How Many Species Are There on Earth and in the Ocean? Camilo Mora1,2*,Derek P.Tittensor1,3,4,Sina Adl1,Alastair G.B.Simpson1,Boris Worm1
波谲云诡的意思1Department of Biology,Dalhousie University,Halifax,Nova Scotia,Canada,2Department of Geography,University of Hawaii,Honolulu,Hawaii,United States of America, 3United Nations Environment Programme World Conrvation Monitoring Centre,Cambridge,United Kingdom,4Microsoft Rearch,Cambridge,United Kingdom
Abstract
The diversity of life is one of the most striking aspects of our planet;hence knowing how many species inhabit Earth is among the most fundamental questions in science.Yet the answer to this question remains enigmatic,as efforts to sample the world’s biodiversity to date have been limited and thus have precluded direct quantification of global species richness, and becau indirect estimates rely on assumptions that have proven highly controversial.Here we show that the higher taxonomic classification of ,the assignment of species to phylum,class,order,family,and genus)follows a consistent and predictable pattern from which the total number of species in a taxonomic group can be estimated.This approach was validated against well-known taxa,and when applied to all domains of
life,it predicts,8.7million(61.3 million SE)eukaryotic species globally,of which,2.2million(60.18million SE)are marine.In spite of250years of taxonomic classification and over1.2million species already catalogued in a central databa,our results suggest that some 86%of existing species on Earth and91%of species in the ocean still await description.Renewed interest in further exploration and taxonomy is required if this significant gap in our knowledge of life on Earth is to be clod.
Citation:Mora C,Tittensor DP,Adl S,Simpson AGB,Worm B(2011)How Many Species Are There on Earth and in the Ocean?PLoS Biol9(8):e1001127.
doi:10.1371/journal.pbio.1001127
Academic Editor:Georgina M.Mace,Imperial College London,United Kingdom
Received November12,2010;Accepted July13,2011;Published August23,2011
Copyright:ß2011Mora et al.This is an open-access article distributed under the terms of the Creative Commons Attribution Licen,which permits unrestricted u,distribution,and reproduction in any medium,provided the original author and source are credited.
Funding:Funding was provided by the Sloan Foundation through the Census of Marine Life Program,
Future of Marine Animal Populations project.The funders had no role in study design,data collection and analysis,decision to publish,or preparation of the manuscript.
Competing Interests:The authors have declared that no competing interests exist.
*E-mail:
Introduction
Robert May[1]recently noted that if aliens visited our planet, one of their first questions would be,‘‘How many distinct life forms—species—does your planet have?’’He also pointed out that we would be‘‘embarrasd’’by the uncertainty in our answer. This narrative illustrates the fundamental nature of knowing how many species there are on Earth,and our limited progress with this rearch topic thus far[1–4].Unfortunately,limited sampling of the world’s biodiversity to date has prevented a direct quantifi-cation of the number of species on Earth,while indirect estimates remain uncertain due to the u of controversial approaches(e detailed review of available methods,estimates,and limitations in Table1).Globally,our best approximation to the total number of species is bad on the opinion of taxonomic experts,who estimates range between3and100million species[1];although the estimations likely reprent the outer bounds of the total number of species,expert-opinion app
roaches have been ques-tioned due to their limited empirical basis[5]and subjectivity[5–6](Table1).Other studies have ud macroecological patterns and biodiversity ratios in novel ways to improve estimates of the total number of species(Table1),but veral of the underlying assumptions in the approaches have been the topic of sometimes heated controversy([3–17],Table1);moreover their overall predictions concern only specific groups,such as incts[9,18–19], deep a invertebrates[13],large organisms[6–7,10],animals[7], fungi[20],or plants[21].With the exception of a few extensively studied ,birds[22],fishes[23]),we are still remarkably uncertain as to how many species exist,highlighting a significant gap in our basic knowledge of life on Earth.Here we prent a quantitative method to estimate the global number of species in all domains of life.We report that the number of higher taxa,which is much more completely known than the total number of species [24],is strongly correlated to taxonomic rank[25]and that such a pattern allows the extrapolation of the global number of species for any kingdom of life(Figures1and2).
