Importance of diversity estimates
Estimates of diversity are important for conservation, agriculture and fisheries. However, our knowledge is very limited. One often reads about threatened extinction of some conspicuous animals or plants, but it is usually not realized that each large species is host to many species of parasites, some of them specific to that host species and therefore doomed to extinction with it, and that many species live in close association with various symbionts. The human species, for example, is host to far more than 100 parasite species, quite a few of these only found in humans. The role of symbionts and parasites in ecosystems is largely unknown, but may be important. Among free-living species there are many, for example large animals at the top of the food chain, that may be essential for the maintenance of an entire ecosystem.
For this discussion, we define diversity as the number of species in a habitat, region or on earth, although for certain ecological studies it can be defined as the number of species per unit area with consideration of the frequency of each species. But what is a species?
What is a species?
A species can be defined in different ways, based on interbreeding or common descent, (e.g., Sluys and Hazevoet 1999  ). For multicellular and most unicellular animals, the biological species concept based on interbreeding (a species comprises all individuals that can interbreed freely in nature producing fertile offspring) is usually satisfactory, although intermediate forms may exist (“Rassenkreise”, geographical species), and different species may be “forced” to interbreed in captivity. However, for microorganisms different concepts may be necessary. Interbreeding is made likely, even if it has not been observed in nature, by a greater within-species than between-species morphological (and molecular) similarity. A clear decision is not always possible, i.e., there will always be ambiguous cases of species identity, which may lead to over- or underestimates of species numbers.
Estimating species richness
Any estimates of local, regional or global diversity must be based on surveys, i.e., counts of species numbers. Recent discoveries of considerable numbers of hitherto undescribed large marine vertebrates (fish) and invertebrates (such as crustaceans, molluscs and squid) around Timor and in the Subantarctic have shown that even large animals are far from completely known. Nevertheless, surveys of large plants and of large animals such as vertebrates on land, in freshwater and in the oceans have been relatively thorough (although far from complete), whereas smaller plants and animals, parasites, and even more so microorganisms are very little and in many habitats hardly known at all. For example, the meiofauna (multicellular non-planktonic organisms too small to be retained on a 1mm sieve) of coasts has been thoroughly examined only around the Island of Sylt in the North Sea, by a large group of researchers from the University of Göttingen over many years. 652 species have been described and identified, but according to estimates, a further 200 (including some unicellular species) are still waiting to be discovered (Armonies and Reise 2000  ). Since many species are very restricted in their geographical distribution, global meiofaunal diversity must be immense.
The deep-sea (which covers about 65% of the Earth’s surface) has a rich, largely unknown fauna. Nematodes (roundworms) are among the most diverse and common species on the deep-sea floor, but only 10 studies of species diversity of deep-sea nematodes had been conducted until 1994 (Lambshead et al. 1994  ), i.e., nematode diversity was known from less than one square meter of sea bed.
According to some estimates, parasites represent more than half of all species of animals on Earth, but they are little known (for details see Rohde 2002  , 2005  ).
Hoberg (2005  ) reports that there are more than 300 species of seabirds, but only 700 species of helminths (digeneans, eucestodes, nematodes and acanthocephalans) are known from 165 bird species, i.e., the helminth (parasitic worm) fauna of about half of all the host species is completely unknown, not to mention the unicellular parasites.
Among parasites of fish (among the best examined host groups in aquatic habitats), only species of parasites of certain groups in some regions have been studied with some thoroughness (depending on the expertise available in a particular country). Among the better known groups are flukes (trematodes), but even here huge gaps remain. For example, according to Cribb (2002  ), trematodes have been recorded only from 62 of the 159 species of the large family of marine fishes, the Epinephelidae: most species were sampled from single localities, and not a single tropical species has been thoroughly examined. Exhaustive surveys of more or less “all” parasite groups of marine fishes are restricted to some northern seas.
