A Forgotten Rodent from the Garden of Eden: What Really Happened to The

Home , Brown rat, Eastern rat, Nesokia, Syrian Desert

Mammalia 2021; 85(2): 103–108

Original study

Boris Kryštufek, Omar F. Al-Sheikhly, Javier Lazaro, Mukhtar K. Haba, Rainer Hutterer, Sayed B. Mousavi and Danijel Ivajnšič* A forgotten rodent from the Garden of Eden: what really happened to the long-tailed nesokia rat in the Mesopotamian marshes? https://doi.org/10.1515/mammalia-2020-0092 extensive than expected with a highly suitable habitat Received June 27, 2020; accepted August 4, 2020; published online covering 15,650 km2 of Mesopotamian marshland in Iraq August 24, 2020 (between Basra and Salah Ad Din provinces) and the Hawizeh Marshes in the adjacent Iranian Khuzestan. Abstract: The greater part of expected mammalian extinctions will be of smaller-bodied mammals, including Keywords: field survey; habitat destruction; habitat rats which are more generally known only as pests and modelling; Iraq; Nesokia bunnii; threatened species. carriers of pathogens. We address the long-tailed nesokia rat, which is among the least studied Palaearctic mammals. The species is known from merely five specimens, collected 1 Introduction between March 1974 and January 1977 within a radius of fl 30 km around Qurna inside the seasonally ooded Meso- Rats of the Rattus division (Murinae) are a specious rodent potamian marshes in southern Iraq. In the 1990s, this assemblage native to southern and south-eastern Asia. extensive aquatic habitat has been deliberately reduced to They are well-known to the lay public as pests and carriers <15% of its original area and the IUCN expressed fear that of parasites and pathogens, and some of them pose a “ ” such a disaster almost certainly caused the extinction of serious threat for human health (Aplin et al. 2003). Less the long-tailed nesokia. Although the interventions after known is that some rats are also listed on the IUCN Red List 2003 reversed the shrinking trend and marshes started to as species of conservation concern (IUCN 2017) and that expand, the continuous presence of the long-tailed nesokia several species were already driven to extinction during the fi could not be unambiguously con rmed. We provide modern history (e.g. Harper 1945, Turvey 2009). Perhaps meagre evidence suggesting that the rat might be still quite surprisingly to many, the majority of mammalian present in the marshes. Next, our habitat modelling shows extinctions over the last 400 years were not of large char- that the area of the long-tailed nesokia might be more ismatic species but of smaller-bodied creatures (Entwistle and Stephenson 2000). Rarity is more common among rodents than would be expected in a random process (Ceballos and Brown 1995), hence the greater part of ex- š č *Corresponding author: Danijel Ivajn i , Faculty of Natural Sciences pected mammalian extinctions during the next few de- and Mathematics, Faculty of Arts, University of Maribor, Koroška 160, 2000, Maribor, Slovenia, E-mail: [email protected] cades will be of smaller-bodied mammals (Entwistle and Boris Kryštufek, Slovenian Museum of Natural History, Prešernova 20, Stephenson 2000). Despite this plain evidence, rodents 1000, Ljubljana, Slovenia, E-mail: [email protected] receive disproportionate low attention with respect to Omar F. Al-Sheikhly, Department of Biology, University of Baghdad, conservation activity and research (Young 1994), which Baghdad, Iraq, E-mail: [email protected] translates into a lack of knowledge on their status and Javier Lazaro, Max Planck Institute for Animal Behavior Am Obstberg 1, 78315 Radofzell, Germany, E-mail: [email protected] conservation needs. In this review, we address one of those Mukhtar K. Haba, University of Baghdad, Baghdad, Iraq, poorly studied species, the long-tailed nesokia, Nesokia E-mail: [email protected] bunnii (Khajuria 1981), which is known from a handful of Rainer Hutterer, Zoologisches Forschungsmuseum Alexander Koenig, individuals collected in the wetlands at the confluence of Adenauerallee 160, 53113 Bonn, Germany, the Tigris and the Euphrates Rivers, where the mytholog- E-mail: [email protected] Sayed B. Mousavi, Khuzestan, Ahvaz, Iran, ical Garden of Eden was located by most scholars (Albright E-mail: [email protected] 1922).

