency (removing maximum and minimum estimates for the trait). We then summarized numerical trait values by calculating the mean of each trait for a given species across studies. Note that, because sample sizes are unavailable for values reported as averages in the TRY and Díaz et al. (2022) data sets, we treated average values as single observations in this averaging process.

The data coverage for our focal species varied substantially across the traits available in the TRY and Díaz data sets (Table S1). We discarded those traits with data coverage for less than 20% of the species list (i.e., stem specific density- 14% coverage; specific leaf area- 14% coverage; leaf nitrogen content per leaf area- 4% coverage; rooting depth- 3% coverage), leaving six traits for functional biodiversity analyses (Table S1). Leaf dry matter content, plant height, leaf nitrogen content per dry mass, leaf area, woodiness, and diaspore mass were retained for further analyses. Because the retained traits still showed data incompleteness, we used the missForest algorithm from the missForest R package to impute missing trait values with 1000 maximum iterations and 1000 trees per forest (Stekhoven and Bühlmann 2012). To reduce the effects of over-fitting, taxonomic data were not included in the fitting process. The missForest algorithm can impute categorical and continuous variables and provides estimates of out-of-bag error for both. The following variables were log-transformed because they skewed towards zero: leaf area, plant height, and diaspore mass. We imputed trait values with and without phylogenetic information, but the eventual imputed values were not substantially different. We performed qualitative diagnostics of the trait imputation by comparing the distribution of the original trait values across species with the distribution of imputed trait values. We further considered the out-of-bag RMSE for continuous variables and PFC for categorical variables to compare different settings for the imputation algorithm. Overall, the following parameters were used in the missForest function: maxiter = 100 and ntree = 1000.

The data coverage for our focal species varied substantially across the traits available in the TRY and Díaz data sets (Table S1). Leaf dry matter content, plant height, leaf nitrogen content per dry mass, leaf area, woodiness, and diaspore mass were retained for further analyses. Plant height, leaf area, and woodiness are three key traits that best predict positive plant responses to environmental threats (Kühn et al. 2021). Leaf dry matter content and leaf nitrogen content per dry mass are strongly connected to rates of growth and photosynthesis (Pérez-Harguindeguy et al. 2013). Diaspore mass was included because of its high coverage and potential positive association with establishment success and competition (Weiher et al. 1999). The imputation process had an acceptable out-of-bag error and visual comparison of the trait distributions before and after imputation indicated an appropriate fit (Table S1). Functional traits also had varied distributions, with leaf dry matter content and plant woodiness having bimodal distributions (as expected for plant woodiness which was scored only as 0 or 1); while leaf area, plant height, and leaf nitrogen content/dry mass were unimodal (Figure S1).

Species occurrence data

While the PNW has no official boundaries, we defined the area as the region described in Figure 1-1 of Turner (2014). The ecoregions of the PNW used in this study were determined from the Commission for Environmental Cooperation North American Environmental Atlas map of Terrestrial Ecoregions, Level II (Commission for Environmental Cooperation 1997). Finally, to determine in which Level II ecoregions the species in our main dataset occur, we used the BIEN_occurrence_species function from the BIEN package in R (Maitner et al. 2018). We determined all of the ecoregion types in which our species occur, and, for each ecoregion type, derived lists of our species found to occur there. We measured the number of species from our PNW dataset that occur at each Level II ecoregion type to provide an upper bound on the number of species in a simulated community for a given ecoregion type.

Simulating plant communities

We created a total of 100,000 simulated communities of plant species by sampling species from the curated species list. These communities are by no means representative of real plant assemblies, but to increase the probability that species in simulated communities also co-occur in natural conditions, we only combined species that grow under similar environmental conditions (as represented by ecoregions). Each community was randomly assigned an ecoregion type, L, from the list of Level II ecoregion types identified in the Pacific Northwest. The number of species assigned to a community, M, was randomly sampled from a uniform distribution between 1 and the maximum number of species in our data set for the given ecoregion type determined from the BIEN database. To assign the set of species to a patch at ecoregion type L with M species, we sampled M species without replacement from the species list for ecoregion type L. By doing this we assumed equal abundances of species within a community. Each simulation produced a set of 1000 communities representing an ecoregion within the PNW. We ran the simulation 100 times, creating a randomized set of communities each time, and calculated biodiversity metrics for each community.

