Biodiversity Journal , 2012, 3 (4): 459-478

Annotated checklist of the (Insecta ) of the islands of the Sicilian Channel, including the first re - cords of Sympetrum sinaiticum Dumont, 1977 and Pantala flavescens (Fabricius, 1798) for Italy

Andrea Corso 1* , Ottavio Janni 2, Maurizio Pavesi 3, Michael Sammut 4, Arnold Sciberras 5 & Michele Viganò 6

1MISC - Via Camastra, 10 - 96100 Siracusa, Italy; e-mail: [email protected] 2Via G.G. D'Amore, n. 21 - 81016 Piedimonte Matese, Caserta, Italy 3Museo di Storia Naturale - Corso Venezia, 55 - 20121 Milano, Italy; e-mail: [email protected] 411, Sqaq Rigu, Birkirkara, BKR 2131, Malta; e-mail: [email protected] 524 'Camilleri Court' flat 5, il-Marlozz Str, Mellieħa (Ghadira), Malta; e-mail: [email protected] 6 MISC - Via Ongetta, 5 - 21010 Germignaga, Varese, Italia ; e-mail: [email protected] *Corresponding author

ABSTRACT In this paper we report data on the historical and recent status of all (In - secta, Odonata) recorded for the Sicilian Channel islands: the Pelagie islands and Pantel - leria, politically belonging to Italy, and Maltese Archipelago islands. The number of species known for the former group of islands raises from 7 to 20. Of these, 2 are new for the Ita - lian fauna, namely the Desert Darter Sympetrum sinaiticum , noticed through likely si - ghtings starting from 2010 on Lampedusa, and confirmed through voucher specimens collected in April 2012, and the Wandering Glider Pantala flavescens , first noticed in Oc - tober 2012 on Lampedusa and Linosa; while Calopteryx sp. cf. haemorrhoidalis , genei , Aeshna mixta , Orthetrum nitidinerve , Orthetrum coerulescens anceps , Crocothemis erythraea , Sympetrum striolatum , S. meridionale , impartita , Trithemis an - nulata and T. kirbyi , already known for Italy, are new for the Italian islands of the Sicilian channel. The Maltese fauna includes at present 18 recorded species; the previously reported Trithemis arteriosa is to be deleted from the list, since the concerned specimen upon re- examination proved to be T. annulata .

KEY WORDS Odonata; Sicilian Channel Islands; Sympetrum sinaiticum , Pantala flavescens ; new for Italy . Received 12.05.2012; accepted 22.12.2012; printed 30.12.2012 Proceedings of the 1 st International Congress “Insularity and Biodiversity”, 11 th -13 th May 2012, Palermo, Italy

INTRODUCTION leria (Consiglio, 1960; Lohmann, 1989; Pavesi & Utzeri, 1995), while material from these islands is Because of local scarcity of aquatic biotopes, re - also briefly discussed in other works (Carchini & sulting in low species richness, the dragonfly fauna Di Domenico, 1992; Carfì & Terzani, 1993). Only (Insecta, Odonata) of the Italian islands of the Sici - 7 species are reported in the literature for the Italian lian Channel has received little attention. Indeed, islands of the Sicilian Channel, including one only there are only three publications specifically dealing known from historical records and considered lo - with the Odonata of the Pelagie islands and Pantel - cally extinct (Pavesi & Utzeri, 1995); of recently 460 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

