Biodiversity and distribution of earthworms in the biogeographic province of the Levant

Abstract

We are reporting an increase from 30 to 52 identified earthworm taxonomic species, including one subspecies, in the Levant bioregion during the last 20 years. The number of earthworm taxonomic species is still underestimated. We classified the earthworms of the Levantine into three categories: endemic species, autochthonous species and introduced peregrine species. The endemic species might have originated before the Messinian stage (7.5–3.5 mya) of the Miocene. We suggest that certain endemic species underwent speciation (and subsequently disappeared) during interglacial (pluvial) periods. These changes occurred when the Levantine corridor transformed into a “culdesac” pathway due to the southern opening closing off, caused by the retreating steppe zone and its replacement with a gradually expanding open park forest. As for the autochthonous species, they arrived in the Levant coming from different biogeographic regions by means of biological dispersal and did not have enough time to speciate. The fact that autochthonous species of the Afrotropical realm and of southeastern Asia origin are missing indicates that such species either were not able to cross the Sahara Desert, the Himalayan and the Tien-Shan orogens or that they were extinct before the end of the Miocene. The establishment of the introduced peregrine species group in the Levant has been facilitated by the availability of human-constructed niches during the current interglacial.

Keywords:

• Evolution
• speciation
• endemism
• dispersals

Introduction

To understand the diversity and distribution of earthworms in the Levant, one has to understand the geological, climatic and tectonic variability to which earthworms were exposed to during their evolution. Based on the current geological and climatic situation, the Levant biogeographic province had been defined as a “stretch of land about 150 km wide, wedged between the Mediterranean Sea and the Syro-Arabian desert belt, stretching from the Taurus Mountains in the north to the Isthmus of Suez in the south” (Pavlíček et al., 2003). This definition roughly corresponds to the definition of the Levant by Por (1975) and to the Syro-Palestinian chorotype in Vigna-Taglianti et al. (2003).

The Levant was established in the Early Miocene (16–23 mya) due to the collisions of the African and Eurasian tectonic plates. In the beginning, the Levant operated as a transitional dispersal corridor for mammals and other animals between Africa and Eu rope. The Levant became a biogeographic province, sufficiently isolated to hamper gene flow after the Pontian and Zagros orogeneses in the Late Miocene (Tchernov, 1988). In the Late Miocene, the Afrotropical mammalian fauna became isolated from North Africa and from the Levant. For instance, Tchernov (1988) argued that marked aridity affected the list of mammals already from around 14 mya. If correct, this conclusion indicates the existence of the core of the Sahara-Syrian desert belt and the limiting of faunal exchange already during the Middle Miocene. With the establishment of the Saharo-Syrian desert belt, the Levantine corridor changed into ”culs-de-sac” (see Table 1 for definition) with the opening in the northern part.

Table 1. Nomenclature of some of the terms used in the article with examples in earthworms.

Cyprus existed as an isolated uplifted oceanic island already in Paleogene (66–23 mya). During the Messinian salinity crisis period (7.5–5.33 mya), Tethys (Mediterranean Sea) almost evaporated (Robertson, 1998). The island had been connected to the adjacent Asiatic continental plates through today’s submerged continental ridges (Robertson, 1998). The progressive desiccation began in the Miocene and became the major climatic trend in the Levant during the Quaternary (approximately, 2.6 mya up to now) (Tchernov, 1988). At the end of the Miocene, the climate was drier and cooler than previously and frost-free winters were nevertheless still supporting a kind of savannah community growing all year round (Tchernov, 1988).

The scenario proposed above would imply that the late Miocene earthworm fauna, if it was at all present, was expected to survive a) the extensive early Pliocene Tethys Sea transgression that brought the fragmentation of the circum-Mediterranean region (Tchernov, 1988), b) the extensive volcanic and tectonic activities that took place in the Levant, mainly during the Quaternary, and c) the glacial periods. In fact, these glacial periods did not allow earthworms to disperse across glaciated areas in the eastern Black Sea region, the Taurus Mountains, the Anatolian Plateau, and some volcanoes and independent mountain chains. The only periods available for their biological dispersal from the adjacent zoogeographic regions to and from the Levant were in the relatively short interglacials (on average 10,000–30,000 years) that occured during the last 800,000 years (PAGES, 2022). Importantly, the interglacials in the East Mediterranean bioregion correspond to pluvial periods (Deuser, et al. 1976).