Higher taxonomy data have been previously ud to quantify species richness within specific areas by relating the number of species to the number of genera or families at well-sampled locations,and then using the resulting regression model to estimate the number of species at other locations for which the number of families or genera are better known than species richness(reviewed by Gaston
&Williams[24]).This method,however,relies on extrapolation of patterns from relatively small areas to estimate the number of species in other ,alpha diversity). Matching the spatial scale of this method to quantify the Earth’s total number of species would require knowing the richness of replicated planets;not an option as far as we know,although May’s aliens may disagree.Here we analyze higher taxonomic data using a different approach by asssing patterns across all taxonomic levels of major taxonomic groups.The existence of predictable patterns in the higher taxonomic classification of species allows prediction of the total number of species within taxonomic groups and may help to better constrain our estimates of global species richness.
Results
We compiled the full taxonomic classifications of,1.2million currently valid species from veral publicly accessible sources(e Materials and Methods).Among eukaryote‘‘kingdoms,’’asss-ment of the temporal accumulation curves of higher ,the牛肉拌面
人无信不立打一字
cumulative number of species,genera,orders,class,and phyla described over time)indicated that higher taxonomic ranks are much more completely described than lower levels,as shown by strongly asymptoting trajectories over time([24],Figure1A–1F, Figure S1).However,this is not the ca for prok
aryotes,where there is little indication of reaching an asymptote at any taxonomic level(Figure S1).For most eukaryotes,in contrast,the rate of discovery of new taxa has slowed along the taxonomic hierarchy, with clear signs of asymptotes for phyla(or‘‘divisions’’in botanical nomenclature)on one hand and a steady increa in the number of species on the other(Figure1A–1F,Figure S1).This prevents direct extrapolation of the number of species from species-accumulation curves[22,23]and highlights our current uncer-tainty regarding estimates of total species richness(Figure1F). However,the increasing completeness of higher taxonomic ranks could facilitate the estimation of the total number of species,if the former predicts the latter.We evaluated this hypothesis for all kingdoms of life on Earth.
First,we accounted for undiscovered higher taxa by fitting,for each taxonomic level from phylum to genus,asymptotic regression models to the temporal accumulation curves of higher taxa (Figure1A–1E)and using a formal multimodel averaging framework bad on Akaike’s Information Criterion[23]to predict the asymptotic number of taxa of each taxonomic level (dotted horizontal line in Figure1A–11E;e Materials and Methods for details).Secondly,the predicted number of taxa at each taxonomic rank down to genus was regresd against the numerical rank,and the fitted models ud to predict the number of species(Figure1G,Materials and Methods).We applied this approach t
o18taxonomic groups for which the total numbers of species are thought to be relatively well known.We found that this approach yields predictions of species numbers that are consistent with inventory totals for the groups(Figure2).When applied to all eukaryote kingdoms,our approach predicted,7.77million species of animals,,298,000species of plants,,611,000species of fungi,,36,400species of protozoa,and,27,500species of chromists;in total the approach predicted that,8.74million species of eukaryotes exist on Earth(Table2).Restricting this approach to marine taxa resulted in a prediction of2.21million eukaryote species in the world’s oceans(Table2).We also applied the approach to prokaryotes;unfortunately,the steady pace of description of taxa at all taxonomic ranks precluded the calculation of asymptotes for higher taxa(Figure S1).Thus,we ud raw numbers of higher taxa(rather than asymptotic estimates)for prokaryotes,and as such our estimates reprent only lower bounds on the diversity in this group.Our approach predicted a lower bound of,10,100species of prokaryotes,of which,1,320are marine.It is important to note that for prokaryotes,the species concept tolerates a much higher degree of genetic dissimilarity than in most eukaryotes[26,27];additionally, due to horizontal gene transfers among phylogenetic clades, species take longer to isolate in prokaryotes than in eukaryotes, and thus the former species are much older than the latter[26,27]; as a result the number of described species of prokaryotes is small (only,10,000species are currently accepted).