Open Access. © 2020 Boris Kryštufek et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License. 104 B. Kryštufek et al.: A forgotten rodent from the Garden of Eden

The first two individuals of the long-tailed nesokia tailed and the short-tailed nesokia rats, concluding that were collected on 23 March and 16 November 1974, Erythronesokia is “not more than – if at all – a subgenus of respectively, and served as the type and paratype for Nesokia, but not a separate genus.” Similar were conclu- Erythronesokia bunnii, a new genus and species (Khajuria sions from a geometric morphometric approach of cranio- 1981). Both specimens were deposited in the mammal dental structures (Kryštufek et al. 2016) and the species is collection at the Natural History Research Centre and now regarded to be one of the two extant species of Nesokia Museum, University of Baghdad, Baghdad (NHRCM), (Denys et al. 2017). The evolutionary history of the long- which was vandalized in 2003 during the war in Iraq (Al- tailed nesokia rat can only be deduced from circumstantial Sheikhly et al. 2015). In 2016, Kryštufek et al. (2017) sur- evidence. The earliest appearance of Nesokia in the fossil veyed the still existing NHRCM small mammal collection record is at about 2 mya in Siwalik, India (Patnaik 2014), for E. bunnii but failed in finding a single voucher spec- and a molecular dating suggests a divergence from Ban- imen. It was therefore concluded that the type material of dicota (as a sister genus) at 1.2 mya (confidence inter- E. bunnii had been ultimately lost. Further three adults val = 0.8–1.6 mya; Fabre et al. 2013). One can expect the were collected between May 1974 and January 1977 and origin of the long-tailed nesokia rat to have occurred in the were published by Al-Robaae and Felten 1990. They are late Middle Pleistocene or early Upper Pleistocene. now in the Naturmuseum Senckenberg, Frankfurt, Ger- The range of the long-tailed nesokia rat is one of the many (SMF), and were described by Kryštufek et al. (2017). smallest among the Palaearctic mammals. All reliable As far as to our knowledge, no further museum vouchers records of occurrence cluster within a narrow perimeter in are known to exist. the Basra province inside the seasonally flooded Meso- The description provided by Khajuria (1981) and the potamianmarshesinsouthernIraq.Inthe1990s,this subsequent reclassification of E. bunnii relied entirely on extensive aquatic habitat, originally covering 15–20 morphology. The species shares with the short-tailed thousand km2, has been deliberately reduced to <15% of nesokia N. indica (Gray, 1832) all traits that distinguish its original area (Al-Ansari et al. 2012). The prevailing Nesokia from the closely related Bandicota. The two species opinion claimed that such an extensive ecosystem of nesokia rats differ in colour, size, relative length of tail, destruction “… almost certainly lead to the global and shape of skull and mandible (Kryštufek et al. 2016, extinction” (UNEP 2001) of two mammals, endemic to 2017). True mice and rats (Murinae) abound with cryptic these marshes, the Iraqi smooth-coated otter Lutrogale species, which are so morphologically similar that they can perspicillata maxwelli Hayman, 1956, and the long-tailed only be discerned by molecular markers. The long-tailed nesokia rat. Attempts initiated in 2003 towards restora- nesokia rat is clearly an exception to this general similarity. tion of the original habitat reversed the shrinking trend It is so distinct from N. indica and the bandicoot rats that its and marshes started to expand at the rate of 800 km2 per species status was nearly unequivocally accepted by all year (Al-Ansari et al. 2012). While the smooth-coated otter, subsequent authors (Figure 1). The generic status of along with the Eurasian otter Lutra lutra (Linnaeus, 1758) Erythronesokia was retained in 1980s and more hesitantly escaped extinction (Al-Sheikhly et al. 2017; Moretti et al. in the 1990s. In a craniometric study, Al-Robaae and Felten 2017) the fate of the long-tailed nesokia rat is more per- 1990 demonstrated a close resemblance between the long- plexing. The animal was last seen in 1976, i.e. long before any drainage occurred. The IUCN Red List assessed the current conservation status of the species as Endangered (Stuart 2008). With this study we aim to set a baseline for further work on this elusive rat in the Mesopotamia marshes of southern Iraq. Our goal is twofold. First, we assess the habitat suitability for the long-tailed nesokia rat in the marshes of Iraq to get an idea of the extension of the appropriate environment of the animal. Second, we explore all possible sources of information to trace any report which might signalize the occurrence of the animal. At last but not least, we would like to raise interest