In this paper, we attempt to infer the evolutionary history of Levantine earthworms from their current distributions, in combination with the geological, climatic and tectonic history of the region.

Material and Methods

We compiled old and current reports of earthworm species occurrences in the Levant and included own records which we obtained during the last two decades (Tables 2–3). We divided the species into three categories: a) endemic, b) autochthonous, and c) introduced peregrine species. We also described the presence of polyploid lineages in eleven species from a total of 52 species. In order to test p values of a single proportion available, we used a calculator available at the following address: http://statkat.com/online-calculators/p-value-binomial-test-single-proportion. php. The exact binomial distribution test evaluates the statistical significance of deviations from a theoretically expected distribution of observations into two categories using sample data (https://sites.utexas.edu/sos/guided/inferential/catego-rical/univeriate/binomial). The null hypothesis (H₀) is saying that the number of compared species (e.g. endemic and introduced peregrine ones) is not biased in favour of tails. In our case, we determined H₀ = 0.33. The alternative hypotheses, H₁, and H₂ are saying that: the number of compared species is biased in favour of one of the two tails. If p > α, then we do not reject the null hypothesis. If p < α, we accept one of the two alternative hypotheses.

Table 2. A list of earthworm taxa recorded in the political regions of the Levant. Dendrobaena pantaleonis and Lumbricus terrestris are excluded from the table because of their improbable occurrence in the Levant. Abbreviations: CY = Cyprus; HT = Hatay (Turkey); IL = Israel; JO = Jordan; LB = Lebanon; SI = Sinai Peninsula (Egypt); SY = Syria), n. r. – a new record, Empty cells: no record known. The numbers correspond to the following references: 1: Csuzdi et al. (1999); 2: Csuzdi & Pavlíček (2002); 3: Csuzdi & Pavlíček (2005a); 4: Csuzdi & Pavlíček (2005b); 5: Csuzdi et al. (2007); 6: Pavlíček & Csuzdi (2005); 7: Pavlíček et al. (1997); 8: Pavlíček et al. (2003); 9: Pavlíček & Csuzdi (2012a); 10: Rosa (1893); 11: Szederjesi et al. (2013a); 12: Szederjesi et al. (2013b); 13: Szederjesi et al. (2016); 14: Szederjesi et al. (2018); 15: Szederjesi et al. (2019); 16: Michaelsen (1910); 17: Michalis (1993).

Table 3. So far unpublished earthworm records for the Levant and the species added to the list of the Levantine fauna during the years 2003 to 2022. det. = identified by; leg. = collected by.

Results

Taxonomic overview. The data from the Levant show that the number of earthworm species and subspecies grew in the last ∼20 years from 32 to 52 and that the number of genera from 16 to 20 (Table 2). The genera/species ratio decreased from 0.5 to 0.38 in the last 20 years. The following five families were present (Tables 2, 3):

1. Acanthodrilidae with three introduced peregrine species (Dichogaster bolaui, Microscolex dubius and M. phosphoreus) originating from South America.