reported ones, only 2 were found to breed, and of other aspects of islands biodiversity, including Odo - these only 1 with a doubtless viable population. nata. This paper presents the results of our field This very low species richness no doubt largely re - work, together with an overview of existing publi - sults from near-absence of permanent or even tem - shed data. As a result of our field work, the number porary water bodies, suitable for Odonata breeding of species recorded for the Pelagie and Pantelleria and larval development. Yet a lot of dragonfly spe - has more than doubled, with 20 species, including cies are strong fliers, able to cover long distances 2 new for the Italian fauna. as occasional vagrants or regular true migrants, ABBREVIATIONS. AC = A. Corso; AS = A. more so when supported by favourable winds, and Sciberras; COM = Comino; GOZ = Gozo; LIN = Li - often also to breed, at least temporarily, even in ar - nosa; LMN = Lampione; LMP = Lampedusa; tificial, newly formed water bodies. Therefore the MAL = Malta; MISC (Malati di Isolitudine allo Sta - little number of recorded species is also due to li - dio Cronico) = a birding and nature group focused mited field work carried out over the years; on the on islands of Mediterranean Basin; MP = M. Pavesi; other hand, the number of actually occurring spe - MS = M. Sammut; MV = M. Viganò; OJ = O. Janni; cies may be subject to rapid changes, due to modi - PNT = Pantelleria. fications of local conditions, e.g. newly created suitable biotopes. Other groups attracted greater interest and have been the object of more MATERIALS AND METHODS detailed studies, in parallel with the situation on mainland Sicily (Corso, personal unpublished data), and nomenclature follow Dijkstra & which resulted in a clear, although limited, increase Lewington (2006) and Dijkstra & Kalkman (2012). in the number of recorded species. We consulted all the available literature on the The islands of the Maltese archipelago, com - Odonata of our study area, and analysed both re - pared to Pantelleria and the Pelagie, despite of in - cent and historical works in order to note any po - tensive habitat destruction still harbour many more pulation trends (increasing/decreasing/stable) and bodies of either fresh or brackish water, and local have a baseline checklist against which to compare Odonata received a greater deal of interest. Lite - our findings. rature about Malta and its satellite islands is varied Three of us (AC, OJ and MV) have visited Pan - and extensive; in addition to historical works (Val - telleria and the Pelagie in spring (February-May), letta, 1949, 1957), an increasing number of papers summer (June-August), autumn (September-No - in recent years focused on single species or on the vember), and a few winter visits in December and whole odonate fauna, including discussions of late January. More specifically, between April 2004 their biology and local status (Ebejer et al., 2008; and November 20 12 they have visited the Pelagie for Sciberras, 2008; 2011; Sciberras & Sammut, 2008; a total of almost 400 days, in the following periods: 2013; Sciberras et al., 2010; Gauci & Sciberras, April-May 2004, September-October 2005, March- 2010). This resulted in a more diverse known odo - May 2006, September-October 2006, March-May natological fauna, as far as certainly or supposedly 2007, September-October 2007, January 2008, breeding species are concerned: not less than 15 March-May 2008, June 2008, October-November of the 18 species so far recorded. Even so, further 2008, March-July 2009, September-November 2009, additions are to be expected, since climate changes January 2010, March-May 2010, October-November may result in formerly occasional or never found 2010, February-April 2011, July-August 2011, Oc - species becoming established, even with large po - tober-November 2011, October-November 2012. As pulations. For example, two of the present-day part of a LIPU-spo nsored study on raptor migration most abundant dragonflies, Orthetrum trinacria across the Mediterranean, they visited Pantelleria for and Trithemis annulata , were first recorded only 30-days periods in April-May 2005-2012 and Au - after 2000. gust-September 2008, during which they prospected From 2006 to 2012, we have visited the Pelagie all potentially suitable habitats, and as general rule islands and Pantelleria with the primary purpose of tried to cover as much ground as possible, in order gathering ornithological data, but we kept a high in - to find both odonates breeding in local water bodies terest in and have paid a great deal of attention to and migrating individuals. Occasional field trips Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 461

were also made by AS (April 2010 and September rain pool at Ponente, within the island’s nature re - 2010) and MP (November 2012). serve, a water-filled, shallow gravel pit at Albero As for Maltese islands, two of the authors (AS, Sole, a temporary pool at the mouth of a mostly MS) are resident on Malta, the largest one, and dry creek at Cala Pulcino, or again man-made were able to monitor the local odonate fauna all small water reservoirs (see below, under Sympe - over the last three decades. All species but very few trum fonscolombii ). Newly built-up water cat - have been photographed; voucher specimens for chments or reservoirs, to be kept filled with some many have also been collected and are now housed water throughout the year, where also aquatic ve - in the private collections of three of the authors getation may develop (e.g. within the Lampedusa (AC, MP and AS). Each specimen collected, whe - Island Natural Reserve), could indeed result in ther mounted or in an envelope, is accompanied by conspicuous increase of breeding species, with the following data: species (and subspecies if ap - some populations possibly becoming established. plicable), sex, locality (including GPS coordinates) Some species may also, at least occasionally, breed and date of collection, number of observed indivi - on Linosa, where no natural water bodies, yet here duals, collector (legit) and responsible for the iden - and there some small reservoirs exist. Lampione tification (det.). Determinations for each record, on the other hand is a small, not inhabited rock, either based on material evidence (voucher speci - where not even the least water catchment exists, men/photo) or only on sighting, rely on the record’s and therefore all Odonata there found are no doubt author himself, unless otherwise stated. migrants. Dragonflies occurrence, sometimes mas - sive, therefore results above all from migratory events, mainly linked to strong winds from sou - RESULTS AND DISCUSSION thern quadrants, namely “libeccio” (from south- west) and “scirocco” (from south-east). The Our study raises to 20 the number of odonate former above all seems to account for most of the species altogether recorded for the Pelagie (20, of influxes, because of much lesser distance, between which 2 tentatively) and Pantelleria (9, of which 1 the islands and the North-African coast, in south- with the only European large viable population), west direction rather than in south-east one. Winds vs. 7 previously reported in the literature (Table 1). from north-east (“grecale”) or north (“tramon - More exactly, as for the Pelagie, recorded species tana”), cooler than southern ones, are usually not richness is as follows: Lampedusa (17, of which 2 considered as source of influxes, yet they may ac - tentatively), Linosa (13, of which 2 tentatively), count for occasional arrivals from Sicily, including and Lampione (4). Two of the species recorded in weak fliers such as zygopterans. the past for Pantelleria, one of which at that time Differences in number of recorded species the - supposedly breeding, have never been found again refore result above all from different amount of ob - there during our survey (see below), while all the servations. Lampedusa, the only one of the Pelagie species previously recorded for the Pelagie were with an airport, is the easiest to reach, even with repeatedly confirmed. stormy sea conditions, while the other two can be It should be kept in mind that there is indeed reached only by ship. Exceedingly low number of no reason for which any species found on one of species recorded for Lampione partly relies on lack these islands would not to be expected, at least oc - of observations, only resulting from short visits, casionally, also in the others. All of them are rather because of non-existing local facilities; on the close to each other, therefore more or less equally other hand, on its most limited extension, compa - exposed to migratory influxes, accounting for most red to Lampedusa and Linosa, therefore on its har - odonate records, these islands being poor (Pantel - dly acting as rest site for migrant individuals. leria) to almost or completely devoid of water bo - Moreover, during strong winds periods, usually the dies (the Pelagie). Very few of these exist on best situation to observe dragonflies on the Pela - Lampedusa, where some species with evidence or gie, reaching the islet may prove impossible. presumption of breeding are recorded, although Pantelleria shows somewhat different condi - none is definitely known to have viable local po - tions. Besides being by far the largest and highest pulations; for instance a fairly large, temporary in altitude (836 m), it has a permanent saline lake, 462 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