Calculating dimensions of biodiversity

Because we did not simulate species abundances,species richness was the sole measure of taxonomic biodiversity calculated for this dataset. Species richness was calculated directly as the number of unique species present within a community. All biodiversity indices used in this study are named, defined, and cited in Table 1.

Phylogenetic diversity

For calculating phylogenetic diversity, we generated a phylogenetic tree for the species list, using the V.PhyloMaker2 R package (Jin and Qian 2022). As the backbone mega-tree, we used the “GBOTB.extended.WP.tre” mega-tree, which combines the GBOTB phylogeny for seed plants (Smith and Brown 2018) and the clade of pteridophytes from the phylogeny of Zanne et al. 2014, and is standardized following the botanical nomenclature of the World Plants database (Hassler 2024). Tips for a new genus were bound to this mega-tree using “scenario 3”, which is the most commonly used scenario according to Jin and Qian (2022). One species, Pterospora andromedea failed to bind to the mega-tree, so was excluded from all community composition simulations. The mega-tree was subsequently pruned to the available species list, using the picante R package (Kembel et al. 2010). Next, we calculated three phylogenetic diversity indices for each simulated community assuming equal abundance of each species. These included Faith’s (1992) phylogenetic diversity index, phylogenetic species variability, and phylogenetic species richness, defined, respectively, as the sum of phylogenetic tree branch lengths connecting all species in the target grouping, the decrease in variance due to phylogenetic relatedness of a trait, and the number of species in a community multiplied by the communities phylogenetic species variability (Helmus et al. 2007). We also calculated phylogenetic versions of six biodiversity indices – Simpson, Gini & Simpson, McIntosh, Shannon, Menhinick, and Margalef – which replace the usual species abundances with the abundances of phylogenetic tree branches occurring within a community (Pavoine 2020). Further details on these indices can be found in the Supplementary Material.

Functional diversity

We used the trait dataset to calculate functional diversity (i.e., functional richness, functional divergence, functional dispersion, functional species evenness, and Rao’s entropy) of each simulated community, assuming equal abundance of each species, using the fundiversity R package (Grenié and Gruson 2023). While typically these functional diversity metrics are applied to abundance data, they are also applicable to presence/absence data (Villéger et al. 2008). To moderate the weight of species with extreme trait values, leaf area, leaf dry matter content, plant height, and diaspore mass were log-transformed. All traits were standardized to zero mean and unit variance.

Diversity based on Traditional Ecological Knowledge

TEK-based biodiversity indices measured the number of unique names by local Indigenous communities and identified uses for the species present in a simulated community. To derive these TEK-based metrics, we simply added all the unique names and identified uses for species in a given community.

Identifying and comparing hotspots

Following (Sabatini et al. 2022), hotspots were defined to be all the simulated communities landing in the upper 95% quantile in the distribution of values for a given biodiversity index. We identified hotspots for all 100 runs of the simulation and across all biodiversity indices. To compare the hotspots identified for each biodiversity index, we computed the precision (or positive predictive value) of one biodiversity index compared to another. To do this, we assigned one index as the baseline (which we think of as the “true identifications”) and the other index as the comparison. The precision then measures the ability of the comparison index to identify the same hotspots as the baseline index. For the model, we let TP (the “true positives”) denote the number of hotspots identified by the baseline index and N (total number), denote the total number of hotspots identified by the comparison index. Then precision is defined as TP / N, the proportion of hotspots identified by the comparison index that are “correct” (also identified by the baseline index). If the comparison index identifies the same hotspots as the baseline index but also identifies many other hotspots (“false positives”), then precision will be lower because N > TP. While there are related methods of quantifying the similarity between binary classifications (such as sensitivity, specificity, or negative predictive value), we chose to use precision because we believe it best corresponds to the comparison process we are interested in. To evaluate whether our choice of classifier impacted our results, we repeated the analysis using the true positive rate (also known as recall or sensitivity, which measures the proportion of positive cases correctly identified as positive) in place of precision. We calculated precision and sensitivity for every possible pairing of biodiversity indices, allowing each to serve as a baseline and a comparison. To assess which biodiversity indices tend to identify the same hotspots (i.e. which measures cluster together), the biodiversity indices were clustered by Ward’s (1963) clustering criterion through the function hclust in the R package stats (R Core Team 2023).

