Multiple lines of evidence confirm that Hume's Owl Strix butleri (A. O. ...

Zootaxa 3904 (1): 028–050 www.mapress.com /zootaxa / Copyright © 2015 Magnolia Press

Article

ISSN 1175-5326 (print edition)

ZOOTAXA

ISSN 1175-5334 (online edition)

http://dx.doi.org/10.11646/zootaxa.3904.1.2 http://zoobank.org/urn:lsid:zoobank.org:pub:FE08BF4A-EB0A-4AA0-A933-5B01449187CE

Multiple lines of evidence confirm that Hume’s Owl Strix butleri (A. O. Hume, 1878) is two species, with description of an unnamed species (Aves: Non-Passeriformes: Strigidae) GUY M. KIRWAN1, MANUEL SCHWEIZER2,4 & JOSÉ LUIS COPETE3 1

Research Associate, Field Museum of Natural History, 1400 South Lakeshore Drive, Chicago, IL 60605, USA. E-mail: [email protected] 2 Naturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland. 3 c/o Martínez de la Rosa 27 Principal 3ª, 08012 Barcelona, Spain. E-mail: [email protected] 4 Corresponding author: E-mail: [email protected]

Abstract Genetic and morphological analyses revealed that the type specimen of Hume’s Owl Strix butleri, the geographical provenance of which is open to doubt, differs significantly from all other specimens previously ascribed to this species. Despite the absence of vocal data definitively linked to the same population as the type specimen, we consider that two specieslevel taxa are involved, principally because the degree of molecular differentiation is close to that seen in other taxa of Strix traditionally recognised as species. Partially complicating this otherwise straightforward issue is the recent description of “Omani Owl S. omanensis” from northern Oman based solely on photographs and sound-recordings. We consider that there is clear evidence of at least some morphological congruence between the butleri type and the phenotype described as “omanensis”. As a result, we review the relative likelihood of three potential hypotheses: that “omanensis” is a synonym of butleri; that “omanensis” is a subspecies of butleri; or that “omanensis” and butleri both represent species taxa. Until such time as specimen material of “omanensis” becomes available for genetic and comparative morphological analyses, we recommend that this name be considered as a synonym of butleri, especially bearing in mind the possibility (not previously considered in detail) that the type of butleri could have originated in Arabia, specifically from Oman. We describe other populations heretofore ascribed to S. butleri as a new species. Key words: taxonomy, mitochondrial DNA, Strix butleri, Strix omanensis

Introduction The range of Hume’s (Tawny) Owl Strix butleri encompasses eastern Egypt, through Sinai, southern and eastern Israel, Palestinian Territories, Jordan and much of the Arabian Peninsula (Mikkola 2012, Dickinson & Remsen 2013; Fig. 1). In addition, there are two possible records on the island of Socotra, off the Horn of Africa (Jennings 2010: 423), and the type specimen was originally believed to have come from “Omara, on the Mekran Coast” (Hume 1878: 316), in what is now southern Pakistan (= Ormara, 25o10’N, 64o34’E; Fig. 1), although some authors have erroneously placed Ormara in southeastern Iran (Goodman & Sabry 1984). Despite being known now to be comparatively widespread, e.g., occurring in at least 33 30-minute × 30-minute quadrants in Arabia (Jennings 2010) and at many localities in Israel (Shirihai 1996), Hume’s Owl was virtually unknown until rather recently (Hüe & Etchécopar 1970). Indeed, the first-ever specimen of butleri (now in the Liverpool World Museums), collected by C. W. Wyatt in Wadi Feiran, Sinai (28o43’N, 33o35’E), in 1864, lay unnamed for more than a decade, until Hume had published his description of butleri (Tristram 1879). There was just one definite record in Arabia prior to the 1970s—a specimen collected in 1950 (Meinertzhagen 1954)—and the species was known from just three atlas squares on the peninsula pre-1984, in western and central Saudi Arabia and southwest Oman (Jennings 2010: 423, although two other records are listed by Goodman & Sabry 1984, and Jennings himself mentioned an aural-only record in February 1948); none in Egypt outside Sinai prior to 1982 (Goodman & Sabry 1984) and just

28 Accepted by P. Rasmussen: 21 Oct. 2014; published: 5 Jan. 2015

one there pre-1997 (Baha el Din & Baha el Din 2001); none in Jordan before 1979 (Andrews 1995); and, although the species is arguably best known in what is now Israel (Shirihai 1996), even here it was known from only c. 10 specimens prior to the 1970s (Mendelssohn et al. 1975). A virtually complete list of specimens was presented by Goodman & Sabry (1984); we have personally examined just one additional example, sent by Aharoni to the Museum für Naturkunde, Berlin, and collected in the environs of Jerusalem, Israel (see Table 1), but a pair sent to Alexander Koenig, also by Aharoni, is now in the Zoologische Forschungsmuseum Alexander Koenig, Bonn (ZFMK). Hume’s type of Strix butleri, now at the Natural History Museum, Tring (BMNH 1886.2.1.994) was, as made clear by its author, originally obtained by a Mr Nash, a friend of Captain (subsequently Lieutenant-Colonel) Edward Arthur Butler, who gave the specimen to Hume (1878: 316). It is obvious that Hume did not know Nash personally and information about him is scarce. Meinertzhagen (1930) described the specimen in question as “a miserable skin, dirty and shockingly prepared”. Ticehurst (1927) stated that Nash was based in Ormara “for many years” and, in a footnote, Butler (1877: 300) mentioned that “through the kindness of G. Nash, Esq., Telegraph Department, who sent a canoe from Ormarra [sic] to Astola about the 19th June, I have received another beautiful series of eggs of Sterna bergii [Swift Tern Thalasseus bergii]”. Since Nash was receiving specimens collected by others, as is clear from this statement, it must be possible that some of these could have originated from across the Gulf of Oman. As early as the 1920s, given the lack of further records from what is now Pakistan, Ticehurst (1927) speculated that the species might only be a vagrant there. Much later, Goodman & Sabry (1984) considered that the continued lack of records in southeastern Iran (where they erroneously placed Ormara), coupled with Hume’s lack of specific designation of a type locality because he knew only where Nash was based, not where the specimen had been obtained, to be sufficient rational to doubt that Ormara really was the type locality. As a result, the species has been variously accepted (e.g., by Roberts 1992) or treated as hypothetical (Rasmussen & Anderton 2005) in the avifauna of Pakistan, although Roberts also noted that few ornithologists have spent much time in Ormara region, especially since the 1920s, whilst other authorities have remarked on the lack of records from neighbouring Iran (e.g., Scott & Adhami 2006, Khaleghizadeh et al. 2011), other than a questionable record from the Mekran coast in 1920 (Scott 2008: 30). Most recently, Roselaar & Aliabadian (2009: 26) postulated “No bird has been found again on the Makran coast, and likely the type specimen was brought to Ormara over sea from Arabia.” The fact that Nash probably had close connections to the port of Ormara, rather than being merely based in a port city, and that various parts of the Mekran [= Makran, i.e. southwest Pakistan and extreme southeast Iran] were governed by the Sultanate of Oman, from Muscat, from 1783 until Pakistan became independent in 1947 (with the enclave of Gwadar remaining under Omani sovereignty until 1955), only increases that possibility. It is perhaps interesting to note that Tristram (1880: 246) stated that the type specimen emanated from “the Mekran coast of Arabia”, although here the possibility of a lapsus seems strong. In 1985, during a visit to the Natural History Museum, Tring while preparing The Birds of Israel (Shirihai 1996), Hadoram Shirihai noted that the type of Strix butleri differed in several respects from other specimens of the same species held in the same museum and at institutions in Israel, as well as with the many live birds he had observed in the latter country. Subsequently, together with the late Prof. Heinrich Mendelssohn, Shirihai had planned to name the Israeli and related populations as a new subspecies of butleri, but due to the death of his coauthor and many other commitments this never happened. Most recently, Robb et al. (2013) described a new species in the Strix butleri complex from the Al Hajar Mountains in northern Oman based solely on photographs and sound-recordings of several individuals, of which one was designated as the type. They named their species “Omani Owl Strix omanensis” and, despite the lack of any specimens or genetic material, we believe that these authors conclusively demonstrated that “omanensis” differs sufficiently in voice and, to a lesser extent, in morphology to be treated as specifically different from populations of butleri elsewhere in the Middle East. Until this remarkable discovery, butleri was known in Oman solely from the southwest of the country in Dhofar (Jennings 2010, Eriksen & Victor 2013), contra the maps in König et al. (1999, 2008) as already noted by Robb et al. (2013: 306). To test whether Shirihai’s idea that a new taxon was involved and to evaluate the validity of the newly described “omanensis”, we re-examined the morphology of the type of butleri, as well as other specimens of butleri in several museums, and investigated their identification and relationships using genetic data.