SPECIES PNT LMP LIN LMN STATUS NOTES

ZYGOPTERA CALOPTERYGIDAE Calopteryx haemorrhoidalis ? accidental

C. virgo meridionalis no records accidental on Maltese islands

COENAGRIONIDAE

Ischnura fountaineae X ? X breeding Linosa records at reservoirs

Ischnura genei ? X breeding? ANISOPTERA AESHNIDAE occasional? large migrant swarms regularly Aeshna mixta X X X observed in Sicily migrant, irregular oviposition and exuviae recorded on Anax imperator X X X X breeding Pantelleria migrant, irregular Anax parthenope X X X X breeding? oviposition recorded on Pantelleria migrant, possibly potentially suitable biotopes on Anax ephippiger X X X X breeding Pantelleria

Orthetrum brunneum no records rare on Malta, breeding not confirmed single record; occasional migration Orthetrum nitidinerve X vagrant recorded on Malta Orthetrum coerul. anceps X vagrant? single record; common on Malta

Orthetrum chrysostigma no records records on Malta, status uncertain vagrant, formerly large population in the past on Orthetrum cancellatum X ? breeding? Pantelleria, extinct migratory species, recently established Orthetrum trinacria X X vagrant on Maltese islands occasional migrant, oviposition recorded on Lampedusa, Crocothemis erythraea X X irr. breeding? breeding not confirmed migrant, irregular Sympetrum fonscolombii X X X X breeding large numbers regularly observed migrant, irregular oviposition recorded on Lampedusa, Sympetrum striolatum X X X breeding? breeding not confirmed Sympetrum meridionale X vagrant single record regular in winter-early spring, no Sympetrum sinaiticum X ? migrant reproductive behaviour scattered records; established on Brachythemis impartita X occasional migrant Italian major islands Trithemis annulata X vagrant recently established on Maltese islands

Trithemis kirbyi X vagrant single record

Selysiothemis nigra no records breeding on Maltese islands single cluster of records; migratory Pantala flavescens X X vagrant species

Table 1. Odonata of the Pantelleria and Pelagie islands. Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 463

SPECIES MAL GOZ COM STATUS NOTES

ZYGOPTERA CALOPTERYGIDAE local populations never recorded, Calopteryx haemorrhoidalis X ? vagrant despite potentially suitable habitats C. virgo meridionalis ? X vagrant

COENAGRIONIDAE

Ischnura fountaineae no records possibly overlooked

Ischnura genei X X X breeding ANISOPTERA AESHNIDAE Aeshna mixta X X X occasional partly declining, displaced by Anax imperator X X breeding A. parthenope ? Anax parthenope X X X breeding exuviae recorded; no viable population Anax ephippiger X X X migrant, irregular breeding known LIBELLULIDAE

Orthetrum brunneum X breeding? rare, breeding not confirmed

Orthetrum nitidinerve X occasional single cluster of records

Orthetrum coerul. anceps X X breeding scattered records, single emergence Orthetrum chrysostigma X breeding? recorded; possibly underestimated declining because of O. trinacria Orthetrum cancellatum X X breeding settlement Orthetrum trinacria X X X breeding recently established, abundant