Figure 1: Colour drawing of long-tailed nesokia rat Nesokia bunnii, on this extremely poorly known mammal which, despite based on museum vouchers in the Naturmuseum Senckenberg. its threatened conservation status, attracts so little Digital painting by Javier Lazaro. interest. B. Kryštufek et al.: A forgotten rodent from the Garden of Eden 105

2 Materials and methods This map served in the habitat suitability/species distribution- modelling step as a weight. Finally, the Weighted Mahalonobis distance was applied to develop the final N. bunnii habitat suitability map. 2.1 Data collection

We examined existing museum vouchers of N. bunnii, i.e. three skins and skulls held in the mammal collection of the Senckenberg For- 3 Results and discussion schungsinstitut und Naturmuseum Frankfurt a. M., Germany (SMF 62925, 87531, 87532). We recorded the entire information associated 3.1 Distributional records and habitat with these vouchers, both from specimen tags and card files. Furthermore, we performed a thorough search in literature sources suitability and on the internet, focussing on information on marshes in general and on small mammals in particular. Over the last several years we set Only four records of the long-tailed nesokia rat are known live traps in habitat suitable for the target species, made observations and they are positioned within a radius of 30 km around on the rodents and their tracks, and interviewed Marsh Arabs (indig- Qurna (Figure 2). These localities are: “Basra” (Khajuria enous inhabitants of the marshes) on their possible sightseeing of an 1981); Saraifa, 30 km north of Qurna; Bani Mansor, 25 km unusual rat. We informed them about the animal’s appearance and š encouraged them to report any encounters and, if possible, document west of Qurna; 5 km north of Qurna (Kry tufek et al. 2017). the sightseeing with a photograph. According to the information on the SMF specimen tags, all vouchers were captured in the marshland. This record concurs well with the mandibular morphology of the 2.2 Habitat modelling long-tailed nesokia which is different from fossorial and Catholic species of Nesokia and Bandicota (Kryštufek In order to model the habitat suitability for N. bunnii, the following et al. 2016), allegedly a consequence of its presumed methodological steps were applied: (1) identification of environmental predictor variables, (2) collection and preparation of environmental aquatic life style (Al-Robaae and Felten 1990; Harrison predictors, (3) species range polygon refinement and (4) habitat suit- and Bates 1991). ability/species distribution modelling. The IUCN polygon for the species’ range estimated the By leaning on the existing literature about Nesokia bunnii’s range to be more extensive, covering the Central (Qurna) habitat preferences, the Global Lakes and Wetlands Database and the Hammar Marshes, the Euphrates–Tigris river sys- Level 3 (GLWD-3) geospatial data were obtained (https://www. worldwildlife.org/pages/global-lakes-and-wetlands-database). tem, between Basra and Al Kut near Baghdad, i.e. an area 2 Next, the long-time smoothed NDVI values corresponding to the of approximately 35,800 km . Our results suggest an even mean NDVI over the period from 1999 to 2012 for summer (June 18) more extensive suitable habitat, covering the majority of were purchased from the Land Cover CCI Climate Research Data Mesopotamian marshland in Iraq, extending for approxi- Package (CRDP) which is a product developed by the Climate Change mately 600 km between Basra and Salah Ad Din provinces. Institute of the European Space Agency (ESA CCI) (http://maps.elie. ucl.ac.be/CCI/viewer/index.php). As human interference in the A wide-ranging habitat fragment was also retrieved in the study area was high, owing to wetland draining in the last decades, Hawizeh Marshes between the Tigris and Karun Rivers in the NDVI variable, gathered from freely available LANDSAT satellite the Iranian Khuzestan. As suggested by the habitat suit- imagery (https://earthexplorer.usgs.gov/), was used as a proxy of ability model, the eastern border might be set by the landscape change. Thus, the NDVI difference between 1995 (June) mountain ranges of Zagros and Kabir Kuh while the west- and 2018 (June) was calculated and used as the next environmental predictor for the N. bunnii habitat suitability modelling. Moreover, ern border putatively coincides with the Syrian Desert. The bioclimatic data, purchased from the WorldClim database (Version highly suitable habitat (habitat suitability >0.75) covers an 2.0), were used as well (http://www.worldclim.org/). Here the tem- estimated 15,650 km2, and is in two major fragments, in the perature and precipitation derived from bioclimatic variables were north (around Samarra) and south (south-eastern Iraq and considered separately. However, all were z-standardized and PCA adjacent Iran). These putative fragments are separated by transformed. Three PCA components (explaining 90% of variability in temperature derived bioclimatic properties) and three PCA com- less suitable habitat (habitat suitability <0.25) in a distance ponents (explaining 87% of variability in precipitation derived of approximately 130 km. bioclimatic properties) were added to the model. The categorical Needless to say, our habitat suitability model has to be wetland database was transformed to a distance variable by using taken with caution. First, the actual sites of occurrence of the Euclidean distance algorithm, in order to meet the requirements the long-tailed nesokia rat are nearly half a century old. of the habitat suitability model incorporated in the TerrSet geospatial software (Eastman 2018). Next, their position could not be geo-referenced with a Prior to habitat suitability modelling, the above mentioned pre- desired precision. Finally, the cluster of positive sites dictor variables and the IUCN range polygon were used to calculate a covers only a small proportion of potentially suitable confidence map in the species range polygon refinement procedure. habitat. 106 B. Kryštufek et al.: A forgotten rodent from the Garden of Eden