Asssment of Possible Limitations
We recognize a number of factors that can influence the interpretation and robustness of the estimates derived from the method described here.The are analyzed below.
Species definitions.An important caveat to the interpretation of our results concerns the definition of species. Different taxonomic ,zoologists,botanists,and bacteriologists)u different levels of differentiation to define a species.This implies that the numbers of species for taxa classified according to different conventions are not directly comparable. For example,that prokaryotes add only0.1%to the total number of known species is not so much a statement about the diversity of prokaryotes as it is a statement about what a species means in this group.Thus,although estimates of the number of species are internally consistent for kingdoms classified under the same conventions,our aggregated predictions for eukaryotes and prokaryotes should be interpreted with that caution in mind. Changes in higher taxonomy.Increas or decreas in the number of higher taxa will affect the raw data ud in our method and thus its estimates of the total number of species.The number of higher taxa can change for veral reasons including new discoveries,the lumping or splitting of taxa due to improved phylogenies and switching from phenetic to phylogenetic classifications,and the detection of synonyms.A survey of2,938 taxonomists with experti across all
major domains of life (respon rate19%,e Materials and Methods)revealed that synonyms are a major problem at the species level,but much less so at higher taxonomic levels.The percentage of taxa names currently believed to be synonyms ranged from17.9(628.7SD) for species,to7.38(615.8SD)for genera,to5.5(634.0SD)for families,to3.72(645.2SD)for orders,to1.15(68.37SD)for class,to0.99(67.74SD)for phyla.The results suggest that by not using the species-level data,our higher-taxon approach is less nsitive to the problem of synonyms.Nevertheless,to asss the extent to which any changes in higher taxonomy will influence our current estimates,we carried out a nsitivity analysis in which the number of species was calculated in respon to variations in the number of higher taxa(Figure3A–3E,Figure S2).This analysis indicates that our current estimates are remarkably robust to changes in higher taxonomy.
Changes in taxonomic effort.Taxonomic effort can be a strong determinant of species discovery rates[21].Hence the estimated asymptotes from the temporal accumulation curves of higher taxa(dotted horizontal line in Figure1A–1E)might be driven by a decline in taxonomic effort.We presume,however, that this is not a major factor:while the discovery rate of higher taxa is declining(black dots and red lines in Figure3F–3J),the rate of description of new species remains relatively constant(grey lines in Figure3F–3J).This suggests that the asymptotic trends among higher 人物神态的词语
taxonomic levels do not result from a lack of taxonomic effort as there has been at least sufficient effort to describe new
Author Summary
Knowing the number of species on Earth is one of the most basic yet elusive questions in science.Unfortunately, obtaining an accurate number is constrained by the fact that most species remain to be described and becau indirect attempts to answer this question have been highly controversial.Here,we document that the taxonomic classification of species into higher taxonomic groups (from genera to phyla)follows a consistent pattern from which the total number of species in any taxonomic group can be predicted.Asssment of this pattern for all kingdoms of life on Earth predicts,8.7million(61.3 million SE)species globally,of which,2.2million(60.18 million SE)are marine.Our results suggest that some86% of the species on Earth,and91%in the ocean,still await description.Closing this knowledge gap will require a renewed interest in exploration and taxonomy,and a continuing effort to catalogue existing biodiversity data in publicly available databas.
species at a constant rate.Secondly,although a majority(79.4%) of experts that we polled in our taxo
微信群主怎么转让nomic survey felt that the number of taxonomic experts is decreasing,it was pointed out that other factors are counteracting this trend.The included,among others,more amateur taxonomists and phylogeneticists,new sampling methods and molecular identification tools,incread international collaboration,better access to information,and access to new areas of exploration.Taken together the factors have resulted in a constant rate of description of new species,as evident in our Figure1,Figure3F–3J,and Figure S1and suggest that the obrved flattening of the discovery curves of higher taxa is unlikely to be driven by a lack of taxonomic effort.
