A review of the research on Canary Islands praying mantises (Mantodea) Author Wieland, Frank Author Schütte, Kai Author Goldberg, Julia text Zootaxa 2014 3797 1 78 102 journal article 45735 10.11646/zootaxa.3797.1.9 6166d638-5f68-4f0b-b0b0-7ba0909ba301 1175-5326 227227 D940560B-D602-44A3-A18F-9CC51ED46FC2 Canary Island Mantodea : Distribution of taxa and their taxonomy Distribution of Canary Island Mantodea . Figure 1 provides an overview of the seven Canary Islands including their respective geological ages, sizes and the current state of knowledge of their mantodean fauna (see also Fig. 3 and Table 4). Today, the most widespread species on the Canary Islands are M. religiosa and H. gracilis , each being found on five islands. While M. religiosa is missing only on Lanzarote and Fuerteventura, H. gracilis did not reach La Palma and El Hierro but is present on all other islands. Among Amelinae , A. gracilis has populations on Gran Canaria , Tenerife and La Palma. Ameles limbata is present on La Palma and Tenerife and has also been listed for Gran Canaria ( Chopard 1954 : 4; Ehrmann 2002 : 59). However, Kaltenbach (1979: 517–518) compared the Gran Canaria specimen with other so-assumed conspecific individuals and found that Chopard's specimen actually belonged to A. gracilis . Therefore, the presence of A. limbata on Gran Canaria is listed as doubtful in Figure 1 and Table 4. Nonetheless, Kaltenbach (1979: 523) still considered A. limbata a valid species (unlike stated by Otte & Spearman 2005 : 145). Distribution data of Pseudoyersinia are inconsistent. It has often been stated that four of its species are single island endemics: P. teydeana on Tenerife, P. canariensis on La Palma, P. pilipes on La Gomera and P. betancuriae on Fuerteventura (e.g., Bacallado Aránega 1984 : 123; García Becerra et al. 1992 : 94; Wiemers 1993 ; Francisco- Ortega et al. 2009: 129). Only P. subaptera was listed for two islands (Gran Canaria and Tenerife, e.g., García Becerra et al. 1992 : 141). However, a careful literature survey shows that P. canariensis has also been mentioned for Lanzarote and Tenerife ( Bland et al. 1996 : tab. 5) and P. subaptera has been listed for Lanzarote ( Gangwere et al. 1972 : 26; García-Becerra et al. 2001: tab. 1). Furthermore, Pseudoyersinia sp. was collected on Lanzarote by Heßler (2000), which was tentatively assumed to belong to P. canariensis (Heßler 2008: 139) . Additional sightings (ootheca; Ziegler, pers. comm. 2010) add to the evidence that a stable population of Pseudoyersinia is present on Lanzarote. Gangwere et al. (1972 : 26) and García-Becerra et al. (2001: tab. 1) assigned the Lanzarote species to P. subaptera , whereas Bland et al. (1996 : tab. 5) assigned it to P. canariensis . Pérez et al. (2003 : 244) identified several specimens collected on Montaña Clara, a very small islet off the coast of Lanzarote, as P. betancuriae . Again, this shows the uncertainty of the taxonomy and distribution of Pseudoyersinia species in the Canary archipelago. The Lanzarote population of Pseudoyersinia might belong to P. betancuriae , as the females are very similar (Stiewe, pers. comm. 2012; preliminary listed as " P. betancuriae ?" in Table 4; mentioned as " Pseudoyersinia sp." in Fig. 1). This is plausible because the islands were connected by land bridges in the past (e.g., Francisco-Ortega et al. 2009 : 127; Hochkirch & Görzig 2009 : 193). However, it is equally possible that Lanzarote was independently populated from mainland Africa by Pseudoyersinia , with the Lanzarote population now possibly forming a distinct species. Many factors are known to be important for the number of species on islands, among them island size, island topology, island age and the distance from the mainland (e.g., Gillespie & Roderick 2002 ; Steinbauer et al. 2012 ). As far as island size is considered ( Fig. 3 a), the distribution data for the Canary Island Mantodea is mostly consistent. Two of the largest islands (Tenerife and Gran Canaria ) hold the largest number of species, whereas the smaller islands with less than 1000 km 2 are home to roughly half the number of species or less. However, Fuerteventura is an exception. It is the second largest island in the archipelago but harbors only three mantodean species (Gran Canaria , which is of approximately the same size has 5 or 6 species). As mentioned before, distance from mainland Africa may also play a role in the mantodean colonization of the Canary Islands in that the closest island might harbor the highest number of species. Here, this is not the case ( Fig. 3 b). Again, Fuerteventura and also Lanzarote show only a medium number of species (three), comparable to two of the farthest islands (La Gomera and La Palma), whereas Gran Canaria and Tenerife have about twice as many species. Only El Hierro is consistent with this prediction, being the farthest from mainland Africa and harboring only one species of Mantodea . The last biogeographic assumption, that the oldest islands have the greatest species diversity, is proven erroneous, too ( Fig. 3 c). The two oldest islands, Fuerteventura (20.5 my old) and