Figure 2: Habitat of the long-tailed nesokia rat N. bunnii in Mesopotamian seasonally ﬂooded marshes in Iraq (A and B; photos by Mukhtar K. Haba and Omar F. Al-Sheikhly) and habitat suitability map for the species based on selected geospatial predictor variables. Blue dots indicate sample sites and the red-outlined polygon represents the IUCN N. bunnii distributional range. The habitat suitability is coloured with cyan (low suitability >0.25) to red colour pallet (high suitability >0.75) for the target species. The position of the area on the map of south–west Asia is shown in the upper right corner of the inset C. B. Kryštufek et al.: A forgotten rodent from the Garden of Eden 107

Figure 3: The two most abundant rat-sized rodents in the Mesopotamian seasonally flooded marshes are (A) the short-tailed nesokia (Nesokia indica; photographed at Hoor Ol-Ezeem on the Iranian side of Hawi- zeh Marsh; photo by Sayed B. Mousavi) and (B) the brown rat (Rattus norvegicus; captured at Mashab in the Eastern Hammar Marsh, Basra Province, Iraq; photo by Omar F. Al-Sheikhly). Compare with the long- tailed nesokia rat in Figures 1, 4B.

3.2 Is the long-tailed nesokia still present in short-tailed nesokia rat which resembles its long-tailed the marshes? relative in rusty dorsal fur (Figure 3A) but shows faint demarcation on ﬂanks between the dorsal and the ventral As already mentioned, the last record of the long-tailed colour and has more delicate feet. The brown rat is nesokia rat is from 1977. In 2007 Haba (2009) surveyed brownish grey (Figures 3B, 4C). The photograph by Haba Mesopotamian marshland mammals, mainly by relying on is so far the strongest indication on the presence of the visual observations of animals or tracks. These techniques long-tailed nesokia rat after the last museum voucher had are not the most suitable for detecting a nocturnal murine been captured near Bani Mansor on 2 January 1977 š and indeed, the survey failed to conﬁrm any evidence of the (Kry tufek et al. 2017). The carcass was found in the ’ long-tailed nesokia rat. After the 2007 survey, we Lissan E jeardah Marsh of the Mysan province, i.e. continued monitoring the putatively suitable habitats in northward of Basra and Qurna and close to the 1970s the marshes by live-trapping and interviewing local in- trapping sites of the species. This leftover was unfortu- habitants. The only large murines we recorded were the nately not preserved. brown rat Rattus norvegicus and the short-tailed nesokia rat Another effort to record the long-tailed nesokia rat is (Figure 3). worth mentioning. The late Jamshid Darvish, a leading During this 2007 survey, Haba published a photo- rodent expert in Iran, informed us that his group searched graph of a half-eaten carcass which was not identified to for the long-tailed nesokia rat in the Khuzestan of Iran, but thespecies,anditwaslabelledasa“Rat” (Figure 13 in their attempts yielded no positive results (personal Haba 2009). Only the sacral region of the body was pre- communication to BK, on 25 February 2011). served, along with hind feet and the tail (Figure 4A). The As shown in this study, the suitable habitat for the most obvious features of this leftover, clearly seen long-tailed nesokia rat might be much more extensive than despite the low resolution of the photograph, are presumed ever before (Figure 2). The proposed range is not (i) robust feet, (ii) tail with a thick base which fairly unlikely as long as one keeps in mind that two otter-species rapidly tapers towards the tip, and (iii) the ochre red mentioned above are widely present in this area (Al- dorsal pelage which is clearly demarcated from the dirt Sheikhly and Nader 2013). We hope that the perspective of white underside. The pelage matches that of the long- surveying a broad area will strengthen interest of conser- tailed nesokia rat in every detail (cf. Figures 1, 4B). The vation groups on the status of the long-tailed nesokia rat whitishundersideandrobustfeetdifferfromthatofthe throughout the Garden of Eden.

Figure 4: Posterior part of the “rat” carcass (A) found on January 2007 in the marshes of Mysan province, Iraq, compared to the long-tailed nesokia rat N. bunnii (B; Natur- museum Senckenberg, voucher SMF 87531) and brown rat Rattus norvegicus (C; Zoo- logisches Forschungsmuseum Alexander Koenig, voucher ZFMK-MAM-1979.760). Photos by Mukhtar K. Haba (A) and Boris Kryštufek (B, C). 108 B. Kryštufek et al.: A forgotten rodent from the Garden of Eden