AN UNNAMED SPECIES OF STRIX (AVES)

Zootaxa 3904 (1) © 2015 Magnolia Press ·

29

30 · Zootaxa 3904 (1) © 2015 Magnolia Press

KIRWAN ET AL.

BMNH 1886.2.1.994

BMNH 1965.M.5235

TAUM 9812

TAUM 10219

BMNH 1965.M.5236

ZMB 31/2017

ZM-HIJ 1468

ZM-HIJ 819

ZM-HIJ 541

ZM-HIJ 817

AMNH 631.939

AMNH 631.940

TAUM 10373

TAUM 10802

TAUM 12161

TAUM 15603

TAUM 16150

TAUM 9428

TAUM 4928

Strix butleri holotype

Strix hadorami holotype

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Mus. no.

Taxon

Hebrew University of Jerusalem.

female

male

male

male?

male

male

female

female

female

?

female

female

male

male

female

Sex

17.01.1963

31.05.1978

09.10.2005

29.09.1972

05.03.1998

21.01.1997

21.07.1980

30.11.1918

30.11.1918

17.05.1970

unknown

unknown

07.01.1973

22.03.1931

16.08.1950

unknown

27.02.1982

28.03.1928

Date

244.0

253.0

248.0

239.0

249.0

221.0

230.0

249.0

252.0

Wing

En Mor, Israel

W Mashash/W Halhul, Israel

Dead Sea, Israel

En Gedi, Dead Sea, Israel

Exp. Zoo, Israel

Exp. Zoo, Israel

Dead Sea, Israel

Wadi Suenit, Jordan

Wadi Suenit, Jordan

240.0

244.0

247.0

242.0

245.0

248.0

231.0

242.0

247.0

St Catharine's monastery, Sinai Mts, Egypt 242.0

near Jerusalem, Israel

-


Wadi Fauwar, Jerusalem, Israel

Mahd Dhahab, Jiddah, Saudi Arabia

Israel

Near Hakikar, Arava, Israel

Wadi Kelt, Israel/Palestinian Territories

Ormara, Mekran coast, Pakistan?

Locality

-

143.0

130.0

-

137.0

135.0

130.0

132.0

134.0

145.0

136.0

134.0

139.0

128.0

136.0

-

-

142.0

135.0

Tail

-

58.6

52.6

-

55.9

54.4

56.2

54.5

54.3

59.9

55.7

53.0

56.0

52.9

54.6

-

-

57.0

53.6

Tail / Wing

-

10.0

12.0

-

-

10.0

-

9.0

12.0

12.0

5.0

7.0

10.0

7.0

5.0

-

-

9.0

15.0

Tail grad.

-

52.3

44.7

49.1

43.7

47.1

45.6

47.0

43.0

48.2

44.8

45.4

49.9

45.2

47.0

-

-

45.3

53.7

Tarsus

14.9

16.4

14.2

15.4

14.5

14.1

-

19.3

17.1

-

-

15.9

15.4

14.9

-

-

16.2

-

Bill depth

55.0

50.0

55.0

50.0

56.0

46.0

43.0

53.0

46.0

52.0

55.0

53.0

49.0

53.0

-

-

53.0

50.0

p1

……continued on the next page

25.5

27.4

26.3

27.8

26.9

25.6

24.5

28.2

24.9

26.9

25.3

26.8

25.1

27.9

-

-

24.5

28.0

Bill length

AMNH = American Museum of Natural History, New York; BMNH = Natural History Museum, Tring; TAUM = Tel Aviv University Museum; ZMB = Museum für Naturkunde, Berlin; ZM-HIJ = Zoological Museum,

G.M.K., except those specimens from AMNH, which were taken by L. Svensson.

TABLE 1. Museum registration numbers, collecting information and measurements (in mm) for the specimens analysed. Note that lengths of primaries are given relative to the wingtip. All measurements taken by



31

TAUM 10219

BMNH 1965.M.5236 male

ZMB 31/2017

ZM-HIJ 1468

ZM-HIJ 819

ZM-HIJ 541

ZM-HIJ 817

AMNH 631.939

AMNH 631.940

TAUM 10373

TAUM 10802

TAUM 12161

TAUM 15603

TAUM 16150

TAUM 9428

TAUM 4928

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami

Strix hadorami female

male

male

male?

male

male

female

female

female

?

female

female

17.01.1963

31.05.1978

09.10.2005

29.09.1972

05.03.1998

21.01.1997

21.07.1980

30.11.1918

30.11.1918

17.05.1970

unknown

unknown

07.01.1973

22.03.1931

16.08.1950

unknown

27.02.1982 male

28.03.1928

TAUM 9812

Date

Strix hadorami

BMNH 1886.2.1.994


Sex

Strix hadorami holotype BMNH 1965.M.5235 female

Mus. no.

Taxon

TABLE 1. (Continued)

19.0

15.0

13.0

14.0

16.0

-

-

15.0

14.0

p2

En Mor, Israel

W Mashash/W Halhul, Israel

Dead Sea, Israel


Exp. Zoo, Israel

Exp. Zoo, Israel

Dead Sea, Israel

Wadi Suenit, Jordan

Wadi Suenit, Jordan

-

19.0

11.0

17.0

15.0

17.0

11.0

12.0

13.0

St Catharine's monastery, Sinai Mts, Egypt 12.0

near Jerusalem, Israel

-


Wadi Fauwar, Jerusalem, Israel

Mahd Dhahab, Jiddah, Saudi Arabia

Israel

Near Hakikar, Arava, Israel

Wadi Kelt, Israel/Palestinian Territories

Ormara, Mekran coast, Pakistan?