2. Criodrilidae with one widely distributed autochthonous species (Criodrilus lacuum) of Palearctic origin.

3. Lumbricidae with 44 species of Holarctic origin.

4. Megascolecidae with two peregrine species (Amynthas cortices and Metaphire californica) of south-eastern Asia origin.

5. Ocnerodrilidae with two peregrine species (Eukeria saltensis and Ocnerodrilus occidentalis) of southern Asia origin.

In the political entities of the Levant, the highest species number is recorded in Israel (N=37), followed by Cyprus (N=24), the Hatay Province of Türkiye (N=22), Jordan (N=21), Lebanon (N=19), Syria (N=15) and the Sinai, Egypt (N=3) (Tables 2, 3). However, the comparison of species richness (Table 2) among the different political entities is biased since it does not consider the size of the regions, the sampling intensity and methodology, and the climate. For instance, much larger collection efforts were made in Israel than in other political entities. This tendency is shown, among others, by the seven endemic species described as new to science from Israel (Table 3) (Dendrobena negevis, D. nevoi, D. rothschildae, D. alexandrii, Healyella jordanis, Perelia galileana, P. shamsi) out of eleven endemic species recorded for the Levant in the total (+ D. orientalis karak, D. omodeoi, P. hatayica, P. makrisi). Nevertheless, we can observe an expected trend in the decrease of species richness from the mesic northern Levant towards the arid, desertic southern Levant following the drought gradient. In the southern Levantine deserts, only springs, wells, water reservoirs and oases can serve as biotopes for the introduced earthworms (E. tetraedra, Ap. caliginosa, and Ap. rosea: Pavlíček et al., 1997). No autochthonous species originated from the southern Levantine deserts (Pavlíček et al., 1997) nor from south-eastern Asia. A faunal exchange between the Levant and south-eastern Asia, if ever it existed in earthworms, had ceased, possibly, due to the uplifting of the huge Tien-Shan and Himalayan orogens during the Miocene (Jia et al., 2020).

Endemic species. We have determined that the number of Levantine endemic species is X=17 (which is 33% of the total of 52 species, Table 4). The null hypothesis (Ho) is set as 17/52 = 0.33. The binomial distribution test (with X=17 or a smaller number, n = 52, P=0.33, and α=0.05) is equal to p=0.546. It is therefore unlikely that we do not have sufficient evidence to claim that the number of the 17 endemic species deviates significantly from the total of 52 species. The endemism occurs in the three lumbricid genera: Dendrobaena (47%, 9 from 19 species belonging to the Levantine members of the genus Dendrobaena), Healyella (50%, one from two species of the Levantine members of the genus Healyella) and Perelia (83%, 5 from 6 species of the Levantine members of the genus Perelia, Tables 2, 4). Since endemic species have not extended beyond the Levant, they could not have a direct impact (i.e. interactions) on the rest of the Palearctic earthworm fauna.

Table 4. Division of the Levantine earthworms according to their chorotype, reproductive mode, and origin in the Levant. Pc = Ploidy complexes. A – autochthonous, E – endemic, I – introduced, E – east, N – north, ? – uncertain. See Table 2 for species authority and year of description.

Autochthonous species of the Levant. The autochthonous species came from different adjacent zoogeographic regions by means of biological dispersal. Biological dispersal is the most important differentiation characteristic between autochthonous and introduced peregrine species. In introduced peregrine species it is facilitated by humans. Most autochthonous earthworms show zoogeographic affinities towards Europe (mainly towards the Balkans and the Carpathian Mts.), Central Asia, the Caucasus Mountains and Anatolia. No Afrotropical earthworm taxa were found in the Levant (including Cyprus). This fact points to the stability of the Saharo-Syrian desert belt and its ability to prevent earthworm fauna exchange between the Levant and the Afrotropics. On the one hand, some of the autochthonous species in the Levant that are also distributed in the adjacent areas might be Levantine escapees (for example D. samarigera, D. succinta and Healyella syriaca) (Tables 2–3). On the other hand, the insular earthworm fauna of Cyprus and the earthworm fauna of northern part of northern Levant (Hatay, Türkiye) contain Dendrobaena pentheri (Csuzdi et al., 2007), an autochthonous element of the Anatolian-Caucasian fauna. This may indicate that D. pentheri is currently entering the Levant. In our opinion, the occurrence of D. pantaleonis recorded in Cyprus (Michalis, 1993) needs reconfirmation as well as the extent of endemism reported so far in Perelia phoebea from Cyprus and Rhodes (Szederjesi et al., 2016) (Tables 2–4).