Acknowledgments

The authors would like to thank Bela Loza and Abdel Holloway for contributions to the conceptualization of this manuscript at an early stage. This work is an outcome from discussions the co-authors had at a workshop sponsored by the U.S. National Science Foundation, and we acknowledge NSF for financial support in paying publication costs. KJMD was supported by NSF Award #2316455.

Conflict of Interests statement

The authors have no conflict of interests to declare.

Contributions statement

KJMD: Conceptualization, methodology, software, validation, formal analysis, investigation, data curation, writing - original draft, writing - review & editing, visualization, supervision, project administration, funding acquisition
EVC: Conceptualization, methodology, software, formal analysis, investigation, data curation, writing - review & editing, supervision, project administration

SA: Conceptualization, methodology, investigation, writing- original draft, writing- review & editing

KC: Conceptualization, investigation, methodology, data curation, writing- original draft, writing- review & editing

LRS: Conceptualization, methodology, investigation, data curation, writing- original draft, writing- review & editing, supervision, project administration, funding acquisition

Armstrong C, Miller J, McAlvay A, et al. 2021. Historical Indigenous land-use explains plant functional trait diversity. Ecol Soc.
Bello F de, Lavorel S, Díaz S, et al. 2010. Towards an assessment of multiple ecosystem processes and services via functional traits. Biodivers Conserv 19: 2873–93.
Brooks TM, Mittermeier RA, Fonseca GAB da, et al. 2006. Global biodiversity conservation priorities. Science 313: 58–61.
Cadotte MW, Carscadden K, and Mirotchnick N. 2011. Beyond species: functional diversity and the maintenance of ecological processes and services. J Appl Ecol 48: 1079–87.
Cadotte MW and Jonathan Davies T. 2010. Rarest of the rare: advances in combining evolutionary distinctiveness and scarcity to inform conservation at biogeographical scales. Divers Distrib 16: 376–85.
Center for Biological Diversity. 2021. Article 8(J) - Traditional Knowledge, Innovations and Practices: Introduction. https://www.cbd.int/traditional/intro.shtml. Viewed 22 Mar 2024.
Chamberlain SA and Szöcs E. 2013. taxize: taxonomic search and retrieval in R. F1000Res 2: 191.
Charnley S, Fischer AP, and Jones ET. 2007. Integrating traditional and local ecological knowledge into forest biodiversity conservation in the Pacific Northwest. For Ecol Manage 246: 14–28.
Commission for Environmental Cooperation. 1997. Ecological Regions of North America: Toward a Common Perspective. Montreal, Quebec: Commission for Environmental Cooperation Secretariat.
Cordero S, Abello L, and Gálvez F. 2022. Rizoma: a new comprehensive database on traditional uses of Chilean native plants. Biodivers Data J 10: e80002.
Crawley MJ and Harral JE. 2001. Scale dependence in plant biodiversity. Science 291: 864–8.
Daily GC, Ceballos G, Pacheco J, et al. 2003. Countryside biogeography of neotropical mammals: Conservation opportunities in agricultural landscapes of Costa Rica. Conserv Biol 17: 1814–26.
Devictor V, Mouillot D, Meynard C, et al. 2010. Spatial mismatch and congruence between taxonomic, phylogenetic and functional diversity: the need for integrative conservation strategies in a changing world. Ecol Lett 13: 1030–40.
Díaz S, Kattge J, Cornelissen JHC, et al. 2022. The global spectrum of plant form and function: enhanced species-level trait dataset. Sci Data 9: 755.