Locality

-

4.0

2.0

3.0

7.0

0.0

4.0

-

-

4.0

3.0

2.0

0.0

2.0

3.0

-

-

8.0

0.0

p3

-

0.0

0.0

0.0

0.0

3.0

0.0

-

-

0.0

0.0

0.0

0.0

0.0

0.0

-

-

0.0

2.0

p4

-

11.0

5.0

7.0

0.0

6.0

0.0

6.0

5.0

6.0

3.0

3.0

5.0

8.0

3.0

-

-

7.0

9.0

p5

-

-

26.0

24.0

20.0

24.0

22.0

25.0

26.0

22.0

27.0

20.0

26.0

26.0

24.0

-

-

27.0

35.0

p6

-

-

43.0

44.0

40.0

45.0

40.0

45.0

46.0

41.0

43.0

42.0

42.0

41.0

45.0

-

-

49.0

54.0

p7

-

-

69.0

71.0

69.0

75.0

78.0

71.0

69.0

73.0

70.0

76.0

70.0

70.0

75.0

-

-

73.0

100.0

p10

-

89.0

65.0

65.0

60.0

85.0

71.0

68.0

64.0

61.0

77.0

72.0

85.0

76.0

79.0

-

-

81.0

88.0

s1

7/8

-

7/8

7/8

7/8

7/8

7/8

7

7/8

7/8

7/8

7/8

7/8

7/8

7/8

-

-

7/8

6/7

p1=

5/6

-

5/6

5/6

5/6

5/6

5/6

5/6

5/6

5/6

5/6

5/6

5/6

5/6

5/6

-

-

5/6

5/6

p2=

-

322.0

323.0

-

304.0

330.0

264.0

310.0

330.0

-

-

-

-

333.0

332.0

-

-

320.0

345.0

Total length

Material and methods Specimen and morphometric analyses. G.M.K. examined specimens of Strix butleri at the following museums: Natural History Museum, Tring (BMNH; n = 3, including BMNH 1886.2.1.994, the type), Museum für Naturkunde, Berlin (ZMB; n = 1), Tel Aviv University Museum (TAUM; n = 9, of which measurements were taken of seven), and Zoological Museum, Hebrew University of Jerusalem (ZM-HIJ; n = 4). Additionally, L. Svensson sent us photographs and measurements of the specimens in the American Museum of Natural History, New York (AMNH; n = 2). In addition to plumage comparisons, the following measurements were taken according to standard protocols (Svensson 1999) using dial callipers and two different-length wing-rules with a perpendicular stop at zero: wing length (from carpal joint to tip applying gentle pressure to the primary-coverts), tail length (from the distal end of the pygostyle to the tip), tail graduation (from tip of longest to tip of shortest rectrix), tarsus length (from the back of the intertarsal joint to the last complete scute before the toes diverge), bill length (from the tip of the maxilla to skull) and bill depth (at the distal edge of feathering), as well as the following measurements relating to wing formula: distance of primaries (pp) 1–7 and p10 to wingtip (note: primaries numbered ascendently, thus p1 = outermost), and distance of secondary 1 (s1) to wingtip, and the relative positions of p1 and p2 versus the other primaries. Note that the tarsus measurement is difficult to take accurately in this species because tarsal feathering typically covers the last complete scute, which means it must be felt for below the feathers. The three Tring specimens, including the butleri type, were also independently examined by J.L.C. We analysed morphometric differences using a multivariate approach. Wing, tail, tarsus and bill lengths, the length of the primaries and the outermost secondary relative to the wingtip were analyzed in individuals for which all measurements were available. The wingtip was variable, formed by p3 or p4 (Table 1), and we therefore included the distance between the tip of p3 or p4 and the tips of the remaining primaries (excluding pp8–9) in our analyses. We used the method of Baur & Leuenberger (2011), which permits the interpretation of principal components (PCs) as ratios, permitting a strict separation of differences in size and shape, which has been problematic with traditional multivariate approaches. The geometric mean of the original measurements was used to define an isometric size axis (“isosize”) (cf. Baur & Leuenberger 2011). Isosize is composed solely of differences in scaling, and size-independent shape variables are then obtained by projecting the measurements orthogonal to isosize. A PCA accounting exclusively for differences in proportions is then computed on the covariance matrix of the shape parameters. The “PCA ratio spectrum” (cf. Baur & Leuenberger 2011) is used to interpret the principal components in shape space. Genetic analyses. M.S. sequenced a partial fragment of the mitochondrial gene cytochrome b (cyt b) of the holotype of Strix butleri and another specimen usually considered to be this taxon from Mahd Dhahab, c. 250 miles north-northeast of Jeddah, western Saudi Arabia (Table 1). These were compared with cyt b sequences of three specimens of S. butleri from Israel and other species assigned to the genus Strix (Table 2). These sequences were collated from GenBank. Total genomic DNA from toe-pad samples of the two skins was isolated using sbeadex® forensic kit (LGC Genomics). To avoid potential contamination, DNA extraction and the preparation of PCR reactions took place in a different laboratory than that used for the treatment of fresh tissue and blood samples. Moreover, laboratory equipment used for preparing these samples was treated with DNA AwayTM (Molecular BioProducts) solution or irradiated with ultraviolet light prior to use. However, no fresh samples of any member of Strigiformes have ever been handled in the laboratories concerned, which eliminated the risk of contamination with DNA from closely related species. Partial sequences of cyt b were amplified with polymerase chain reaction (PCR) using primers specifically designed for this study (Table 3). Primers were designed with Primer3 0.4.0 (Rozen & Skaletsky 2000) as implemented as a plug-in in Geneious Pro 7.0.5 (Drummond et al. 2013). A negative control was used to check for potential contamination. PCR reaction volumes were 50 µl containing 25 µl QIAGEN Multiplex PCR Kit (Qiagen), 3 µl genomic DNA, 1µl of each primer with a concentration of 10µM and 20 µl ddH2O. PCR was performed on a SensoQuest Labcycler. The cycling conditions for amplification were: initial denaturation of 95°C for 15 minutes followed by 40 cycles of denaturation at 94°C for 30 seconds, annealing for 1.5 minutes at 57°C, and extension at 72°C for 1.5 minutes, with a final extension at 72°C for 10 minutes. PCR products were examined by gel electrophoresis to confirm amplification of the target fragments. Target fragments were then excised from gels and cleaned using the Wizard® SV Gel and PCR Clean-UP System (Promega). Sequencing was


KIRWAN ET AL.

carried out by Microsynth (Balgach, Switzerland) and LGC Genomics (Berlin, Germany). Sequence preparation and editing of sequences was performed with Geneious Pro. Individual sequences were checked by searching for apparent stop codons after the translation into amino acids. Alignment of the sequences (all of which were truncated to the same length) was performed using the MAFFT algorithm (Katoh et al. 2002) implemented as a plug-in in Geneious Pro using default settings. We constructed an unrooted median-joining network using the software Networks 4.6.2.1 (Bandelt et al. 1999) to illustrate similarities among haplotypes. To reconstruct phylogenetic relationship among the different taxa, we employed a maximum likelihood (ML) search using RAxML 7.2.8 (Stamatakis 2006, Stamatakis et al. 2008) implemented as a plug-in in Geneious Pro. We used the rapid bootstrapping option with 1000 replicates and search for the best-scoring ML tree with all free model parameters estimated by the software (substitution rates, gamma shape parameter, base frequencies). Rufous-legged Owl Strix rufipes was chosen as outgroup following the results of Wink et al. (2009). Corrected genetic distances were calculated in Geneious Pro based on the patristic distances of a neighbour-joining tree using the HKY distance model. TABLE 2. Specimens used for genetic analyses. Species

Voucher number

Locality

GenBank accession number

Strix butleri, holotype

BMNH 1886.2.1.994

Putatively Ormara, southwest Pakistan

KM459027

Strix hadorami

BMNH 1965.M.5236

Mahd Dhahab, NNE of Jeddah, Saudi Arabia

KM459028

Strix hadorami

Israel

AJ003912

Strix hadorami

Israel

AJ003913

Strix hadorami

Israel

EU348994

Strix aluco

AJ003907

Strix aluco

AJ003909

Strix aluco

AJ004052

Strix aluco

AJ004054

Strix aluco sylvatica

AJ004051

Strix aluco wilkonskii

AJ004045

Strix nebulosa

AJ004059

Strix nebulosa lapponica

AJ004058

Strix rufipes

AJ004060

Strix uralensis liturata

AJ004063

Strix uralensis macroura

AJ004062

Strix woodfordii

AJ004065

Strix woodfordii

AJ004066

Strix woodfordii nigricantior

EU348995

Strix woodfordii woodfordii

AJ004064

TABLE 3. Primers used for the amplification of the cyt b fragments analysed in this study. Forward

5'-TAGCCCAAATCATCACTGGC-3'

Reverse

5'-TGAAGAAGAGTGAGGCTCCG-3'

Forward

5'-ACTCCGCAACCTCCATGCAA-3'

Reverse

5'-AAGGCAGTTGCTATAAGAGT-3'



33

FIGURE 1. Map showing the range of Strix butleri (in this paper named S. hadorami), showing the type localities of S. butleri ( ), “S. omanensis” (X) and S. hadorami (•).

FIGURE 2. Left: Scatterplot of first against second principal component (PC1 and PC2) in shape space. Right: Scatterplot of isosize against first principal component in shape space. The holotype of Strix butleri is clearly separated in shape PC1 and isosize from the other specimens formerly assigned to this taxon and described here as a new species, Strix hadorami.


KIRWAN ET AL.

Results Morphometric analyses. The butleri type was clearly separated from other specimens in shape PC1 following the principal component analysis (PCA) performed in isometry- free shape space, although the two overlapped in shape PC2 (Fig. 2). According to the “PCA ratio spectrum”, shape PC1 was dominated by the ratio of p2 and p5 (Fig. 3). These two measurements were found not to be the most dominant ratios of the “allometry ratio spectrum” (Fig. 3) and can thus be considered to be only marginally affected by allometry, although there seems to be a slight correlation between shape PC1 and isosize (Fig. 2).

FIGURE 3. PCA ratio spectrum for the first principal component in shape space (PC1) and allometry ratio spectrum. Ratios calculated from characters lying far apart in the P C A r a t i o spectrum are responsible for a large part of the variance of a component, whereas ratios of characters lying close to each other in the spectrum contribute little to the total variance. The allometry ratio spectrum is interpreted accordingly. Shape PC1 is dominated by the ratio of p2 and p5, however, it is only marginally effected by allometry, as these two measurements are not the most dominant ratios of the allometry ratio spectrum.

Genetic analyses. We succeeded in sequencing two overlapping fragments leading to a sequence length of 218 base pairs (bp) of the holotype of Strix butleri and the specimen from Mahd Dhahab. Translation of the sequences into amino acids did not reveal any unexpected stop codons, and the final alignment for the phylogenetic analyses contained no indels. A second independent DNA extraction was performed using the remaining toe-pad material of the butleri type. The first fragment was then sequenced again and the previous results were confirmed. AN UNNAMED SPECIES OF STRIX (AVES)


35

The butleri type was found to be genetically clearly distinct from the published samples of S. butleri from Israel, whereas the specimen from Mahd Dhahab possessed an identical sequence to the latter. A GenBank BLAST search revealed the butleri type to have a unique haplotype (see below), so contamination with DNA from another taxon therefore seems highly unlikely. Pollution of the results due to the sequencing of nuclear mitochondrial paralogs (Numts) appears also unlikely (cf. Bensasson et al. 2001). Had this been the case, we would have expected stop codons or even reading frame shifts when translating the sequence into amino acids, and it could be expected that more than just one substitution led to amino acid changes in comparison with closely related taxa (see below). Moreover, the amplification of same-sized nuclear fragments is usually more problematic in rather old samples with degraded DNA, hence the amplification of a Numt could be expected to be less probable than the true mitochondrial fragment in our case. Finally, the sequence of the Mahd Dhahab specimen was identical to published sequences of S. butleri, demonstrating the specificity of the designed primers. The butleri type differed in 18 substitutions in 218 bp from the remaining S. butleri samples, however these led to just one amino acid change, i.e. a substitution of Leucine to Threonine at amino acid position 34 of the alignment. The genetic distinctiveness of the butleri type in relation to other taxa of the genus Strix is illustrated in the median-joining network (Fig. 4.). Corrected genetic distance between the butleri type and other samples assigned to the species was 10.6% (Table 4); they ranged from 10.9% to 13.7% between the butleri type and other Strix taxa, and 7.0–14.1% between our other samples of S. butleri and other Strix taxa. In the ML analysis, the butleri type was found to represent the sister lineage of a clade consisting of African Wood Owl S. woodfordii and the remaining samples of S. butleri (Fig. 5). The close relationship between the latter two was already recovered by Heidrich & Wink (1994) and Wink et al. (2009). However, not all relationships among species were supported in the ML tree.