Introduced peregrine species. The determined number of Levantine introduced species is X=21 (i.e. 40.4% of the 52 species, Table 4). By means of a binomial distribution test (with X=21, n=52 and P=0.33), we found out that the probability of finding the observed number of successes X=21 or a larger number, if the null hypothesis were true, is equal to p = 0.162. It is therefore unlikely that we have sufficient evidence to claim that the number of X=21 introduced species is biased in favour of tails and thus behave differently from the remaining total number of species that is higher than expected. However, we determined the number of the introduced peregrine species X1=16 and peregrine one as X2=21 (Table 4). The relationship between X2 and X1 is 0.905. The binomial distribution test (with X=16 or a larger number, n=21, P=0.33, and α =0.05) is equal to p=0.000, and it is therefore likely that we have sufficient evidence to claim the number of the introduced peregrine species is biased in favour of tails that is lower than expected. The high overlap indicates that the majority of the introduced peregrine species are different to the introduced ones and it is possible to differentiate between these two categories (Table 2). Therefore, the introduced peregrine species can be regarded as one earthworm category (see definition in Table 2).

We expect that the introduced peregrine species were immigrating to the Levant helped by human-facilitated dispersal during the Holocene interglacial (approximately the last 12,000 years of the Earth’s history). If some of them had been introduced to the Levant by several unknown endozoochoric agent(s), they should not be called introduced species but rather autochthonous ones. Introduced peregrine species often populate specific anthropogenic niches such as greenhouses (e.g. Amynthas corticis), showers and water pipes (e.g. Dichogaster bolaui) (Csuzdi et al., 2008), wet compost heaps (e.g. Eisenia fetida), artificial canals, etc.

Polyploidy. If we were to categorize all ploidy lineages as 'valid' taxonomic species, we would need to include approximately 15 additional lineages to the species already listed in Table 5. It is nevertheless questionable whether different polyploid linegaes should be taxonomically named as species since we do not know the extent of gene flow between them. However, triploids were found to be able to hybridize only with one another or escape the evolutionary constraints by their genome duplication (Table 5). In the four widely distributed taxonomic species (Ap. trapezoides, Bimastos rubidus, Eisenia fetida and Octolasion cyaneum), uniparental reproduction (e.g. autogamy and embryonic cloning) has been recorded. The species complexes Dendrobaena semitica, D. byblica and D. veneta need taxonomic revisions. So far, only a part of D. veneta (Szederjesi et al., 2019) was revised.

Table 5. Tentative list of macroevolution events among ploidy lineages in earthworms of the Levant.

Discussion

The Levantine biota consists of endemic, autochthonous, and introduced peregrine earthworm species. Whereas the endemic species originated locally and did not spread beyond the borders of the Levant, the appearance of relatively numerous endemic species and their mosaic distribution in the mesic part of the Levant in a few genera only (Dendrobaena, Perelia, and Healyella) seems to be related to speciation events perhaps associated with climate, orography, pedology (soil type) and possibly with the unbalanced biotic interactions established by colonization disharmonies.

Missing faunal associations. Two groups of missing associations have been observed in autochthonous species. Missing autochthonous migrants from south-eastern Asia are pointing toward the powerful ceasing role of the uplifted Hindu-Kush and Himalayan ranges during the Miocene Vallesian (11.6–9 mya). They ceased the biotic interchange between the Indian subcontinent, Europe and Africa (the latter through the Levant). In addition, it seems that, at least, from the Middle Miocene onwards, the Sahara-Syrian desert belt has been a powerful isolation barrier between the mesic Levant and the humid Afrotropics. This finding supports the existence of a core Sahara-Syrian desert belt already in the Middle or Late Miocene or points to an extinction of Afrotropical faunal and floral elements occurring in the Levant as a result of the growing desertification. In a contrast to the southern, eastern and western sides of the Levant which were impenetrably closed to an earthworm migration by desert or by the Mediterranean Sea, the northern opening intermittently worked during the inter-glaciation periods in the Quaternary. We have to keep in mind that, at least during the last glacial period (corresponding to Würm in the Alps: about 110,000–15,000 ya), the territory of the Levant was not covered by glaciers and that steppe vegetation was abundant. The first Pre-Würm interglacial period is the Eemian (∼ 30,000–115,000 ya) at the end of the Penultimate Glacial Period (∼194,000–135,000 ya). It ended at the beginning of the dry (in the Levant) Würm Glacial Period (∼115,000–11,700 ya).