Acknowledgments: The authors are grateful to Irina Ruf the pand had its day? Cambridge Univ. Press, Cambridge, and Katrin Krohmann (Senckenberg Forschungsinstitut pp. 120–139. und Naturmuseum Frankfurt a. M., Germany) who enabled Fabre, P.-H., Pagès, M., Musser, G.G., Fitriana, Y.S., Fjeldsä, J., Jennings, A., Jønsson, K.A., Kennedy, J., Michaux, J., Semiadi, G., access to specimens. Supriatna, N., Helgen, K.M. (2013). A new genus of rodent from Author contribution: All the authors have accepted Wallacea (Rodentia: Muridae: Murinae: Rattini), and its responsibility for the entire content of this submitted implication for biogeography and Indo-Pacific Rattini manuscript and approved submission. systematics. Zool. J. Linn. Soc. 169, 408–447. Research funding: Funding for this research was provided Haba, M.K. (2009). Mesopotamian marshland mammals. Marsh Bull. 4: 179–189. by the Slovenian Research Agency (Grant P1-0255 for B.K. Harper, F. (1945). Extinct and vanishing mammals of the Old World. and Grant P6-0372 for D.I.). Special Publication No. 12. American Committee for International Conflict of interest statement: The authors declare no Wild Life Protection. New York Zoological Park, New York, conflicts of interest regarding this article. pp. 1–850. Harrison, D.L., and Bates, P.J.J. (1991). The mammals of Arabia. Harrison Zoological museum Publ, Sevenoaks, pp. 1–354. IUCN. (2017). The IUCN red list. Available at: https://www.iucnredlist. References org/. Downloaded on 18 February 2020. Khajuria, H. (1981). A new bandicoot rat, Erythronesokia bunnii gen. et Al-Ansari, N., Knutsson, S., and Ali, A. (2012). Restoring the Garden of sp. nov. (Rodentia: muridae), from Iraq. Bull. Nat. Hist. Res. Eden. Iraq. J. Earth Sci. Geotech. Eng. 2: 53–88. Centre Baghdad 7: 157–164. Albright, W.F. (1922). The location of the Garden of Eden. Am. J. Semitic Kryštufek, B., Al-Sheikhly, O.F., and Hutterer, R. (2017). A Lang. Lit 39: 15–31. redescription of the long-tailed Nesokia, Nesokia bunnii, and Al-Robaae, K., and Felten, H. (1990). Was ist Erythronesokia Khajuria, designation of a neotype (rodentia: muridae). Zootaxa 4216: 1981 (Mammalia: Rodentia: Muridae)? Z. Säugetierkd 55: 253–259. 