Completeness of taxonomic inventories.To account for yet-to-be-discovered higher taxa,our approach fitted asymptotic regression models to the temporal accumulation curve of higher taxa.A critical question is how the completeness of such curves will affect the asymptotic prediction.To address this,we performed a nsitivity analysis in which the asymptotic number of taxa was calculated for accumulation curves with different levels of completeness.The results of this test indicated that the asymptotic regression models ud here would underestimate the number of predicted taxa when very incomplete inventories are ud(Figure3K–3O).This underestimation in the number of higher taxa would lower our prediction of the number of species
Table1.Available methods for estimating the global number of species and their limitations.
Ca Study Limitations Macroecological patterns
Body size frequency distributions.By extrapolation from the frequency of large to small species,May[7]estimated10to50million species of animals.May[7]suggested that there was no reason to expect a simple scaling law from large to small species.Further studies confirmed different modes of evolution among small species[4]and inconsistent body size frequency distributions among taxa[4].
Latitudinal gradients in species.By extrapolation from the better sampled temperate regions to the tropics,Raven[10]estimated3to5million species of large organisms.May[2]questioned the assumption that temperate regions were better sampled than tropical ones;the approach also assumed consistent diversity gradients across taxa which is not factual[4].
Species-area relationships.By extrapolation from the number of species in deep-a samples,Grassle&Maciolek[13]estimated that the world’s deep afloor could contain up to10million species.Lambshead&Bouchet[12]questioned this estimation by showing that high local diversity in the deep a does not necessarily reflect high global biodiversity given low species turnover.
Diversity ratios
Ratios between taxa.By assuming a global6:1ratio of fungi to vascular plants and that there are,270,000species of vascular plants, Hawksworth[20]estimated1.6million fungi species.Ratio-like approaches have been heavily critiqued becau,given known patterns of species turnover,locally estimated ratios between taxa may or may not be consistent at the global scale[3,12]and becau at least one group of organisms should be well known at the global scale,which may not always be true[15].Bouchet[6]elegantly demonstrated the shortcomings of ratio-bad approaches by showing how even for a well-inventoried marine region,the ratio of fishes to total multicellular organisms would yield,0.5million global marine species whereas the ratio of Brachyura to total multicellular organisms in the same sampled region would yield,1.5million species.
Host-specificity and spatial ratios.Given50,000known species of tropical trees
and assuming a5:1ratio of host beetles to trees,that beetles reprent40%of
the canopy arthropods,and that the canopy has twice the species of the ground,
Erwin[9]estimated30million species of arthropods in the tropics.
Known to unknown ratios.Hodkinson&Casson[18]estimated that62.5%of the
bug(Hemiptera)species in a sampled location were unknown;by assuming that7.5%–10%of the global diversity of incts is bugs,they estimated between1.84and2.57million species of incts globally.
Taxonomic patterns
Time-species accumulation curves.By extrapolation from the discovery record
it was estimated that there are,19,800species of marine fishes[23]and,11,997 birds[22].This approach is not widely applicable becau it requires species accumulation curves to approach asymptotic levels,which is only true for a small number of well-described taxa[22–23].劳伤腰痛
Authors-species accumulation curves.Modeling the number of authors describing species over time allowed rearchers to estimate that the proportion of flowering plants yet to be discovered is13%to18%[21].This is a very recent method and the effect of a number of assumptions remains to be evaluated.One is the extent to which the description of new species is shifting from using taxonomic experti alone to relying on molecular methods(particularly among small organisms[26])and the other that not all authors listed on a manuscript are taxonomic experts, particularly in recent times when the number of coauthors per taxa described is increasing[21,38],wh
ich could be due to more collaborative rearch[38]and the acknowledgment of technicians,field assistants, specimen collectors,and so on as coauthors(Philippe Bouchet,personal communication).