Dı́az S and Cabido M. 2001. Vive la différence: plant functional diversity matters to ecosystem processes. Trends Ecol Evol 16: 646–55.
do Nascimento Fernandes de Souza and Hawkins JA. 2020. Ewé: a web-based ethnobotanical database for storing and analysing data. Database 2020.
Drew JA. 2005. Use of Traditional Ecological Knowledge in Marine Conservation. Conserv Biol 19: 1286–93.
Ens EJ, Pert P, Clarke PA, et al. 2015. Indigenous biocultural knowledge in ecosystem science and management: Review and insight from Australia. Biol Conserv 181: 133–49.
Ens EJ, Rossetto M, and Costello O. 2023. Recognising Indigenous plant-use histories for inclusive biocultural restoration. Trends Ecol Evol 38: 896–8.
Ens E, Scott M, Daniels (OAM), et al. 2020. Cross-Cultural Guide to Some Animals and Plants of South East Arnhem Land. Batchelor Press Institute.
Forest F, Grenyer R, Rouget M, et al. 2007. Preserving the evolutionary potential of floras in biodiversity hotspots. Nature 445: 757–60.
Funk JL, Larson JE, Ames GM, et al. 2017. Revisiting the Holy Grail: using plant functional traits to understand ecological processes. Biol Rev Camb Philos Soc 92: 1156–73.
Gini C. 1912. Variabilità e mutabilità: contributo allo studio delle distribuzioni e delle relazioni statistiche. [Fasc. I.]. Tipogr. di P. Cuppini.
Gorenflo LJ, Romaine S, Mittermeier RA, and Walker-Painemilla K. 2012. Co-occurrence of linguistic and biological diversity in biodiversity hotspots and high biodiversity wilderness areas. Proc Natl Acad Sci U S A 109: 8032–7.
Grenié M and Gruson H. 2023. fundiversity: a modular R package to compute functional diversity indices. Ecography 2023.
Hanspach J, Jamila Haider L, Oteros‐Rozas E, et al. 2020. Biocultural approaches to sustainability: A systematic review of the scientific literature. People Nat (Hoboken).
Helmus MR, Bland TJ, Williams CK, and Ives AR. 2007. Phylogenetic measures of biodiversity. Am Nat 169: E68–83.
Hidayati S, Ghani BAA, Giridharan B, et al. 2018. Using Ethnotaxonomy to Assess Traditional Knowledge and Language Vitality. Ethnobiology Letters 9: 33–47.
Jackson R and Jain N. 2006. Mountain cultures, keystone species: exploring the role of cultural keystone species in Central Asia. Snow Leopard Conservancy.
Kattge J, Bönisch G, Díaz S, et al. 2020. TRY plant trait database - enhanced coverage and open access. Glob Chang Biol 26: 119–88.
Kühn N, Tovar C, Carretero J, et al. 2021. Globally important plant functional traits for coping with climate change. Frontiers of Biogeography 13.
Lamanna C, Blonder B, Violle C, et al. 2014. Functional trait space and the latitudinal diversity gradient. Proc Natl Acad Sci U S A 111: 13745–50.
Maitner BS, Boyle B, Casler N, and Condit R. 2018. The bien r package: A tool to access the Botanical Information and Ecology Network (BIEN) database. Methods Ecol Evol.
Moerman DE. 1998. Native American Ethnobotany Database. http://naeb.brit.org/. Viewed 28 May 2024.
Mourão JS, Araujo HFP, and Almeida FS. 2006. Ethnotaxonomy of mastofauna as practised by hunters of the municipality of Paulista, state of Paraíba-Brazil. J Ethnobiol Ethnomed 2: 19.
Myers N, Mittermeier RA, Mittermeier CG, et al. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853–8.
Pavoine S. 2020. adiv: An r package to analyse biodiversity in ecology. Methods Ecol Evol 11: 1106–12.
Pérez-Harguindeguy N, Díaz S, Garnier E, et al. 2013. New handbook for standardised measurement of plant functional traits worldwide. Aust J Bot 61: 167–234.
Poiani KA, Richter BD, Anderson MG, and Richter HE. 2000. Biodiversity Conservation at Multiple Scales: Functional Sites, Landscapes, and Networks. Bioscience 50: 133–46.