167–187. Al-Sheikhly, O.F., and Nader, I.A. (2013). The status of Iraq smooth- Kryštufek, B., Janžekovič, F., Hutterer, R., and Klenovšek, T. (2016). coated otter Lutrogale perspicillata maxwelli Hayman 1956 and Morphological evolution of the skull in closely related bandicoot Eurasian otter Lutra lutra Linnaeus 1758 in Iraq. IUCN Otter Spec. rats: a comparative study using geometric morphometrics. Group Bull. 30: 18–30. Hystrix It. J. Mamm 27, https://doi.org/10.4404/hystrix-27.2- Al-Sheikhly, O.F., Haba, M.K., Barbanera, F., Csorba, G., and Harrison, 11639. D.L. (2015). Checklist of the mammals of Iraq (chordata: Moretti, B., Al-Sheikhly, O.F., Guerrini, M., Theng, M., Gupta, B.K., mammalia). Bonn Zool. Bull. 64: 33–58. Haba, M.K., Khan, W.A., Khan, A.A., and Barbanera, F. (2017). Al-Sheikhly, O.F., Haba, M.K., Faaza, N.A., Barbanera, F., Yoxon, G., Phylogeography of the smooth coated otter (Lutrogale McLennan, B., Zook, D.B., Al-Kanani, K.A., and Abdul Gafor, S.A. perspicillata): distinct evolutionary lineages and hybridization (2017). First photographic evidence of smooth-coated otter with the Asian small clawed otter (Aonyx cinereus). Sci. Rep. 7: (Lutrogale perspicillata maxwelli) and Eurasian otter (Lutra lutra) 41611. in Iraq since 1950s. Otter 3: 15–20. Patnaik, R. (2014). Phylogeny of Siwalik murine rodents: Aplin, K.P., Brown, P.R., Jacob, J., Krebs, C.J., and Singleton, G.R. implications for Mus-Rattus divergence time. J. Palaeont. Soc. (2003). Field methods for rodent studies in Asia and the Indo- India 59: 15–28. Pacific. ACIAR Monograph No. 100. Australian Centre for Stuart S.N. 2008. Nesokia bunnii. The IUCN Red List of Threatened International Agricultural Research, Canberra, pp. 1–60. Species 2008: e.T14660A4453417. Available at: https://doi.org/ Ceballos, G., and Brown, J.H. (1995). Global patterns of mammalian 10.2305/IUCN.UK.2008.RLTS.T14660A4453417.en. Downloaded diversity, endemism, and endangerment. Cons. Biol. 9: 559–568. on 18 February 2020. Denys, C., Taylor, P., and Aplin, K.P. (2017). Family Muridae (True mice Turvey, S.T. (2009). Holocene mammal extinctions. In: Turvey, S.T. and rats, gerbils and relatives). In: Lacher, T.E., and Mittermeier, (Ed.). Holocene extinctions. Oxford Univ. Press, Oxford, R.A. (Eds.). Handbook of the mammals of the World, Vol. 7. D.E. pp. 41–62. WilsonRodents II. Lynx Edicions, Barcelona, pp. 536–884. UNEP. (2001). The mesopotamian marshlands: demise of an Eastman, R. (2018). TerrSet. Clark University, Worcester, MA. ecosystem. Early Warning and Assessment Technical Report, Entwistle, A.C., and Stephenson, P.J. (2000). Small mammals and UNEP/DEWA/TR.01–3 Rev. 1. Nairobi, Kenya. the conservation agenda. In: Entwistle, A.C., and Dunstone, N. Young, T.P. (1994). Natural die-offs of large mammals: implications for (Eds.). Prirorities for conservation of mammalian diversity: has conservation. Cons. Biol. 8: 410–418.