Analysis of expert estimations.Estimates of,5million species of incts[15] and,200,000marine species[14]were arrived at by compiling opinion-bad estimates from taxonomic experts.Robustness in the estimations is assumed from the consistency of respons among different experts.Erwin[5]labeled this approach as‘‘non-scientific’’due to a lack of verification.Estimates can vary widely,even tho of a single expert[5,6]. Bouchet[6]argues that expert estimations are often pasd on from one expert to another and therefore a robust estimation could be the‘‘same guess copied again and again’’.
doi:10.1371/journal.pbio.1001127.t001
through our higher taxon approach,which suggests that our species estimates are conrvative,particularly for poorly sampled taxa.We reason that underestimation due to this effect is vere for prokaryotes due to the ongoing discovery of higher taxa (Figure S1)but is likely to be modest in most eukaryote groups becau the rate of discovery of higher taxa is rapidly declining (Figure 1A–3E,Figure S1,Figure 3F–3J).什么叫组织
Since higher taxonomic levels are described more completely (Figure 1A–1E),the resulting error from incomplete inventories should decrea while rising in the taxonomic hierarchy.Recalculating the number of species while omitting all data from genera yielded new estimates that were mostly within the intervals of our original estimates (Figure S3).However,Chromista (on Earth and in the ocean)and Fungi (in the ocean)were exceptions,having inflated predictions without the genera data (Figure S3).This inflation in the predicted number of species without genera data highlights the high incompleteness of at least the genera data in tho three cas.In fact,Adl et al.’s [28]survey of expert opinions reported that the number of described species of chromists could be in the order of 140,000,which is nearly 10times the number of species currently catalogued in the databas ud here (Table 1).The results suggest that our estimates for Chromista and Fungi (in the ocean)need to be considered with caution due to the incomplete nature of their data.
Subjectivity in the Linnaean system of classification.
Different ideas about the correct classification of species into a taxonomic hierarchy may distort the shape of the relationships we describe here.However,an asssment of the taxonomic hierarchy shows a consistent pattern;we found that at any taxonomic rank,
the diversity of subordinate taxa is concentrated within a few groups with a long tail of low-diversity groups (Figure 3P–3T).Although we cannot refute the possibility of arbitrary decisions in the classification of some taxa,the consistent patterns in Figure 3P–3T imply that the decisions do not obscure the robust underlying relationship between taxonomic levels.The mechanism for the exponential relationships between nested taxonomic levels is uncertain,but in the ca of taxa classified phylogenetically,it may reflect patterns of diversification likely characterized by radiations within a few clades and little cladogenesis in most others [29].We would like to caution that the databa we ud here for protistan eukaryotes (mostly in Protozoa and Chromista in this work)combines elements of various classification schemes from different ages—in fact the very division of the organisms into ‘‘Protozoa’’and ‘‘Chromista’’kingdoms is non-phylogenetic and not widely followed among protistologists [28].It would be valuable to revisit the species estimates for protistan eukaryotes once their global catalogue can be organized into a valid and stable higher taxonomy (and their catalogue of described species is more complete—e above).
Discussion
Knowing the total number of species has been a question of great interest motivated in part by our collective curiosity about the diversity of life on Earth and in part by the need to provide a reference p
oint for current and future loss of biodiversity.Unfortunately,incomplete sampling of the world’s biodiversity combined with a lack of robust extrapolation approaches
has
Figure 1.Predicting the global number of species in Animalia from their higher taxonomy.(A–F)The temporal accumulation of taxa (black lines)and the frequency of the multimodel fits to all starting years lected (graded colors).The horizontal dashed lines indicate the connsus asymptotic number of taxa,and the horizontal grey area its connsus standard error.(G)Relationship between the connsus asymptotic number of higher taxa and the numerical hierarchy of each taxonomic rank.Black circles reprent the connsus asymptotes,green circles the catalogued number of taxa,and the box at the species level indicates the 95%confidence interval around the predicted number of species (e Materials and Methods).