Crocothemis erythraea X X X breeding strongly declining, locally disappeared, Sympetrum fonscolombii X X X breeding overcome by O. trinacria no records in last years, overcome by Sympetrum striolatum X formerly breeding? O. trinacria Sympetrum meridionale no record

Sympetrum sinaiticum no records possibly overlooked scattered records on Linosa, Brachythemis impartita no records established on Italian major islands Trithemis annulata X X breeding recently established

Trithemis kirbyi no records

Selysiothemis nigra X X breeding

Pantala flavescens no records possibly overlooked

Table 2. Odonata of the Maltese islands. 464 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

with the only known viable population of Odonata ther records for Gozo exist, one of which was chec - throughout th ese islands, and a fair number of man- ked by AS and proved identical to Marsascala spe - made water reservoirs and catchments, where occa - cimen. Yet the latter, although badly damaged by sional breeding of some species was observed. The dermestid beetles (only three wings and small de - lower number of recorded species, compared with bris of thorax still remain), from the picture rather Lampedusa or even Linosa, can therefore only result resembles a C. haemorrhoidalis . from inadequate prospecting, particularly in some Wing length/width, considered by Sciberras & periods (see above); little doubt exists that absence Sammut (2008) as diagnostic, proved an unreliable of observations during the appropriate periods ac - feature, since considerable variations may be obser - tually accounts for non-existing local records of e.g. ved even within a single population throughout the Sympetrum sinaiticum or Pantala flavescens . season, depending on emergence period: the earlier Two species reported for Pantelleria in the lite - the emergence, the broader the wing, and also the rature (Pavesi & Utzeri, 1995) were not confirmed larger the individual size (Gallesi et al., in prepara - during our survey. Orthetrum cancellatum , recorded tion). On the other hand, wing venation, dark blac - in 1875 in large numbers and probably breeding at kish instead of bright metallic blue, remnants of that time, was never found again and is very likely thorax cuticle also dark black-bluish, basal clear to be locally extinct. O. trinacria , recorded in 1984 area of wings light tan instead of almost colourless upon a single individual, was possibly never establi - hyaline, all argue for C. haemorrhoidalis and shed. Breeding was recorded or supposed throu - against virgo s.l. However, having the specimen ghout different islands for some species, yet at been found dead in a rocky pool, these colour fea - present a definitely viable population is only known tures may also result from a post mortem alteration. on Pantelleria: Ischnura fountaineae . Sympetrum si - More recently, Sciberras & Sammut (2013) dealt naiticum and Pantala flavescens are new for Italy; with some additional collection materials; besides Calopteryx sp. cf. haemorrhoidalis , Ischnura genei , the above material, they also report and figure a Aeshna mixta , Orthetrum nitidinerve , O. coerule - fairly preserved specimen from Marsascala, clearly scens anceps , Crocothemis erythraea , Sympetrum recognizable as C. haemorrhoidalis (along with a striolatum, S. meridionale , Brachythemis impartita , true meridionalis , see below). It is somewhat sur - Trithemis annulata and T. kirbyi are first recorded prising that no local population was ever noticed, for the Italian islands of the Sicilian Channel. since in Maltese islands some flowing-water bodies The list of species recorded on Maltese islands exist, very likely unsuitab le for the highly rheophi - at present remains at 18, the most recent being Or - lous and microthermophilous C. v. meridionalis , thetrum chrysostigma ; at least 15 are certainly or but potentially matching C. haemorrhoidalis re - supposedly breeding (Table 2). quirements. Therefore the loca l occurrence of C. Here follows a systematic and commented list haemorrhoidalis is to be regarded as occasional, of the recorded species. consequent to either human accidental introduction (e.g. by ship) or to arrival of single straggler indi - viduals, possible supported by natural events, such Subordo ZYGOPTERA Selys, 1854 as strong winds (Sciberras & Sammut, 2008) or Family CALOPTERYGIDAE Selys, 1850 even floating debris drift. The likely origin areas are Sicily or North Africa; in both of them C. haemor - Calopteryx haemorrhoidalis (Vander Linden, rhoidalis is common and widespread, often with 1825) large populations. A single female on Linosa, 12.X.2009, not far Sciberras & Sammut (2008; reported by Boudot from Cala Mannarazza, near a small fountain in a et al., 2009) ascribed to C. virgo (Linnaeus, 1758), private garden (AC), was only briefly observed and after some doubts about its possible belonging to could not be caught; we cannot assign it to a definite haemorrhoidalis , one specimen, found dead in a species, yet what could be noticed argues for the rocky pool at Marsascala, Malta, considering it as most likely C. haemorrhoidalis . This is the only possibly intermediate between ssp. meridionalis known record of a Calopterygidae for the Italian is - Selys, 1873 and ssp. festiva Brullé, 1832; few fur - lands of the Sicilian Channel, where no potentially Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 465