Lanzarote (15.5 my old) are populated by only three species each. Gran Canaria (14.5 my old) and Tenerife (11.5 my old) hold five (or six) and seven (or eight) ( Table 1 ) species, respectively. Although of a similar age as the former two islands, La Gomera (12 my old) has only three Mantodea species. Interestingly, the second youngest island of the archipelago, La Palma (1.7 my old), is home to four species, whereas El Hierro (1.1 my old) harbors only M. religiosa , as expected for a relatively young volcanic island. Except for island size none of the considered factors sufficiently explains the current number of species on the islands, excluding El Hierro. While the low number of species on La Gomera (despite its age) is most likely due to its small size, the low species number on Lanzarote and Fuerteventura is probably related to their arid climate, that is heavily influenced by the islands' proximity to the arid area of North-western mainland Africa (de Nascimento et al. 2009 ; Greve et al. 2012 : 1), making them the driest islands within the archipelago. Additionally, they do not comprise any high elevations, therefore causing a lesser impact of the trade winds with respect to cloud formation ( Juan et al. 2000 : Box 2). Also, some authors assume a complicated history of eradications of island faunae because of volcanism (e.g., Emerson 2003 : 11; see also Steinbauer et al. 2012 ), another factor that may have had a strong influence on the distribution of Mantodea in the archipelago. Additional to the geological and climatic factors, human impact must have considerably influenced species composition and distribution on all islands for the past 4000 years (de Nascimento et al. 2009 ) and has to be taken into account when considering the evolutionary history of today's species. Endemism. As could be expected, only the flightless Mantodea species show a high degree of endemism on single islands, whereas the fully winged species are widely distributed. If the current taxonomy is correct, four of the five Pseudoyersinia species [ P. teydeana , P. pilipes , P. betancuriae (only if presence on Lanzarote is incorrect) and P. canariensis (if not present on Lanzarote and Tenerife)] are each restricted to a single island, but P. subaptera can be found on two (or three) islands. The slightly larger Ameles gracilis and A. limbata , of which the females are flightless, are present on three islands [Gran Canaria (uncertain for A. limbata ), Tenerife and La Palma]. The remaining Mantodea species, well capable of flying in both sexes, have colonized four ( B. mendica ) or five islands of the archipelago ( H. gracilis and M. religiosa ). While all Amelinae found on the Canary Islands are probably endemic to the archipelago, the remaining Mantodea species are also widespread elsewhere. Mantis religiosa is almost cosmopolitan ( Ehrmann 2002 ; Berg et al. 2011 ) and apart from the Canary Islands, the distribution of H. gracilis spans northern Africa from Algeria to Saudi Arabia and Oman ( Ehrmann 2002 ; Roy 2004 ). Blepharopsis mendica is also widespread in northern Africa, reaching Turkey in the north and India in the East ( Ehrmann 2002 , 2011 ; Roy 2004 ). The Empusa pennata -problem. Brullé (1839: 76) mentioned the presence of " Mantis pauperata Thunberg " for the Canary Islands [later synonymized with Empusa pennata (Thunberg, 1815) ; see Ehrmann 2002 : 128; Otte & Spearman 2005 : 49]. Since then, several other authors mentioned the presence of E. pennata in the Canary Islands without adding further comments, among them Willemse (1936: 89–101) , Kaltenbach (1963 : 588, 1976: 166), Ehrmann (2002: 128) , Roy (2004: 8) , and Battiston et al. (2010 : 98). However, some authors, for example Bolívar (1893a: 52) , Chopard (1954: 4) and Roy (1987a: 118) , expressed their doubts regarding the presence of E. pennata on the Canary Islands ( Roy 1987b did not list the species at all in his Tab. 1 ), indicating that the existence of E. pennata in the archipelago is rather dubious. This controversy was further elucidated by Kaltenbach (1979: 530) . According to him, it is likely that Brullé (1839) confounded Empusa pennata ( Fig. 2 a) with Hypsicorypha gracilis ( Burmeister, 1838 ) , a very similar looking empusid ( Fig. 2 b) that Burmeister (1838) had described at the same time and that Brullé (1839) probably had not been aware. Furthermore, Kaltenbach (1979) argued, this would be even more likely given the fact that Brullé (1839: 76) determined the species from nymphal specimens (compare Fig. 2 ). Later on, the presence of E. pennata was again announced by van der Heyden (1991) who claimed to have collected a juvenile specimen on Gran Canaria , thus seemingly solving the long-lasting uncertainty. Van der Heyden (1991: fig. 1) provided a picture showing the specimen which, however, unambiguously showed a juvenile B. mendica . Van der Heyden (1991) was hence soon corrected in the same journal by Schroeder (1991) who showed that van der Heyden (1991) had simply misidentified the specimen. As already stated more than 30 years ago by Kaltenbach (1979: 531) , we do not