doi:10.1371/journal.pbio.1001127.g001
yielded highly uncertain and controversial estimates of how many species there are on Earth.In this paper,we describe a new approach who validation against existing inventories and explicit statistical nature adds greater robustness to the
estimation of the number of species of given taxa.In general,the approach was reasonably robust to various caveats,and we hope that future improvements in data quality will further diminish problems with synonyms and incompleteness of data,and lead to even better (and likely higher)estimates of global species richness.
Our current estimate of ,8.7million species narrows the range of 3to 100million species suggested by taxonomic experts [1]and it suggests that after 250years of taxonomic classification only a small fraction of species on Earth (,14%)and in the ocean (,9%)have been indexed in a central databa (Table 2).Closing this knowledge gap may still take a lot longer.Considering current rates of description of eukaryote species in the last 20years (i.e.,6,200species per year;6811SD;Figure 3F–3J),the average number of new species described per taxonomist’s career (i.e.,24.8species,[30])and the estimated average cost to describe animal species (i.e.,US $48,500per species [30])and assuming that the values remain constant and are general among taxonomic groups,describing Earth’s remaining species may take as long as 1,200years and would require 303,000taxonomists at an approximated cost of US $364billion.With extinction rates now exceeding natural background rates by a factor of 100to 1,000[31],our results also suggest that this slow advance in the description of species will lead to species becoming extinct before we know they even existed.High rates of biodiversity loss provide an urgent incentive to increa our knowledge of Earth’s remaining species.
Previous studies have indicated that current catalogues of species are biad towards conspicuous species with large geographical ranges,body sizes,and abundances [4,32].This suggests that the bulk of species that remain to be discovered are likely to be small-ranged and perhaps concentrated
in hotspots and less explored areas such as the deep a and soil;although their small body-size and cryptic nature suggest that many could be found literally in our own ‘‘backyards’’(after Hawksworth and Rossman [33]).Though remarkable efforts and progress have been made,a further closing of this knowledge gap will require a renewed interest in exploration and taxonomy by
both
Figure 2.Validating the higher taxon approach.We compared the number of species estimated from the higher taxon approach implemented here to the known number of species in relatively well-studied taxonomic groups as derived from published sources [37].We also ud estimations from multimodel averaging from species accumulation curves for taxa with near-complete inventories.Vertical lines indicate the range of variation in the number of species from different sources.The dotted line indicates the 1:1ratio.Note that published species numbers (y -axis values)are mostly derived from expert approximations for well-known groups;hence there is a possibility that tho estimates are subject to bias arising from synonyms.
doi:10.1371/journal.pbio.1001127.g002
Table 2.Currently catalogued and predicted total number of species on Earth and in the ocean.
Species
Earth Ocean Catalogued
Predicted
±SE
Catalogued
Predicted
±SE
Eukaryotes Animalia 953,4347,770,000958,000171,0822,150,000145,000Chromista 13,03327,50030,5004,8597,4009,640Fungi 43,271611,000297,0001,0975,32011,100Plantae 215,644298,0008,2008,60016,6009,130Protozoa 8,11836,4006,6908,11836,4006,690Total 1,233,500
8,740,000
三年级上册口算题100道1,300,000
193,756
2,210,000
182,000
Prokaryotes Archaea 502455160110Bacteria 10,3589,6803,4706521,320436Total 10,86010,1003,6306531,320436Grand Total
1,244,360
8,750,000
1,300,000
194,409
2,210,000
182,000
Predictions for prokaryotes reprent a lower bound becau they do not consider undescribed higher taxa.For protozoa,the ocean databa was substantially more complete than the databa for the entire Earth so we only ud the former to estimate the total number of species in this taxon.All predictions were rounded to three significant digits.
doi:10.1371/journal.pbio.1001127.t002