suitable habitat for viable populations is known to viour was observed, possibly because of stormy exist; small man-made fountains or springs, such as wind throughout the day, and no females with co - the concerned one, may only prove temporary at - loration other than immature bright orange were no - tractive sites for straggler individuals. ticed. However, 1 female from Pantelleria with the following data: VII.1954, leg. E. Moltoni, with ma - Calopteryx virgo ssp. meridionalis Selys, 1873 ture olive-greenish coloration, was already housed, without determination, in the Museo di Storia Na - Formerly recorded for the Maltese islands (Sci - turale, Milano (Pavesi & Utzeri, 1995). Ragusa berras & Sammut, 2008; again dealt with in Sciber - (1875), about Odonata noticed on Pantelleria at the ras & Sammut, 2013) upon poorly preserved lake, made no mention of any Ischnura ; the species material, considered as possibly intermediate bet - may of course have been overlooked, yet it is well ween C. virgo meridionalis and C. virgo festiva possible that the species is a more recent coloniser, Brullé, 1825. This record is herewith tentatively not occurring there at that time. Besides supposed ascribed to C. haemorrhoidalis (see above). Howe - changes in physical conditions (namely salinity), ver, the same authors (Sciberras & Sammut, 2013) the reported huge population of Orthetrum cancel - also found in a collection another specimen from latum may have prevented a small zygopteran from Dwejra (Gozo), clearly to be assigned to C. virgo becoming established (see also below, under this meridionalis . No other record is known, nor any species). As for the Pelagie, a very teneral female evidence exists that the species ever occurred with (MP det.) was found on Linosa, Monte Rosso, breeding populations in the Maltese islands. Since 19.IX.2010 (AS), where some Roman age small re - virgo meridionalis is not known for Sicily (nor it servoirs (“gebbie”) exist. Further observations are does occur in southern Italian mainland, replaced needed to assess whether small breeding popula - by forms supposedly referable to festiva ), the likely tions of this species and the following one locally origin areas is only North Africa, where the species occur. Not reported for the Maltese islands, possibly only occurs in few, scattered places. because overlooked. From 2006 to 2012, we have monitored the Specchio di Venere population, the only one known Family COENAGRIONIDAE for Italy and Europe, in order to evaluate population trends, identify actual or potential threats and Ischnura fountaineae Morton, 1905 launch a genetic study (R. Ana Sanchéz, unpubli - shed). The population appears to be numerically The species in literature is often reported as I. stable, and although a slight decrease was noted in fountainei ; however, since it is dedicated to a 2012, it may simply reflect later emergence rather woman, Miss Margaret E. Fountaine, under provi - than an actual decline; indeed, all our observations sions of International Code of Zoological Nomen - were made in the months of April and May. Al - clature, 4th edition, art. 31.1.2, the correct spelling though Pavesi & Utzeri (1995) report 1 male and 1 is fountaineae , as correctly reported e.g. in Dijkstra female collected by Ratti on 2.V.1984, we never ob - & Lewington (2006) or in the website Fauna Euro - served emergence before 14.V. One of the said au - paea. thors (Utzeri) found no adults or larvae on Lohmann (1989) reported the first Italian (and 28.IV.1991, so that they speculated on possible European) record for Pantelleria, at the volcanic local repeated extinctions of the species because of lake named Bagno dell’Acqua (the locality name exceedingly arid conditions and therefore high sa - used by the author and by subsequent ones) or line concentration in some years, followed by sub - Specchio di Venere, on the basis of sightings made sequent recolonisation; which may have accounted on 14 and 15.VIII.1984; subsequently larvae, exu - for supposed absence of the species in the said pe - viae and adults were seen and collected by Pavesi, riod. The same authors, on the other hand, do not Ratti, Carchini, Di Domenico and others (Pavesi & exclude the cycle may be primarily univoltine, with Utzeri, 1995). More precisely, a very large number late mass emergences, and partly (occasionally) se - of adults, partly emerging, and exuviae was recor - mivoltine, with some larvae delaying emergence ded on 7.VIII.1985 (MP); no reproductive beha - until following year. 466 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