Problems with diversity estimates
(1) Latitudinal bias and gradients
How do we arrive at estimates of global diversity? This account is partly based on May (1990  ), who points out that methods using projections from past trends estimating species numbers suffer from the fact that past studies were largely conducted in high latitude environments, but the tropical fauna may have very different patterns of diversity. May referred to the differences between tropical and temperate diversity patterns, but we also must consider – in the oceans – gradients with depth, and that relationships are not always linear. For example, Boucher and Lambshead (1995  ) found a nonlinear relationship between depth and species richness for marine nematodes, the bathyal and abyssal being richest. Also, longitudinal gradients are important. For example, a primary centre of marine diversity is found in the West Pacific Ocean around SE Asia, with species numbers declining with increasing distance from it. – More generally, we should be careful not to generalize from insufficient small surveys, as pointed out in the next section.
(2) Unjustified generalizations from few surveys
Grassle and Maciolek (1992  ) sampled along a 176 km transect on the continental slope of the northeast Atlantic Ocean, recording 798 species of macroinvertebrates. They extrapolated from these data and concluded that the global richness of deep-sea soft sentiments is in the order of 10 million species. Koslow et al. (1997  ) tested their method by extrapolating global deep sea richness of fish from a survey along the continental slope of western Australia, arriving at a number of 60,000 species. However, only about 2650 deep-sea fish species have been described to date and, adding the number of species estimated as not yet described, there should not be more than 3000-4000 species. The authors concluded that the method of Grassle and Maciolek should not be used. Major errors may be introduced because of our ignorance of habitat specificity and geographical ranges of species. – However, is a comparison of a fish survey with one on invertebrates justified? And how reliable is the estimate of a total of 3000-4000 species?
(3) Specificity and geographical distribution
The specificity of animals (e.g., of insects for certain tree species, or parasites for certain fish species) is largely unknown, and estimates based on recording insect species (comprising the majority of animal species on land) from particular plant species, or parasites from particular host species, and multiplying them by the number of plant or host species are therefore invalid. – Knowledge of the geographic distribution of species is essential for determining regional and global diversity, in order to avoid repeated counting of the same species in different regions. It is – however – unknown for most species, e.g. for species in the deepsea and meiofauna.
(4) Fractional extension
Hodkinson and Casson (1990  ) estimated the number of tropical insect species by determining the fraction of species in a well known group that had been recorded earlier, and applied this fraction to calculate total diversity. They found that 37% of species in a Sulawesi National Park was already known prior to their survey, and arrived at an estimate of a total number of 2.7 million insect species. However, such generalizations are almost certainly wrong.
(5) Size-diversity relationship
May (1990  ) discussed the empirical rule, according to which “for each tenfold reduction in length (1000-fold reduction in body weight) there are 100 times the number of species”, pointing out that the rule breaks down at body lengths below 1cm, which, however, may be due to our incomplete knowledge of small terrestrial animals.
We also have to consider the proportion of synonyms (“species aliases” . i.e., several names for the same species) (May and Nee 1995  ). Insects have been most thoroughly studied in this respect: synonymy rates were commonly about 20%, but exceeded 50% in some groups. Importantly, rates may be even higher since studies are continuing and it is unlikely that all synonyms have already been discovered. Taking the approximate proportion of synonyms into account, the number of recorded species world-wide, according to May and Nee, should be downgraded to about one million.
(7) Intermediate forms
Genuine intermediate cases between species and subspecies exist, a problem that has been well studied in some terrestrial animals (e.g., birds). Examples are “Rassenkreise” (geographical species), in which most adjacent populations interbreed but populations at the ends of the species’ range that are in secondary contact do not.
(8) Cryptic species
Molecular studies may reveal the existence of morphologically identical or very similar sibling species, or they may show that “species” distinguished on the basis of minor morphological differences may really be single species. For example, Etter et al. (1999  ) compared local populations of four deep-sea mollusc species from narrow depth ranges, only tens to hundreds of kilometers apart, and not separated by topographical barriers. Their genetic divergence was found to be similar to that found between recognized coastal marine and aquatic mollusc species. This suggests (but does not prove, because of the possibility of intermediate forms in areas between the populations) that sibling species may exist and deep-sea diversity may be much higher than usually estimated on the basis of morphology.