New species description Genetic divergence between the butleri type and the other specimens usually ascribed to this species, from western Saudi Arabia and Israel, respectively, was equal to the range of divergence among well-accepted species in the same genus. The specimens from Israel and Saudi Arabia were even more similar and probably more closely related to African Wood Owl S. woodfordii (especially to S. w. nigricantior) than to the butleri type. However, a longer fragment of cyt b and, especially, additional genetic markers must be analyzed to reveal the phylogenetic relationship among taxa within the genus Strix with confidence. Nonetheless, genetic distances based on our fragment were of similar level to those computed using a longer fragment of 1041 bp for the same specimens except the butleri type, and can thus be considered broadly representative for cyt b. The magnitude of genetic differentiation between the butleri type and other specimens previously ascribed to butleri is not only within the range of that among species in the genus Strix, it also clearly exceeds the variation within several other owl species. Intraspecific genetic divergence in the mitochondrial gene cox1, which has a similar or even slightly higher substitution rate than cyt b (e.g. Pacheco et al. 2011), was less than 3% in 21 of 24 owl species analyzed (including five species of Strix) and reached a maximum of 4.09% in the geographically widespread Tengmalm’s Owl Aegolius funereus (Nijman & Aliabadian 2013). Wink et al. (2009) even claimed that cyt b distances exceeding 2% were typical of different species of owls and genetic distances between sister species of Otus were as low as 3.5% in the mitochondrial gene ND2, which evolves at a faster rate than cyt b. Although genetic divergence might be of limited value for species delimitation in some cases (see, for example, Price 2008, Fregin et al. 2012), the level of genetic distinctiveness of the butleri type indicates that it could represent a reproductively isolated entity, and is not inconsistent with species rank. This, coupled with the morphological (see below) and morphometric differences (see above) that we uncovered, indicates that a new species-level taxon is involved, which we name:

Desert Tawny Owl

Strix hadorami, sp. nov. Type locality. Lower Wadi Kelt (=Wadi Al-Qelt/Wadi Qilt), northeast of Jerusalem, Israel/Palestinian Territories (31o50’N, 35o24’E).


KIRWAN ET AL.

Holotype. Natural History Museum, Tring, UK (BMNH 1965.M.5235; Figs. 6–7, 10–11), female collected by B. Aharoni at Wadi Kelt, northeast of Jerusalem, Israel/Palestinian Territories, on 28 March 1938 (ex. Meinertzhagen collection). Three labels: oldest inscribed “Strix butleri Hume / ♀ Wadi-Kelt / 28.III.1938 / Legit. Dr B. Aharoni”; Meinertzhagen label: “Wadi Kelt Palestine 28.3.38 ♀” reverse “Strix butleri 1965.M.5235”; BMNH label: “Strix butleri Wadi Kelt, Palestine, 28 March 1938 ♀” reverse “Reg. no. 1965.M.5235 R. Meinertzhagen Coll. N”. See Notes for further details on collector and veracity. Diagnosis. Most obviously differentiated from the type of Strix butleri by the presence of pale rufous crossbarring over the breast and belly, contrasting clearly with the rest of the underparts even in very pale examples of the species (Fig. 8). In contrast, the type of S. butleri (Fig. 9) has variable, much darker longitudinal streaks on the underparts, sometimes with a slightly anchor-shaped terminus at the feather tips (Fig. 10), but lacking any evidence of obvious barring. Whereas the wing formulae of Israeli, Egyptian and Saudi Arabian birds are similar in that consistently p1 = p7/8, in the type of S. butleri p1 = p6/7. Furthermore, their tail patterns also differ; whereas the type of S. butleri has four visible dark bars on rectrices 2–4, of which the central two are c. 17 mm wide with 7 mm-wide pale bars, the outermost bar is 9–10 mm wide and the innermost c. 13 mm wide, but in most specimens of S. hadorami the five dark and pale tail bars (including the tail tip) are almost of equal width (albeit the former are subtly narrower), both measuring c. 6–9 mm. Other morphological differences are possible; for example, while the facial discs of birds in Israel and Egypt are uniformly pale (grey-white), that of the type of S. butleri appears to be paler mainly on the lower rim and darker grey above (Fig. 11). However, we should note that the disc around the left eye has been almost entirely destroyed and that this remark therefore pertains only to right half of the face. The issue of how hadorami differs morphologically from “Strix omanensis” is fully evaluated in the Discussion, but fundamentally the new species described here appears to differ from “omanensis” in many of the same characters that separate it from the butleri type. It merits emphasizing that Robb et al. (2013) comprehensively demonstrated

FIGURE 4. Median-joining haplotype network based on 218 bp of the mitochondrial gene cytochrome b (cyt b). Black dots represent mutation steps and circles different haplotypes with size of circles proportional to haplotype frequencies. Green circles signify median vectors.



37

TABLE 4. Corrected genetic distances (%) between the holotype of Strix butleri (BMNH 1886.2.1.994) and S. hadorami and additional taxa within this genus. Strix butleri holotype

Strix hadorami

Strix woodfordii

Strix aluco

Strix nebulosa

Strix uralensis


-

Strix hadorami

10.6

0

Strix woodfordii

10.0−11.7

7.0−8.7

0−1.9

Strix aluco

11.2

11.6

11.1−12.7

0

Strix nebulosa

13.7

14.1

13.5−15.2

12.9

0

Strix uralensis

10.3

10.7

10.2−11.9

10.6

13.0−13.1

0−0.9

Strix rufipes

12.9

13.4

12.8−14.5

13.2

15.7

10.5−10.6

Strix rufipes

-

FIGURE 5. Best-scoring maximum likelihood tree with bootstrap values > 50 indicated at each node.


KIRWAN ET AL.

that: (i) the northern Oman population discovered by them differs vocally from what all previous authors have considered to be butleri, to the extent that they cannot be considered the same species; (ii) the vocalizations of “omanensis” do not represent dialectal or other variation of another Strix species, previously unknown in Arabia; and (iii) “omanensis”, morphologically, most closely recalls what all previous authors have recognized as S. butleri. Given the vocal distinctiveness of the northern Omani population, it is clearly not the same species as we describe here. We can add nothing to the discussion of vocalizations in Robb et al. (2013) and therefore do not consider this issue further here.

FIGURE 6. Holotype of Desert Tawny Owl Strix hadorami (BMNH 1965.M.5235), female collected by B. Aharoni at Wadi Kelt, northeast of Jerusalem, Israel, 28 March 1938 (ex. Meinertzhagen collection) (Harry Taylor © Natural History Museum, Tring).