Despite intensive searches, we never could find demic of major (and some of the smaller) Italian is - any larvae except for the days immediately prece - lands and Maltese archipelago, does not occur in ding emergence. Considering the high numbers of North Africa. individuals occurring in the site, this very likely de - AS also observed two Ischnura individuals on pends on larvae behaviour, to remain hidden inside Lampedusa, Spiaggia dei Conigli, 16.IX.2010, not thick cane tufts until ready to emerge, and may have collected, therefore not to be identified. They may been the real reason for the reported lack of fin - belong to either previous species, as well as to I. sa - dings. Considering observed emergences starting harensis Aguesse, 1958. Considering the proximity from May, and definitive coloration of females with the Tunisian coast, and the zoogeographic com - found in July, yet totally absent in the huge popula - position of the Pelagie, local occurrence, at least oc - tion observed in August, there is no reason not to casional, of I. saharensis is not unlikely; the species suppose a bivoltine/multivoltine cycle, as usual in occurs with very large breeding populations e.g. on Ischnura . On the other hand, this elusive larval Djerba island (AC, unpubl.). The above spottings, habit makes even more difficult to monitor the sta - as for number of species known for each island (see tus of the Pantelleria population, under current si - above), are treated as belonging to a single species. tuation threatened with extinction. Although the Specchio di Venere is a Nature Reserve, SPA (Spe - cial Protection Area) and SCI (Site of Community Subordo ANISOPTERA Selys, 1854 Importance), it is in fact exposed to continuous and Family AESHNIDAE Rambur, 1842 very strong human pressure, from both local people and tourists swimming in the pond and walking Aeshna mixta Latreille, 1805 along the edges, so trampling the breeding area of the species. Moreover, the cultivated fields imme - It was not previously recorded for the Pelagie or diately adjacent to the lake are regularly and heavily Pantelleria (Pavesi & Utzeri, 1995). During our subject to pesticides and fertilizers application. Ur - study, we found it on Pantelleria, Punta Spadillo, gent actions are strongly required for an effective, 19.V.2011, 1 male (AC). This is the only record not only virtual protection of this unique biotope; known to us for the island. Single spotting were no - the best preserved part of the area, delimited follo - ticed in 2010 on Lampedusa, Cala Calandra, 15.IX wing researchers’ indications, should be closed to and on Linosa, Monte Rosso, 19.IX (AS). In au - people’s transit and access should be only allowed tumn 2012, a rather relevant influx has been noticed for scientific research, while the remaining could on the Pelagie, with notably up to 5 a day on Linosa remain open for recreational purposes. Such mea - on last week of October and first week of Novem - sures should be of course strictly enforced with ade - ber, of both sexes (but mostly females), and single quate patrol service by rangers (see also Pavesi & individuals on Lampedusa by the end of October Utzeri, 1995). (AC, OJ, MV, I. Maiorano & G. Soldato). No si - ghting was noticed on Lampedusa between 23.XI Ischnura genei (Rambur, 1842) and 29.XI (MP). In April and May, dozens and so - metimes hundreds of immature individuals of this Recorded for the Maltese islands already in past species and of Aeshna affinis Vander Linden, 1820 literature (Ebejer et al., 2008; Boudot et al. 2009), can be seen migrating along the eastern coast of Si - where it occurs with healthy populations on the is - cily (from Capo Peloro to Capo Passero), and many lands of Malta, Gozo and Comino. Never recorded are seen coming in off the sea in the Syracuse area, for Pantelleria. As for the Pelagie, two specimens, probably arriving from the Sicilian Channel. The male and female (MP det.), were collected on Li - paucity of records for Pantelleria and the Pelagie is nosa, Cala Mannarazza, 18.IX.2010 (AS). Although therefore rather puzzling and may only be due to the breeding is not confirmed (yet possible, see lack of observers during its peak migration period. above), the record is quite interesting, being an evi - The present records confirm the species as highly dence that migratory influxes from Sicily, mediated migratory and able to cover large distances (Boudot by northern (“tramontana”) or north-eastern (“gre - et al., 2009), therefore also to colonise new areas or cale”) winds, may occasionally occur. I. genei , en - to appear far out of range. Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 467

Figures 1-6. Ischnura fountaineae , Specchio di Venere, Pantelleria (AC). Figure 1. Mature male. Figure 2. Immature female, C-type. Figure 3. Mature male, display. Figure 4. Immature female, C-type, display. Figure 5. Mature female, C-type. Figure 6. Mating pair (female C-type). 468 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