Speciation in the Levantine Earthworms. We are expecting that speciation leading to the distribution of local endemism, accompanied by an unknown number of extinctions, took and is still taking place in “culs-de-sac” of different sizes established mainly during the interglacial (pluvial) periods (Deuser et al., 1976), including the recent one. The Levantine endemic species never left the Levant and they, therefore, have no direct impact on the rest of the Holarctic earthworm fauna. However, some of the autochthonous species in the Levant and the ones distributed in adjacent areas might be Levantine emigrants (e.g. Dendrobaena samarigera, D. succinta and Healyella syriaca) or immigrants (e.g. D. pentheri).

The frozen niche variation model (Williams, 1975; Vrijenhoek, 1984) suggests that in a population composed of multiple different clones, each clone uses only a fraction of the total niche to avoid inter-clonal competition. The occurrence of a clonal structure in the local earthworm populations was verified experimentally (e.g., Jaenike et al., 1980).

The major advantage of asexual and unisexual reproduction modes, in comparison with amphimictic ones, is the lower cost of progeny. Nevertheless, the number of asexual (mitotic) divisions between two consecutive meiosis might be under the Hayflick limit control (Hayflick & Moorhead 1961; definition in Table 1).

Autogamy (see Table 1 for definition) was recorded among the widespread Levantine earthworms E. fetida, B. rubidus, and O. cyaneum (Table 4). The mitotic reproduction process known in earthworms, except for the lost tissue regeneration, is embryonic cloning, e.g., polyembryony and twinning. Polyembryony is produced on embryos established amphimictically (for example in L. terrestris) whereas the twinning is uniparental (Noli et al., 2017). Uniparental polyembryony might be facilitated by restitutional automixis (e.g. in O. cyaneum) or by complete automixis (Gorelick & Carpione, 2009; Engelstädter, 2017) (Table 1). The body size variability might be seen as an example of positive natural selection operating on the ecological or population levels towards more efficient resource utilization and decreasing intraspecific competition (Gorelick et al., 2011). At the molecular level, there should be higher proportions of cytosine methylation (Gorelick et al., 2011) and hence a higher T mutation rate in neopolyploids and their diploid ancestors (Salmon et al., 2005; Gorelick et al., 2011). A higher mutation rate should allow a quicker response to environmental challenges and to the ability of neopolyploids to occupy harsher environments, a property known in earthworms as geographic parthenogenesis.

In fact, seven of the Levantine polyploid species display triploids (Amynthas corticis, Ap. caliginosa, Ap. rosea, Ap. trapezoides, Eiseniella tetraedra, Octodrilus transpadanus, and Octolasion cyaneum). Only two Levantine polyploid species (Bimastos rubidus, and C. lacuum) (Table 4) do not show triploids, but these still might be found. Does it mean that in some earthworm species, the first step of speciation might be associated with triploidization? Chromosome duplications are nevertheless also present (Table 5).

Another cytological process which might occur in earthworms is aneuploidy (Vitturi et al., 2000) and hybridization (Pavlíček et al., 2012b, 2017; Plytycz et al., 2018). Aneuploidy has, as far as we know, not been recorded in any of the Levantine earthworms (Pavlíček et al., 2017, Table 1). We have no doubt, nevertheless, that aneuploidization and hybridization took place in some earthworms and other Oligochaeta.

While working in the field, most biologists do not know whether the earthworm specimens they sampled reproduced by means of amphimixis, autogamy, uniparentally or asexually (mitotically), and whether two reproducing individuals collected in the same locality belong to the same clone or to the same generation (see Table 1 for definition). Epigenetic demarcation of the generations might be the practical methodology to use in order to determine if the clones have been isolated one from one another or have been interbreeding (Vitturi et al., 2000; Gorelick et al., 2011).

Disclosure Statement

No potential conflict of interest was reported by the authors.