Description of holotype. Colours correspond to those in Smithe (1976). Upperparts mixture of drab (closest to Buff #24), cinnamon (closest to Cinnamon #39), light sandy-buff (closest to Buff #53) and tawny (closest to Raw Umber #23), forming intricately mottled pattern with very narrow (barely noticeable) dark mesial streaks (closest to Fuscous #21). Mantle and back slightly darker and more tinged cinnamon-grey (closest to Smoke Gray #45) or pale isabelline (closest to Pale Pinkish Buff #121D), nape and head lighter through dominance of sandy-buff and pale ochre colour, but still spotted or diffusely blotched darker brown (closest to Antique Brown #37). Underparts from throat to belly pale (perhaps closest to Pale Horn Color #92) with light rufous barring (closest to Antique Brown #37), scattered dark mesial streaks very few and inconspicuous (Antique Brown #53). Vent and undertailcoverts near-white and unpatterned. Facial discs mainly pale sandy-grey or off-white with brownish suffusion (between Smoke Gray #45 and Drab #27), being only very slightly barred or mottled brown over a crescent-shaped area basically restricted to above and behind the eyes. Discs bordered slightly warmer brown (between Drab #27 and Clay Color #26), short feathers barred darker brown, but barring or spotting of these feathers subdued and not very prominent. Tight “trousers” down to divergence of toes unmarked buff-white (closest to Cream #54 but even paler). All remiges sandy-ochre (closest to Buff #124 but paler) barred dark brown above and below (closest to Dark Brownish Olive #129), and ground colour overall slightly paler below. Primaries have five visible bars, secondaries four (or three, the innermost often just covered by the tips of the greater wing-coverts); pp2–5 (numbered from outermost) emarginated, pp2–4 obviously so. Central bars are broader (up to 15 mm), basal and distal bars narrower (c. 9 mm). Central bars on underside also darker than inner and outer ones. Barring of tail slightly denser than remiges, and central rectrices have six bars, although the basal one probably would be covered by tail-coverts and hence invisible in the field. As on the wings, the central bars are broadest, but unlike the wingAN UNNAMED SPECIES OF STRIX (AVES)


39

feathers, the barring of the tail is almost equally strong from the base to tip. Upper primary-coverts medium brown (close to Ground Cinnamon #239) with two or three darker brown cross-bars (Dark Brownish Olive #129). Alula similar but more diffusely patterned, with pale buff patch on the outer web near tip. Secondary coverts paler than primary coverts, with much paler sandy-buff cross-bars and edges. Under primary coverts buff-white (closest to Cream #54 but marginally paler) with sparse dark brown streaks (again close to Dark Brownish Olive #129) near wing bend. Rest of underwing coverts similarly pale buff with no or little obvious pattern. Bare parts (colours in life unknown): bill largely yellow-straw, whole culmen and lower mandible appearing yellowish, only sides of upper mandible shaded horn-coloured. Toes thinly feathered on upper ridges, elsewhere bare, mainly yellowishlooking on the dried skin. Claws similarly pale yellowish straw. Measurements of holotype. Wing 249 mm, tail 142 mm, tail/wing ratio (100) 57.0, tail graduation 9 mm, bill (to skull) 24.5 mm, bill depth (at feathers) 16.2 mm, and tarsus 45.3 mm. Total length 320 mm. Wing formula: p1 (outermost primary) < wt (wingtip) 53 mm, and =pp7/8; p2 < wt 15 mm, =pp5/6; p3 < wt 8 mm; p4 longest; p5 < wt 7 mm; p6 < wt 27 mm; p7 < wt 49 mm; p10 < wt 73 mm; s1 (outermost secondary) < wt 81 mm.

FIGURE 7A. Dorsal (A) and ventral (B) views of specimens of, left to right, Desert Tawny Owl Strix hadorami (BMNH 1965.M.5236), the holotype of S. hadorami (BMNH 1965.M.5235) and the type of Hume’s Owl S. butleri (BMNH 1886.2.1.994) (Guy M. Kirwan © Natural History Museum, Tring).


KIRWAN ET AL.

FIGURE 7B. Dorsal (A) and ventral (B) views of specimens of, left to right, Desert Tawny Owl Strix hadorami (BMNH 1965.M.5236), the holotype of S. hadorami (BMNH 1965.M.5235) and the type of Hume’s Owl S. butleri (BMNH 1886.2.1.994) (Guy M. Kirwan © Natural History Museum, Tring).

Geographic distribution. All populations previously ascribed to Strix butleri, other than the type specimen, can be confidently considered to represent the new taxon, based on specimens examined by us and L. Svensson, and photographs (see, for example, Robb et al. 2013). Thus S. hadorami occurs in the Eastern Desert of Egypt, from Wadi Rabdeit (22o10’N, 36o26’E), close to the border with Sudan, north to Mons Porphyrites (27o15’N, 33o18’E), northwest of the port of Hurghada (Baha el Din & Baha el Din 2001, Clouet 2011), as well as the Sinai Desert, e.g. around St Katherine’s Monastery, through southern and eastern Israel, in the Negev and Judean Deserts (where known from at least 76 localities: Shirihai 1996), to Jordan, where apparently confined to the eastern fringe of the Jordan (Rift) Valley and the Rum Desert in the southeast of the country (Andrews 1995), thence south across western and, more locally, northern, central and northeastern Saudi Arabia, apparently throughout Yemen, and east to southwest Omani Dhofar (Jennings 2010; Fig. 1). It is presumably absent from northern Oman, where it is replaced by “Strix omanensis” (Robb et al. 2013, van Eijk 2013). Specimens examined. In addition to the type specimen, the following material was examined (coordinates, where available, taken from Goodman & Sabry 1984). Those measured (all by G.M.K.) are as follows (all from Israel unless otherwise stated): BMNH 1965.M.5236, male (often suspected to be juvenile), Mahd Dhahab gold mine (23o29’N, 40o50’E), 250 miles north-northeast of Jeddah, Saudi Arabia, 16 August 1950; ZMB 31/2017, female, Wadi Fauwar, Jerusalem, 31 March 1922; ZM-HIJ 1468, female, En Gedi, Dead Sea, 7 January 1973; ZM-



41

HIJ 819, sex unknown, locality and date unknown; ZM-HIJ 541, female, near Jerusalem, date unknown; ZM-HIJ 817, female, St Katharine’s Monastery, Sinai Mts, Egypt (28o31’N, 33o57’E), 17 May 1970; TAUM 10373, male, Dead Sea, 21 July 1980; TAUM 10802, male (?), Exp. Zoo, died 21 January 1997; TAUM 12161, male, Exp. Zoo, died 5 March 1998; TAUM 15603, sex unknown, En Gedi (31o28’N, 35o22’E), Dead Sea, 29 September 1972; TAUM 16150, male, Dead Sea, 9 October 2005; TAUM 9428, juvenile (sex unknown), junction of Wadi Mashash and Wadi Halhul (31o35’N, 35o22’E), 31 May 1978. Two additional specimens (both mounted) were examined (e.g. for wing formula), but not measured by G.M.K.: TAUM 4928, female, En Mor (30o50’N, 34o47’E), 17 January 1963, and TAUM 16322, male, En Fara (31o50’N, 35o18’E), Dead Sea, 26 February 1910. In addition, L. Svensson measured and sent photographs of the following specimens: AMNH 631,939 (male) and AMNH 631,940 (female), both collected at Wadi Suweinat (31o51’N, 35o17’E), Jordan Valley, on 30 November 1918 (see Table 1). For mensural data see Table 1. Mass: males 140.0–183.5 g (n = 4, an additional captive bird, thought to be male, was labelled as being 284 g), females 214–220 g (n = 2), unsexed 220 g (n = 1); one juvenile 215 g.

FIGURE 8. Male and female specimens of Desert Tawny Owl Strix hadorami (AMNH 631,939 and AMNH 631,940), both collected in Wadi Suenit, Jordan Valley, Israel, November 1918 (Lars Svensson © American Museum of Natural History, New York).

Variation. As already noted by Goodman & Sabry (1984), there is no evidence of pronounced sexual dimorphism in S. hadorami, in size, wing formula, or plumage, although females are on average slightly larger. However, there is reasonably pronounced (apparently) individual variation in coloration, especially of the upperparts, from darker (e.g. BMNH 1965.M.5236) to much paler (see Figs. 7A, 12). The dark barring on the underside of the outermost primaries is c. 12–17 (exceptionally 20) mm wide proximally, 9–11 distally, whereas the intervening pale bars are narrower, c. 6–11 mm wide, while the dark and pale tail bars are almost equally wide (the dark bars subtly narrower), both measuring c. 6–9 mm, although in some specimens, e.g. BMNH 1965.M.5236, the dark bars are obviously broader than this. Some specimens show emarginations only on pp 3–5 (e.g. TAUM 10373).


KIRWAN ET AL.

FIGURE 9. Type of Hume’s Owl Strix butleri (BMNH 1886.2.1.994), from Ormara, southwest Pakistan (Harry Taylor © Natural History Museum, Tring).