Recorded for the Maltese islands by Boudot et cancellatum female perched on the ground, to block al. (2009) upon a single specimen (Ebejer et al., it with legs over its wings and to feed upon it (MP, 2008; Sciberras, 2008). During the spring of 2012 unpubl.). At the same time this species also forms a total of six specimens were recorded for Malta, an important food source for migrating birds, espe - Comino and Gozo (Sciberras & Sammut, 2013). cially Hobby Falco subbuteo Linnaeus, 1758 and European Bee-eater Merops apiaster Linnaeus, Anax imperator Leach, 1815 1758 which are regularly seen catching it (MS, per - sonal data). Pavesi & Utzeri (1995) report a single record of 1 female, collected at Mursia, Pantelleria, on Anax parthenope (Selys, 1839) 9.IX.1994. During our study, we have repeatedly ob - served it in May at various localities on Pantelleria: Recorded for Lampedusa (common in August, Specchio di Venere, Punta Spadillo, Rukia, Rekhale, mainly during southern winds periods) and for Pan - Grazia, Scauri, Ghirlanda, Pian di Monastero. Both telleria (a single record in 1994) by Pavesi & Utzeri sexes have been observed, with preponderance of (1995) which consider it as regular and potentially females. There are also records, documented with breeding, although with no evidence of the latter. photos, in VI-VII.2010 and 2011, and X.2011 (A. During our study it was repeatedly observed in Belvisi). At least some of these individuals may be numbers, besides on Lampedusa, also on Linosa stragglers from Africa, or even Sicily. However, and Lampione (AC, OJ, MV, AS & I. Maiorano), breeding was photographically documented in 2011, e.g. on VII.2009 and 2010. On some days, in spring when a few females were observed laying eggs and (March to May), it arrives erratically by hundreds exuviae were found in a water catchment pool in the from southern quadrants. This mass of individuals locality of Arenella (A. Belvisi). The species has forms an important food source for the breeding also been repeatedly observed, from April to Sep - Eleonora's Falcon Falco eleonorae Gené, 1839, and tember, on Lampedusa and Linosa (AC, OJ, MV, AS for many other migratory birds (Corso, 2011), in - & I. Maiorano), always in limited numbers compa - cluding passerines (MISC, unpubl). At least attempt red to the other two reported Anax species. Also of breeding was documented on Pantelleria at the spotted on Lampione, 5.IV.2010 (AS). Specchio di Venere, where in 2009 it was observed In the Maltese islands it was always reported in laying eggs. No reason however does appear, for old literature as common to very common, deemed which water catchments, reported as oviposition to be the most abundant Aeshnidae; this possibly site for A. imperator (see above), could not be also resulted either from being the species in the past suitable for A. parthenope . much more common, or from confusion with the In the Maltese islands it is reported as common sometimes similar A. parthenope . Although it is in - (Ebejer et al., 2008), found by Sciberras (2008) to deed common and widespread, with several bree - be the commonest local aeshnid, generally domi - ding sites, as for abundance it is at present exceeded nant in coastal areas, including a number of bioto - by the latter. In most coastal water bodies, the for - pes formerly dominated by A. imperator . Mass merly dominant A. imperator in recent years has migrations have been reported, as well as small been progressively overcome by A. parthenope numbers mixed with large migrating swarms of A. (Sciberras, 2008). It can be observed searching for ephippiger (Sciberras, l.c.). Definitely recorded for prey in a variety of habitats, including cultivated Malta and Gozo. It is to note that records only based fields or urban areas, even far away from water. It on collection materials may lead to largely undere - catches several , especially Diptera, but also stimate its frequency, individuals being often excee - larger ones, including various Lepidoptera, e.g. La - dingly difficult to catch. siocampa quercus (Linnaeus, 1758) and Vanessa cardui (Linnaeus, 1758) , and other Odonata. Some - Anax ephippiger (Burmeister, 1839) times it catches preys standing on vegetation or ground (Sciberras, 2008), either lifting them or de - Recorded for Lampedusa only by Pavesi & Ut - vouring them on the spot; in Italy a female was ob - zeri (1995), with a single record in IV.1987. The served to hawk upon a motionless Orthetrum same authors however suggested that it may have Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 469

Figure 7. Anax imperator , mature male, found on the small pool at Arenella, Pantelleria, 6.X.2011, A. Belvisi. Figure 8. Idem, egg-laying female, 1.VII.2011. Figure 9. A. imperator , mature male, Specchio di Venere, Pantelleria, 28.VI.2011, A. Belvisi. Figure 10. Exuvia of A. imperator , Arenella, Pantelleria, 4.X.2011, A. Belvisi. Figure 11. A. ephippiger , mature fe - male, caught at Specchio di Venere, Pantelleria, 20.IV.2011 during a passage of thousands through the Sicilian Channel (AC). Figure 12. Idem, mature male. 470 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