Pollock LJ, Thuiller W, and Jetz W. 2017. Large conservation gains possible for global biodiversity facets. Nature 546: 141–4.
Rangan H, Bell KL, Baum DA, et al. 2015. New genetic and linguistic analyses show ancient human influence on baobab evolution and distribution in Australia. PLoS One 10: e0119758.
R Core Team. 2023. R: A Language and Environment for Statistical Computing.
Redvers N, Poelina A, Schultz C, et al. 2020. Indigenous Natural and First Law in Planetary Health. Challenges 11: 29.
Robinson JM, Gellie N, MacCarthy D, et al. 2021. Traditional ecological knowledge in restoration ecology: a call to listen deeply, to engage with, and respect Indigenous voices. Restor Ecol 29.
Sabatini FM, Jiménez-Alfaro B, Jandt U, et al. 2022. Global patterns of vascular plant alpha diversity. Nat Commun 13: 4683.
Salmón E. 2000. Kincentric Ecology: Indigenous Perceptions of the Human-Nature Relationship. Ecol Appl 10: 1327–32.
Simpson EH. 1949. Measurement of Diversity. Nature 163: 688–688.
Smith SA and Brown JW. 2018. Constructing a broadly inclusive seed plant phylogeny. Am J Bot 105: 302–14.
Sobrevila C. 2010. The role of indigenous peoples in biodiversity conservation : the natural but often forgotten partners. The World Bank.
Stekhoven DJ and Bühlmann P. 2012. MissForest--non-parametric missing value imputation for mixed-type data. Bioinformatics 28: 112–8.
Tengö M, Brondizio ES, Elmqvist T, et al. 2014. Connecting diverse knowledge systems for enhanced ecosystem governance: the multiple evidence base approach. Ambio 43: 579–91.
Tietje M, Antonelli A, Forest F, et al. 2023. Global hotspots of plant phylogenetic diversity. New Phytol 240: 1636–46.
Tobes I, Carrillo-Moreno C, Guarderas-Flores L, et al. 2022. Ethnoichthyology and Ethnotaxonomy of the Kichwa Indigenous People of Arawanu (Arajuno), in the Ecuadorian Amazon. Frontiers in Ecology and Evolution 10.
Turner NJ. 2014. Ancient pathways, ancestral knowledge: Ethnobotany and ecological wisdom of indigenous peoples of northwestern north America. Montréal, QC, Canada: McGill-Queen’s University Press.
van Andel, van ‘t Klooster, Quiroz D, et al. 2014. Local plant names reveal that enslaved Africans recognized substantial parts of the New World flora. Proceedings of the National Academy of Sciences 111: E5346–53.
Velázquez-Rosas N, Silva-Rivera E, Ruiz-Guerra B, et al. 2018. Traditional Ecological Knowledge as a tool for biocultural landscape restoration in northern Veracruz, Mexico: a case study in El Tajín region. Ecol Soc 23.
Venail P, Gross K, Oakley TH, et al. 2015. Species richness, but not phylogenetic diversity, influences community biomass production and temporal stability in a re‐examination of 16 grassland biodiversity studies. Funct Ecol 29: 615–26.
Villéger S, Mason NWH, and Mouillot D. 2008. New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology 89: 2290–301.
Weiher E, Werf A van der, Thompson K, et al. 1999. Challenging Theophrastus: A common core list of plant traits for functional ecology. J Veg Sci 10: 609–20.
Wheeler HC, Danielsen F, Fidel M, et al. 2020. The need for transformative changes in the use of Indigenous knowledge along with science for environmental decision‐making in the Arctic. People Nat (Hoboken) 2: 544–56.
Wilder BT, O’Meara C, Monti L, and Nabhan GP. 2016. The Importance of Indigenous Knowledge in Curbing the Loss of Language and Biodiversity. Bioscience 66: 499–509.
Zanne AE, Tank DC, Cornwell WK, et al. 2014. Three keys to the radiation of angiosperms into freezing environments. Nature 506: 89–92.