Etymology. It is a special pleasure to name this bird for Hadoram Shirihai, a much-valued colleague and collaborator for c. 20 years. Although Hadoram’s ornithological interests are staggeringly wide-ranging, his name is arguably particularly synonymous with this wonderful owl of wild places in the Middle East. He discovered, when still a young boy, a live but poisoned specimen in En Gedi (Shirihai 1975), which became the first individual to be held in captivity and is now a skeleton in the Tel Aviv University Museum. During the two following decades he showed this species to literally thousands of birdwatchers in southern Israel, including the first two authors of this paper, and was unquestionably the first person to note the morphological discrepancies between the type specimen of butleri versus Israeli and other populations, during preparation of his monumental work The Birds of Israel (Shirihai 1996). The vernacular name reflects the fact that, among Strix owls, this species is most closely associated with treeless, arid habitats, and it also closely mirrors the Hebrew name for the species, ‫לילית מדבר‬. The name hadorami is a noun in its genitive form. Ecology. Strix hadorami occurs in rocky desert areas with ravines, cliffs and small caves, particularly in deep wadis with crevices for nesting and roosting and some vegetation (Shirihai 1996, Baha el Din & Baha el Din 2001). It is most frequent in limestone cliffs, but also occurs in granitic and basaltic regions of Saudi Arabia, and sandstone regions of Jordan (Andrews 1995). The species is present from about sea level in southern Oman and around the Dead Sea to an elevation of c. 2800 m in southwest Saudi Arabia. Its breeding range overlaps that of other owls, e.g., Little Owl Athene noctua, Spotted Eagle Owl Bubo africanus milesi and Desert Eagle Owl B. ascalaphus (Jennings 2010). Analysis of pellets in Israel (Wadi Nekarot) reveals that its diet mostly comprises rodents and small insectivores (gerbils, jirds, mice and shrews), less frequently geckos, some passerines (e.g. Desert Lark Ammomanes deserti and House Sparrow Passer domesticus) and arthropods such as scorpions, grasshoppers and beetles (Leshem 1981). Shirihai (1996) recorded the minimum distance between breeding pairs as 400 m in the Judean Desert but 2.5 km in the Eilat Mountains. The breeding season is March–August in Israel (Shirihai 1996), but earlier in Dhofar (southern Oman), where pairs generally cease vocalizing as early as March (Jennings 2010). Egg-laying in Arabia is estimated to occur from early February to late April. Nesting data are scant: just one pair has been studied in detail in Israel, a total of five eggs was laid between 6 and 14 May, with an incubation period of 34–39 days; eggs hatched on 10–21 June and the fledging period was 30–40 days (Shirihai



43

1996); another clutch, in eastern Saudi Arabia, numbered three fresh eggs on 25 February, and Aharoni (1931) considered three to be the maximum clutch size. A juvenile still with some down was seen on 17 July near Riyadh, Saudi Arabia (Jennings 2010). The Arabian population has been estimated at c. 3000 pairs (Jennings 2010) and in Israel there were c. 200 pairs in the late 1980s (Shirihai 1996). However, recent surveys (2006–09) in Israel suggest a decline (http://www.ynet.co.il/articles/0,7340,L-3803208,00.html; N. Weiss pers. comm.), with some former territories occupied now by Bubo ascalaphus. Another very recent survey (January–February 2014, by H. Shirihai, M. San Román and G.M.K.) found evidence of the species in just four of ten wadis where the species was known formerly in the Judean Desert.

FIGURE 10. Close-ups of the underparts patterns of the type of Hume’s Owl Strix butleri (BMNH 1886.2.1.994; A) and Desert Tawny Owl Strix hadorami (BMNH 1965.M.5236; B), demonstrating the barring typical of the latter taxon, which is absent in the type of butleri (Guy M. Kirwan © Natural History Museum, Tring).


KIRWAN ET AL.

FIGURE 11. Close-up of the right-hand facial disc of the type of Hume’s Owl Strix butleri (BMNH 1886.2.1.994), showing the bicoloured pattern, which appears almost identical to that shown by the type of “omanensis”, with a paler crescent of feathers surrounding the lower half of the eye closest to the bill, cf. images 375–376, for example, in Robb et al. (2013: 280) (Guy M. Kirwan © Natural History Museum, Tring).

Notes. B. [Bathsheba] Aharoni was the daughter of Israel Aharoni, who is widely considered the father of Levantine zoology. A Jew born in Vidzi, in what is now Belarus, and schooled in Prague, Israel Aharoni immigrated to the Near East sometime between 1901 and 1904 (Shirihai 1996), and the earliest bird specimen we have found attributable to him is dated January 1906. Although his first name is usually Anglicized as Israel, his contemporary papers are typically credited to ‘J. Aharoni’. Little seems to be known about his wife Yodit; although it has been suggested that she assisted his zoological research (Beolens & Watkins 2006), we believe that the latter authors confused the wife and daughter. Given that the oldest label is clearly inscribed B. Aharoni, we can only assume that his daughter should be credited with collecting the holotype, although the possibility exists that the father was in fact responsible. Although our holotype formed part of the Meinertzhagen collection (and therefore might be tainted by suggestions of fraud), its provenance can be deemed legitimate on the basis of an independent investigation by R. Prŷs-Jones (pers. comm.) and P. C. Rasmussen. This analysis showed that in the many cases in which original labels of other collectors are retained on Meinertzhagen specimens, there is no evidence of data tampering, and furthermore, in this case Meinertzhagen’s Aharoni specimen matches the two AMNH Aharoni specimens of this species in unusual details of specimen preparation (Rasmussen & Prŷs-Jones ms; further details are available from these authors upon request). Their examination also shows that BMNH 1965.M.5236 (see below), another Meinertzhagen specimen (which also was not collected by him), must also be presumed to bear correct data.

Discussion That there are two species-level taxa in what has been regarded as Strix butleri would be a straightforward matter, except for two issues: (1) the uniquely differentiated type is of unknown geographical provenance, and lacks vocal data; (2) the recent description of “S. omanensis” from the mountains of northern Oman, based solely on photographs and sound recordings, but without a specimen or genetic sample; this taxon appears to differ quite obviously in vocal characters from the taxon that we here describe as S. hadorami and, to a lesser extent, in



45

morphology. As a result, we now discuss three alternative hypotheses in an attempt to understand the relationship between S. butleri sensu stricto and “S. omanensis”, as follows: (i) that “omanensis” is a subspecies of butleri; (ii) “omanensis” and butleri both represent species taxa; and (iii) that “omanensis” is a synonym of butleri.

FIGURE 12. Five specimens of Desert Tawny Owl Strix hadorami collected in eastern Israel (Guy M. Kirwan © Tel Aviv University Museum).

Is “Strix omanensis” a subspecies of S. butleri?—If the type specimen of butleri truly is differentiated from “omanensis”, as Robb et al. (2013) claim it is (although see below), and the type of butleri really was collected on the opposite side of the Gulf of Oman, the possibility exists that “omanensis” is a subspecies of butleri. This scenario, wherein separate subspecific populations exist in northern Oman (“omanensis”) versus southern Pakistan, and perhaps southeastern Iran (butleri), would be supported, if a population that does not differ significantly in vocalizations compared to “omanensis” is proven to exist in the latter region, and/or molecular data become available for “omanensis” that demonstrate a close relationship to the butleri type. However, there is some evidence of a biogeographical pattern wherein avian taxa that occur on both sides of the Gulf of Oman are represented by the same subspecies on opposite sides of the Strait of Hormuz, yet replaced by a different taxon in southern Oman (Dhofar), e.g., Laughing Dove Streptopelia senegalensis cambayensis versus S. s. senegalensis (Gallagher & Woodcock 1980, Dickinson & Remsen 2013), Long-billed Pipit Anthus similis decaptus versus A. s. arabicus (Alström et al. 2003, Kirwan & Grieve 2010), and Purple Sunbird Cinnyris asiaticus brevirostris versus another member of what is perhaps part of the same superspecies, Palestine Sunbird C. osea (Cheke & Mann 2008). Are “Strix omanensis” and S. butleri separate species?—This hypothesis presupposes the same variables as the


KIRWAN ET AL.