have evidence for E. pennata being present on the Canary Islands. It is very unlikely that such an impressive species (ca. 6–7 cm body length; e.g. Giglio-Tos 1927 : 639), that is also a good flyer and is frequently attracted by light, could have been overlooked for more than 170 years, especially considering the high density of tourists and researchers on the Canary Islands. Therefore, the presence of the species should be considered a misidentification with H. gracilis . The year of publication of Ameles gracilis and A. limbata . Brullé (1839) described Ameles gracilis and Ameles limbata (both then in the genus Mantis ). There has been some confusion in the literature regarding the true publication dates of these two Ameles species. They have been listed for the years 1838 (e.g., Kaltenbach 1979 : 523, 524; García & Oromí 1999 : 103; García Becerra et al. 2001 : 124, 128), 1839 (e.g., Otte & Spearman 2005 : 144), and 1840 (e.g., Ehrmann 2002 : 59; Agabiti et al. 2010 : 4; Battiston et al. 2010 : 74). In Barker-Webb and Berthelot’s ( 1836–1844 ) second volume (part 2) of their Histoire naturelle des Iles Canaries , the publication date was given on the front page as 1836–1844 (in Roman numbers). Stearn (1937) then researched the true publication dates of the text and plates. He found that the text part of the "Animaux articulés"-chapter (of which Brullé was the single author of the insect section) was published in 1839 ( Stearn 1937: 55 ). Plate 5, however, where both mantid species were shown in combination with their names, was already published in 1838 ( Stearn 1937: 55 ). According to Article 12 in the International Code of Zoological Nomenclature (ICZN), names published before 1931 are available if an indication is given. A figure on a plate in combination with a name is a clear indication in the sense of Article 12. Therefore, the plate predates the written descriptions and the names Mantis gracilis and Mantis limbata (today assigned to Ameles ) were made available by Brullé in 1838. Homonymy of Mantis limbata Hahn, 1835 and Mantis limbata Brullé, 1838 . Gurney (1947) was the first author to recognize and discuss the homonymy of Mantis limbata Hahn, 1835 and Mantis limbata Brullé, 1838 . If Ameles limbata would actually prove to be a valid species, he argued, then the name of this junior primary homonym would have to be replaced ( Gurney 1947 ). Kaltenbach (1979: 523) discussed the problem and stated that in his view such a replacement would be unnecessary because limbata Brullé had been removed from Mantis in 1904 [ sic : Krauss 1892 ] and limbata Hahn had been transferred to Stagmomantis by Saussure (1872) [ sic : Saussure 1873 , see below]. As they are very distinct from each other and are even assigned to different subfamilies, Kaltenbach (1979) did not see any need to pursue a nomenclatural change. Koçak & Kemal (2008) decided to act on the limbata -problem. They argued that a) Mantis limbata Brullé is a junior primary homonym of Mantis limbata Hahn , b) Ameles limbata is taxonomically distinct from other Ameles species and c) Kaltenbach (1979) had not acted correctly with regard to the International Code of Zoological Nomenclature (ICZN). However, Koçak & Kemal (2008) did not provide any detail on their nomenclatural act with respect to the Code. Instead, they introduced the name Ameles canaria as a new replacement name for A. limbata Brullé. Article 23.9.5 of the Code clearly states that “when an author discovers that a species-group name in use is a junior primary homonym of another species-group name also in use, but the names apply to taxa not considered congeneric after 1899, the author must not [our emphasis] automatically replace the junior homonym; the case should be referred to the Commission for a ruling under the plenary power and meanwhile prevailing usage of both names is to be maintained.” A literature survey clearly shows that limbata Hahn was transferred to the genus Stagmomantis by Saussure (1873 ; cited as 1872 by Kaltenbach, but the true publication date of the Mantodea chapter was 1873, see Crosnier & Clark 1998 : Tab. 6) and limbata Brullé was transferred to the genus Ameles by Krauss (1892 ; Kaltenbach 1979 referred to Kirby 1904 who had transferred limbata Brullé to the genus Parameles ). As the two limbata species are very distinct from each other (morphologically, biogeographically and taxonomically) and have not been considered as congeneric after 1899, the requirements of Art. 23.9.5 of the Code are met. Therefore, the introduction of a replacement name by Koçak & Kemal (2008) is herein considered as incorrect. It merely adds to the confusion of the Amelinae taxonomy (see chapter "The Amelinae problem") and creates nomenclatural instability. Furthermore, confusion was enhanced by Koçak & Kemal (2008) because instead of correctly assigning the name Mantis limbata to ‘ Hahn, 1835 ’, the authors assigned it to ‘de Haan, 1835’, a publication that does not exist. A case regarding the preservation of Mantis limbata Brullé is in preparation for submission to the International Commission of Zoological Nomenclature.