Table 1. Diversity indices calculated in this study, definitions, and citations for their first introduction and the specific implementation used in our study. See the Supplementary Material for further details on the calculation of these indices.

Diversity Index	Definition	Citation
Species Richness	The number of species within a defined region.	Krebs, C. J. (2000). Ecology: The Experimental Analysis of Distribution and Abundance (5th ed., p. 434). Benjamin Cummings.
Simpson’s Phylogenetic Diversity	The probability that two individuals sampled from a population at random will belong to the same species.	Simpson, E.H. (1949) Measurement of diversity. Nature, 163, 688.
Gini & Simpson’s Diversity*	The complement of the inverse of Simpson’s diversity index (1-1/D) (see above).	Gini, C. (1912) Variabilita e mutabilita. Studi economicoaguridici delle facoltta di giurizprudenza dell, Universite di Cagliari III, Parte II.
McIntosh’s Diversity*	Expresses the heterogeneity of a sample/community/population in geometric terms.	McIntosh, R.P. (1967) An index of diversity and the relation of certain concepts to diversity. Ecology, 48, 392–404
Shannon’s Diversity*	An estimate of species diversity incorporating the number of species in a given area (richness) and relative abundance (evenness).	Shannon, C.E. (1948) A mathematical theory of communication. Bell System technical journal, 27, 379–423, 623–656.
Menhinick’s Diversity*	The number of different species in a population divided by the square root of the number of individuals in the population.	Menhinick, E.F. (1964) A Comparison of Some Species-Individuals Diversity Indices Applied to Samples of Field Insects. Ecology, 45, 859–861
Faith’s Phylogenetic Diversity	The sum of the branch lengths of a phylogenetic tree connecting all species in the target assemblage.	Faith, D.P. (1992). Conservation evaluation and phylogenetic diversity. Biological Conservation, 61, 1–10
Margalef’s Diversity*	The number of equal species within a defined region relative to the amount of individuals	Margalef, R. (1972) Homage to Evelyn Hutchinson, or why is there an upper limit to diversity? Transactions of the Connecticut Academy of Arts and Sciences, 44, 211–235
Recorded number of Indigenous botanical uses (Ethnobotany)	The recorded foods, drugs, dyes, fibers, and religious uses of plants by Indigenous peoples.	Balick, M.J. Transforming Ethnobotany for the New Millennium. Ann. Mo. Bot. Gard. 1996, 83, 58.
Recorded number of Indigenous names (Ethnotaxonomy)	The names associated with all parts of a plant used by local Indigenous groups	Berlin, B., 1973. The relation of folk systematic to biological classification and nomenclature. Annual Review of Ecology and Systematics 4, 259–271 Berlin, B., 1992. Ethnobiological Classification: Principles of Categorization of Plants and Animals in Traditional Societies, first ed. Princeton University Press, Princeton.
Phylogenetic species variability (PSV)	Quantifies how phylogenetic relatedness decreases the variance of a hypothetical trait shared by all species in a community.	Helmus M.R.,Bland T.J.,Williams C.K.& Ives A.R. (2007) Phylogenetic measures of biodiversity. American Naturalist,169, E68-E83
Phylogenetic species richness	The number of species in a community, multiplied by the community’s PSV (see above).	Helmus M.R.,Bland T.J.,Williams C.K.& Ives A.R. (2007) Phylogenetic measures of biodiversity. American Naturalist,169, E68-E83
Functional species richness	The amount of functional space occupied by the species of the community.	Mouillot, D., et al. (2005) Functional regularity: a neglected aspect of functional diversity. Oecologia, 142, 353-359.
Functional species evenness	The distribution of abundance across functional trait space within a community.	Villéger, S., Mason, N. W. H., & Mouillot, D. (2008). New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology, 89(8), 2290–2301.
Functional species divergence	The degree to which abundance distribution in niche spaces maximizes divergence in functional characters within the community.	Mason, N. W., Mouillot, D., Lee, W. G., & Wilson, J. B. (2005). Functional richness, functional evenness and functional divergence: the primary components of functional diversity. Oikos, 111(1), 112-118.
Functional species dispersion	The mean distance in multidimensional trait space of individual species to the centroid of all species.	Laliberté, E. and P. Legendre (2010) A distance-based framework for measuring functional diversity from multiple traits. Ecology 91299:305
Rao’s entropy (Q)	A measure of diversity based on a proportion of the abundance of species and the dissimilarity among them.	Rao, C.R. (1982). Diversity and dissimilarity coefficients: a unified approach. Theoretical Population Biology, 21, 24:43.
*See the supplementary materials for definitions and citations of these metrics applied to phylogenetic diversity.

There is NO Competing Interest.

Supplementarymaterial.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Comparisons of biodiversity metrics reveal dissimilarities in biodiversity hotspot identification

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Materials and Methods

Declarations

References

Table 1

Additional Declarations

Supplementary Files

Status:

Version 1