previous one, but under this scenario, a population of butleri exists, again presumably in southern Pakistan, that differs markedly in vocalizations from “omanensis”, and/or once molecular data become available they demonstrate the lack of a close relationship between the two populations. Is “Strix omanensis” a synonym of S. butleri?—In describing the first-named taxon, Robb et al. (2013) examined the same three specimens at Tring that formed part of our research, as well as a sample of photographs of wild birds from across the range of butleri sensu lato. Robb et al. (2013: 293) considered “omanensis” to differ morphologically from butleri sensu lato as follows: (i) bicoloured facial disc (dark grey above and pale grey below the eye) versus plain whitish, buffish or pale grey; (ii) very cold dark grey-brown upperparts with only vaguely demarcated spotting, versus pale greyish, rufous or brown with well-demarcated spotting; (iii) buffish-white underparts, becoming white ventrally, marked only with long vertical dark lines, versus yellowish to buff underparts, principally barred and sometimes also streaked vertically; (iv) broad dark trailing edge to underside of wing and wingtip, as a result of dark bands on remiges becoming broader and pale bands duskier towards outer wing, versus dark and pale bands of virtually even width, producing an overall paler outer wing; and (v) undertail has outer two bands broad, inner two narrow, resulting in overall dark appearance, whereas butleri sensu lato has outer two bands broad, but inner two only marginally narrower, producing a more even-coloured and overall paler appearance to the undertail. However, in terms of relating these differences to the type specimen of butleri, only the third difference is clearly claimed to exist between the latter specimen and the northern Omani population photographed by Robb et al. (2013: 282). Our own extensive examination of the butleri type versus the specimens herein designated as belonging to S. hadorami, conducted independently by G.M.K. and J.L.C., with additional but again separate input from L. Svensson, reveals the following. The facial discs of the butleri type are damaged, especially on the left side of the face, but on the right side there is some evidence that the disc is clearly bicoloured (Fig. 11), being darker above and behind the eye, but with an obviously paler crescent-shaped patch of feathers below the eye and that side of the eye closest to the bill (compare images of “omanensis” in Robb et al. 2013: 280). The upperparts are clearly not “very dark ... greyish brown” in the butleri type but, as Robb et al. (2013) noted, variation is considerable in the coloration of the upperparts in birds from Israel, Saudi Arabia and elsewhere, and furthermore the only information we have concerning the coloration of these plumage tracts in “omanensis” is based on images taken using flash photography, of which those published were all assumed to relate to a single individual, thereby precluding any analysis of morphological variation within that population. The colour differences reported in Robb et al. (2013) revolved around relative saturation of warm or cold, buff and brown colours, which can easily be affected by nocturnal flash photography. Thirdly, and importantly, the claim that the type of butleri is barred on the underparts is refuted by our three independent examinations; the type clearly lacks barring of the same type as is evident in all other specimens (i.e. hadorami) we have examined and the many photographs published by Robb et al. (2013), although some of the longitudinal streaks do terminate in slight hooked tips, with lateral extensions, at the feather tips (Fig. 10). With respect to the differences in the remiges cited by Robb et al. (2013), which can easily be compared in their paper only via the underwing (e.g. pl. 380 versus pl. 389–390), see Figs. 7 and 9, wherein it seems that the type of butleri has a darker outer wing (like “omanensis”) than specimens of hadorami, in which the pale bands are more obvious. Finally, the undertail pattern of the butleri type is not identical to the pattern described by Robb et al. (2013) for “omanensis” (see Diagnosis), but as no specimens pertaining to the latter were measured by them, it is impossible to be certain that they differ in this respect, especially as it is clear that the undertail pattern of the type does differ from hadorami, and is basically darker than the latter. In addition to the points of congruence between the butleri type and what Robb et al. (2013) named “omanensis”, we must also reiterate the possibility that Hume’s type specimen came from Arabia, specifically Oman. Although this has been postulated several times, due principally to the lack of subsequent records from the northern side of the Gulf of Oman (Roselaar & Aliabadian 2009, Robb et al. 2013), the fact that Nash was obviously obtaining specimens from other people and had port contacts, together with the strong trade links that would have existed between the two coasts at that time, gives new and apparently previously unconsidered weight to the possibility of an Omani provenance for the specimen. However, it seems unlikely that any proof of this will ever now emerge. In the absence of additional data that might even partially validate either of the alternative hypotheses, we strongly believe that the most parsimonious and scientifically probable conclusion is that “omanensis” is synonymous with butleri. However, we admit the plausibility of either of the two other scenarios (albeit both are



47

evidentially weak at the present time) and that it will be defensible to consider “omanensis” as a taxon inquirenda until such time, at least, as genetic data become available for the northern Omani population. Finally, much of this confusion is unlikely to have resulted if a type specimen of “omanensis” was available and could be compared morphologically and genetically to that of butleri. Robb et al. (2013) cited three reasons why they elected not to collect a bird: the perceived rarity of their new taxon; collecting a bird would make it more difficult to study “omanensis” in the field; and the taking of a specimen would attract local disapproval; we evaluate each of these reasons in turn. Firstly, it seems unlikely that the “omanensis” population will prove to be especially rare, or even endangered. Indeed, at least two new localities for birds with the same vocalization were reported (one of them retrospectively) within just two months of the “omanensis” description (van Eijk 2013; http:/ /gryllosblog.wordpress.com/2014/02/), while Jennings (2014) pointed out the relative uniformity in habitat between northern Oman and highlands in the United Arab Emirates, where an unidentified Strix was heard in April 2006, in Wadi Wurrayah. What is here referred to as S. hadorami was, until a few decades ago, widely considered to be one of the most poorly known birds in the Middle East, yet has since proven to be locally common. Therefore, we see no reason to assume that “omanensis” must be rare; rather, it may be merely geographically restricted. Secondly, although it is possible that taking a specimen might have made it more difficult to study “omanensis” in the field, this would have been true only for a brief time. Thirdly, no evidence that legal collection of a specimen would have been “unacceptable” locally was proffered. Peterson (2014) recently emphasized the extreme importance of type specimens, even in modern-day ornithology, a sentiment that we broadly support. Given that non-destructive sampling can sometimes elucidate taxonomic problems concerning new species described without such material (e.g., Nguembock et al. 2008), we believe that, at the very least, Robb et al. (2013) should have refrained from describing their new taxon until they had a vouchered genetic sample that they had compared with the type of S. butleri. Nevertheless, it is also true that a complete analysis of relevant specimen material has enabled a more thorough understanding of the present case.

Acknowledgements The following staff at the Natural History Museum, Tring, greatly enabled this work, including facilitating molecular analysis of the butleri type and other specimens: Mark Adams, Robert Prŷs-Jones, Hein van Grouw and Alison Harding. Sylke Frahnert and Pascal Eckhart (ZMB, Berlin) kindly permitted access to specimen material there. In addition, G.M.K. also thanks the curators of the Steinhardt Museum of Natural History and National Research Center, at the Dept. of Zoology, Tel Aviv University (Daniel Berkowic and Dr Roi Dor) and the National Natural History Collections, Hebrew University of Jerusalem (Dr Nir Sapir, Ron Efrat and Prof. Ran Nathan) for access to specimens in their care, and Hadoram Shirihai for hosting his visit. MS thanks Stefan T. Hertwig for valuable discussions concerning the genetic results. Similarly, we thank Paul Sweet (AMNH, New York) for permitting L. Svensson to examine additional specimens of S. butleri on our behalf. Svensson also assisted with the diagnosis of the new species and provided many helpful ideas during the early stages of this manuscript’s preparation. Noam Weiss kindly shared some of his recent, unpublished data concerned the current status of these owls in Israel. We are grateful to Normand David for his advice on our nomenclature. Our referees, Daniel Lane, Robert Prŷs-Jones, James V. Remsen and one anonymous reviewer, as well as the editor, Pam Rasmussen, contributed to the final manuscript; we especially thank Prŷs-Jones for discussing the case with us and providing unpublished information pertaining to some of the BMNH specimens of this owl.

References Aharoni, J. (1931) Drei neue Vögel für Palastina und Syrien. Ornithologische Monatsberichte, 39 (6), 171–173. Alström, P., Mild, K. & Zetterström, D. (2003) Pipits and Wagtails of Europe, Asia and North America. Christopher Helm, London, 496 pp. Andrews, I.J. (1995) The Birds of the Hashemite Kingdom of Jordan. Privately published, Musselburgh, Scotland, 185 pp. Baha el Din, S. & Baha el Din, M. (2001) Status and distribution of Hume’s (Tawny) Owl Strix butleri in the Eastern Desert of Egypt. Bulletin of the African Bird Club, 8 (1), 18–20. Bandelt, H.J., Forster, P. & Rohl, A. (1999) Median-joining networks for inferring intraspecific phylogenies. Molecular


KIRWAN ET AL.