Global diversity estimates
All these difficulties have led to drastically different estimates of the number of species. Concerning the ocean, for example, Lambshead (1993  ) states that about 160,000 marine species have been described, and he estimated that the total for macrofauna could reach 10 million species, with the total meiofauna “an order of magnitude higher”. In contrast, Briggs (1994  ) estimated that terrestrial species diversity is about 12 million, plus or minus one million, but that marine diversity appears to be less than 200,000, and this in spite of the much larger area of the seas and the older evolutionary age.
May (1990  ) believes that the number of species on Earth “ is currently uncertain to within a factor of 10 or more.” Even the number of described species is uncertain, but may be around 1.8 million (Stork 1988  ).
We conclude that most species of multicellular, unicellular and micro-organisms have not been described. Even approximate estimates of global diversity are therefore premature. Urgently needed are extensive and intensive surveys of all groups from a multitude of habitats. In particular, the meiofauna, deep-sea fauna, parasites and microorganisms need special attention, as do the Tropics with their enormously diverse terrestrial and aquatic habitats. Furthermore, molecular techniques need to be applied for discovering cryptic species, and for checking whether morphologically similar “species” are indeed different species.
Sluys, R. and Hazevoet, C.J. (1999). Pluralism in species concepts: dividing nature at its diversity joints. Species diversity 4, 243-256. Armonies, W. and Reise, K. (2000). Faunal diversity across a sandy shore. Marine Ecology Progress Series 196, 49-57. Lambshead, P.J.D., Elge, B.M., Thistle, E., Eckman, J.E. and Barnett, P.R.O. (1994). A comparison of the biodiversity of deep-sea marine nematodes from three stations in the Rockall Trough, Northeast Atlantic, and one station in the San Diego Trough, Northeast Pacific. Biodiversity Research 2, 95-107. Rohde, K. (2002). Ecology and biogeography of marine parasites. Advances in Marine Biology, 43, 1-86. Rohde, K. ed. (2005). Marine Parasitology. CSIRO Publishing Melbourne and CABI, Oxon. Wallingford Hoberg, E.P. (2005). Marine birds and their helminth parasites. In K. Rohde, ed. Marine Parasitology, pp. 414-420. CSIRO Melbourne and CABI Wallingford, Oxon. Cribb, T.H. (2002). The trematodes of groupers (Serranidae: Epinephelidae): Knowledge, nature and evolution. Parasitology, 124, S3-S42. May, R.M. (1990). How many species? Philosophical Transactions of the Royal Society of London 330, 293-304. Boucher, G. and Lambshead, P.J.D. (1995). Ecological biodiversity of marine nematodes in samples from temperate, tropical, and deep-sea regions. Conservation Biology 9, 1594-1604. Grassle, J.F. and Maciolek, N.J. (1992). Deep-sea species richness: regional and local diversity estimates from quantitative bottom samples. American Naturalist 139, 313-341. Koslow, J.A., Williams, A. and Paxton, J.R. (1997). How many demersal fish species in the deep sea – a test of a method to extrapolate from local to global diversity. Biodiversity & Conservation 11, 1523-1532. Hodkinson, I.D. and Casson, D. (1990). A lesser predilection for bugs: Hemiptera (Insecta) diversity in tropical rain forests. Biological Journal of the Linnaean Society 43, 101-109. May, R.M. and Nee, S. (1995). The species alias problem. Nature 378, 447-448. Etter, R.J., Rex, M.A., Chase. M.C. and Quattro, J.M. (1999). A genetic dimension to deep-sea diversity. Deep-Sea Research I 46, 1095-1099. Lambshead, P.J.D. (1993). Recent developments in marine benthic biodiversity research. Océanis 19, 5-24. Briggs, J.C. (1994). Species Diversity: Land and Sea Compared. Systematic Biology 43, 130-135. Stork, N.E.(1988). Insect diversity: facts, fiction and speculation. Biological Journal of the Linnean Society 35, 321-337.