Biology and Evolution, 16 (1), 37–48. http://dx.doi.org/10.1093/oxfordjournals.molbev.a026036 Baur, H. & Leuenberger, C. (2011) Analysis of ratios in multivariate morphometry. Systematic Biology, 60 (6), 813–825. http://dx.doi.org/10.1093/sysbio/syr061 Bensasson, D., Zhang, D.X., Hartl, D.L. & Hewitt, G.M. (2001) Mitochondrial pseudogenes: evolution’s misplaced witnesses. Trends in Ecology & Evolution, 16 (6), 314–321. http://dx.doi.org/10.1016/s0169-5347(01)02151-6 Beolens, B. & Watkins, M. (2006) Whose Bird? Men and Women Commemorated in the Common Names of Birds. Christopher Helm, London, 400 pp. Butler, E.A. (1877) Astola, a summer cruise in the Gulf of Oman. Stray Feathers, 5, 283–304. Cheke, R.A. & Mann, C.F. (2008) Family Nectariniidae (sunbirds). In: del Hoyo, J., Elliott, A. & Christie, D.A. (Eds.), Handbook of the Birds of the World. Vol. 13. Penduline-tits to Shrikes. Lynx Edicions, Barcelona, pp. 196–320. Clouet, M. (2011) Sur l’avifaune des massifs du désert oriental Égyptien. Alauda, 79 (3), 237–240. Dickinson, E.C. & Remsen, J.V. (Eds.) (2013) The Howard and Moore Complete Checklist of the Birds of the World. Vol. 1. Fourth Edition. Aves Press, Eastbourne, 461 pp. Drummond, A.J., Ashton, B., Buxton, S., Cheung, M., Cooper, A., Duran, C., Field, M., Heled, J., Kearse, M., Markowitz, S., Moir, R., Stones-Havas, S., Sturrock, S., Thierer, T. & Wilson, A. (2013) Geneious v5.4. Available from: http:// www.geneious.com/ (accessed 19 November 2014) van Eijk, P. (2013) Presumed second locality for Omani Owl. Dutch Birding, 35 (6), 387–388. Eriksen, J. & Victor, R. (2013) Oman Bird List: The Official List of the Birds of the Sultanate of Oman. Seventh Edition. Sultan Qaboos University, Muscat, 288 pp. Fregin, S., Haase, M., Olsson, U. & Alstrom, P. (2012) Pitfalls in comparisons of genetic distances: a case study of the avian family Acrocephalidae. Molecular Phylogenetics and Evolution, 62, 319–328. http://dx.doi.org/10.1016/j.ympev.2011.10.003 Gallagher, M. & Woodcock, M.W. (1980) The Birds of Oman. Quartet Books, London, 310 pp. Goodman, S.M. & Sabry, H. (1984) A specimen record of Hume’s Tawny Owl Strix butleri from Egypt. Bulletin of the British Ornithologists’ Club, 104 (3), 79–84. Heidrich, P. & Wink, M. (1994) Tawny Owl (Strix aluco) and Hume’s Tawny Owl (Strix butleri) are distinct species: evidence from nucleotide sequences of the cytochrome b gene. Zeitschrift für Naturforschung, 49, 230–234. Hüe, F. & Etchécopar, R.-D. (1970) Les Oiseaux du Proche et du Moyen Orient. N. Boubée & Cie, Paris, 948 pp. Hume, A.O. (1878) Asio butleri, sp. nov.? Stray Feathers, 7, 316–318. Jennings, M.C. (Ed.) (2010) Atlas of the breeding birds of Arabia. Fauna of Arabia, vol. 25. King Abdulaziz City for Science & Technology & Fauna of Arabia, Riyadh & Frankfurt, 751 pp. Jennings, M. (2014) A new species of owl from Oman. The Phoenix, 30, 1–2. Katoh, K., Misawa, K., Kuma, K. & Miyata, T. ( 2002) MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Research, 30 (14), 3059–3066. http://dx.doi.org/10.1093/nar/gkf436 Khaleghizadeh, A., Scott, D.A., Tohidifar, M., Musavi, S.B., Ghasemi, M., Sehhatisabet, M.E., Ashoori, A., Khani, A., Bakhtiari, P., Amini, H., Roselaar, C., Ayé, R., Ullman, M., Nezami, B. & Eskandari, F. (2011) Rare birds in Iran in 1980–2010. Podoces, 6 (1), 1–48. Kirwan, G.M. & Grieve, A. (2010) Studies of Socotran birds V. On the validity of Anthus similis sokotrae and a few remarks on Arabian and northeast African populations of Long-billed Pipit A. similis. Sandgrouse, 32 (1), 43–49. König, C., Weick, F. & Becking, J.-H. (1999) Owls: A Guide to the Owls of the World. Pica Press, Robertsbridge, 462 pp. König, C., Weick, F. & Becking, J.-H. (2008) Owls: A Guide to the Owls of the World. Second Edition. Christopher Helm, London, 528 pp. Leshem, J. (1981) The occurrence of the Hume’s Tawny Owl in Israel and Sinai. Sandgrouse, 2, 100–102. Meinertzhagen, R. (1930) Nicoll’s Birds of Egypt. Vol. 1. Hugh Rees, London, 348 pp. Meinertzhagen, R. (1954) Birds of Arabia. Oliver & Boyd, Edinburgh & London, 624 pp. Mendelssohn, H., Yom Tov, Y. & Safriel, U. (1975) Hume’s Tawny Owl (Strix butleri) in Judean, Negev and Sinai Deserts. Ibis, 117 (1), 110–111. http://dx.doi.org/10.1111/j.1474-919x.1975.tb04193.x Mikkola, H. (2012) Owls of the World: A Photographic Guide. Christopher Helm, London, 512 pp. Nguembock, B., Fjeldså, J., Couloux, A. & Pasquet, E. (2008) Phylogeny of Laniarius: molecular data reveal L. liberatus synonymous with L. erlangeri and “plumage coloration” as unreliable morphological characters for defining species and species groups. Molecular Phylogenetics and Evolution, 48, 396–407. http://dx.doi.org/10.1016/j.ympev.2008.04.014 Nijman, V. & Aliabadian, M. (2013) DNA barcoding as a tool for elucidating species delineation in wide-ranging species as illustrated by owls (Tytonidae and Strigidae). Zoological Science, 30 (11), 1005–1009. http://dx.doi.org/10.2108/zsj.30.1005 Pacheco, M.A., Battistuzzi, F.U., Lentino, M., Aguilar, R., Kumar, S. & Escalante, A.A. (2011) Evolution of modern birds revealed by mitogenomics: timing the radiation and origin of major orders. Molecular Biology and Evolution, 28 (6),



49

1927–1942. http://dx.doi.org/10.1093/molbev/msr014 Peterson, A.T. (2014) Type specimens in modern ornithology are necessary and irreplaceable. Auk, 131 (3), 282–286. http://dx.doi.org/10.1642/auk-13-204.1 Price, T. (2008) Speciation in Birds. Roberts & Co., Greenwood Village, Colorado, 470 pp. Rasmussen, P.C. & Anderton, J.C. (2005) Birds of South Asia: The Ripley Guide. Vol. 2. Smithsonian Institution, Washington DC & Lynx Edicions, Barcelona, 683 pp. Robb, M., van den Berg, A.B. & Constantine, M. (2013) A new species of Strix owl from Oman. Dutch Birding, 35 (5), 275–310. Roberts, T.J. (1992) The Birds of Pakistan. Vol. 1. Oxford University Press, Karachi, 598 pp. Roselaar, C.S. & Aliabadian, M. (2009) Review of rare birds in Iran, 1860s–1960s. Podoces, 4 (1), 1–27. Rozen, S. & Skaletsky, H.J. (2000) Primer3 on the WWW for general users and for biologist programmers. In: Misener, S. & Krawetz, S.A. (Eds.), Bioinformatics Methods and Protocols: Methods in Molecular Biology. Vol. 132. Humana Press, Totowa, New Jersey, pp. 365–386. Scott, D.A. (2008) Rare birds in Iran in the late 1960s and 1970s. Podoces, 3 (1/2), 1–30. Scott, D.A. & Adhami, A. (2006) An updated checklist of the birds of Iran. Podoces, 1 (1/2), 1–16. Shirihai, H. (1975) “I found a Hume’s Owl”. Kenaff Renanim (Jerusalem Birdwatching Club), 1, 10. [in Hebrew] Shirihai, H. (1996) The Birds of Israel. Academic Press, London, 682 pp. Smithe, F.B. (1976) Naturalist’s Color Guide. American Museum of Natural History, New York, 21 pp. Stamatakis, A. (2006) RAxML-VI-HPC: Maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22, 2688–2690. http://dx.doi.org/10.1093/bioinformatics/btl446 Stamatakis, A., Hoover, P. & Rougemont, J. (2008) A rapid bootstrap algorithm for the RAxML web servers. Systematic Biology, 57 (5), 758–771. http://dx.doi.org/10.1080/10635150802429642 Svensson, L. (1999) Identification Guide to European Passerines. Fourth edn. Privately published, Stockholm, 368 pp. Ticehurst, C.B. (1927) The birds of British Baluchistan. Journal of the Bombay Natural History Society, 32, 64–97. Tristram, H.B. (1879) Letter to the Editor: Asio butleri and Caprimulgus tamaricis. Stray Feathers, 8, 416–417. Tristram, H.B. (1880) Letter to the Editor. Ibis, Series 4, 4, 245–246. Wink, M., El-Sayed, A.A., Sauer-Gürth, H. & Gonzalez, J. (2009) Molecular phylogeny of owls (Strigiformes) inferred from DNA sequences of the mitochondrial cytochrome b and the nuclear RAG-1 gene. In: Johnson, D.H., Van Nieuwenhuyse, D. & Duncan, J.R. (Eds.), Proceedings Fourth World Owl Conference. Oct–Nov 2007, Groningen, The Netherlands. Ardea, 97 (4), 581–591.


KIRWAN ET AL.