been largely overlooked, and that possible breeding Cala Croce, 14.IX.2010 (AS), obviously vagrant. As in small natural or artificial water catchments, as for brunneum or coerulescens anceps , no suitable well as in the Bagno dell’Acqua, was to be verified, habitat does exist anywhere on these islands. considering its ability to breed in temporary pools Boudot et al. (2009) do not report it for the Mal - in arid environments, because of very rapid larval tese islands. Only recorded by Sciberras et al. development. We have found it to be common du - (2010), who report males and females for different ring all our visits to Pantelleria and the Pelagie, in - localities on Malta, over a very short period only, cluding Lampione (AC, OJ, MV & I. Maiorano). 18-22.VII.2008, without any observed reproductive On some spring days, especially in March and behaviour; never noticed since. Therefore the same April, but occasionally from late February, thou - authors regard its occurrence as resulting from mi - sands of individuals can be seen arriving off the sea gration not followed by successful breeding, which from North Africa. Much like A. parthenope and is very likely correct. Sympetrum fonscolombii , they are an important food source for breeding Eleonora’s Falcon Falco Orthetrum coerulescens anceps (Schneider, 1845) eleonorae and for many other species of migratory birds (Corso, 2011), including passerines (MISC, On the Pelagie or Pantelleria only once obser - unpubl.). For example, in March-April 2011 an im - ved, on Lampedusa, Punta Sottile, 14.IX.2010 pressive influx of thousands of individuals was re - (AS), most likely a vagrant. O. coerulescens is re - corded in Sicily (Corso, personal data) and on corded for the Maltese islands by Boudot et al. 28.IV.2011 almost two thousands individuals were (2009), who refer to clinal populations, seemingly estimated at the Specchio di Venere, with many spe - intermediate between nominal subspecies and coe - cimens collected (AC). Breeding was not documen - rulescens anceps (cf. anceps /cline obsolescent ); yet ted; however, since the other two Anax were according to Sciberras (2008; 2011), Ebejer et al. definitely found ( imperator ) or supposed ( parthe - (2008) and Gauci & Sciberras (2010), Maltese po - nope ) to breed, the same habitats may prove suita - pulations should be ascribed to ssp. anceps . Their ble also for A. ephippiger . taxonomic status and variability indeed deserve an On Maltese islands the species is a regular mi - in-depth study. In fact, according to Boudot et al. grant from March to October, and breeding has re - (2009), mainland Sicily populations appear inter - cently been confirmed on Malta (Sciberras, 2011). mediate between coerulescens and anceps . Mauer - Much like in Sicily, this author reports that the lar - sberger (1994) and Dyatlova (2006) are of the same gest influx of migrants was noted in March 2011, opinion. On the other hand, studies carried out with about 4,000 individuals estimated on throughout Sicily from 2006 to 2012, as for acces - 18.III.2011 on the island of Gozo (Maltese archipe - sory genitalia have shown a number of intermediate lago). specimens, together with others seemingly referable to the nominal subspecies, but none really matching ssp. anceps as Sardinian or North African Family LIBELLULIDAE Rambur, 1842 do (Corso, personal data). Two Maltese specimens collected by AS, from the pictures were found (by Orthetrum brunneum (Fonscolombe, 1837) AC and MP) to be, one obviously a true anceps , the other a chrysostigma (see below). Unfortunately Not recorded for the Pelagie or Pantelleria, part of Maltese Orthetrum material, sent for study where on the other hand no suitable biotopes exist. by AS to AC, never reached the latter and is defini - In the Maltese islands it is rare and only occurs in tely lost. It is therefore at present rather difficult to some freshwater streams on Malta (Ebejer et al., know whether some other anceps records actually 2008). refer to chrysostigma .

Orthetrum nitidinerve (Selys, 1841) Orthetrum chrysostigma (Burmeister, 1839)

A stream-dweller, on the Pelagie or Pantelleria Not recorded for the Pelagie or Pantelleria, only recorded upon a single sighting, on Lampedusa, where nevertheless its occurrence may me expected Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 471

(Pavesi & Utzeri, 1995), because of its ability to amount. Increased salinity because of rainfall shor - breed also in temporary pools in desertic environ - tage may have led to disappearance of O. cancella - ments, so presumably also in man-made water cat - tum , providing in the same time a suitable, almost chments. predators-free habitat to the highly tolerant Boudot et al. (2009) do not report it for the Mal - Ischnura fountaineae (Pavesi & Utzeri, 1995). No tese islands too. First records for the Maltese archi - further record for the island exists, despite small pelago, all from Malta, are reported by Gauci & water catchments and reservoirs (see above, under Sciberras (2010), a total of 4 females, from 2008 to Anax imperator ) being a well-known habitat of this 2010. One of these, found by Gauci at a man-made species. Not positively recorded for the Pelagie. On small freshwater pond in the Ghadira Nature Re - 20.XI.2012, a supposed female of this species was serve, a saline marshland, on 12.VI.2010, from the observed, but not caught or photographed, on Lam - photo is clearly newly emerged, no doubt locally, pedusa, Albero Sole, at a water-filled gravel pit and most likely in the very same spot, although exu - (AC). Few days later, on 23.XI.2012, quite close to via was not found. This record was also discussed the previous spot, a large libellulid, tentatively re - on Forum Natura Mediterraneo on 19-20.XII.2012, ferred to this species (possibly the very same indi - between Charles Gauci (“Selys”) and MP (“gom - vidual), was briefly observed at some distance, phus”). No other records exist; this may argue for perched on the ground of a small dirty road (MP). only temporary breeding, resulting from occasional It disappeared before a clear sighting was possible. arrivals and followed by local extinction. However No other individual was seen. a fully mature male collected by AS, first believed In the Maltese islands it is recorded for Malta to be coerulescens anceps , from the picture was re - and Gozo (Ebejer et al., 2008) as common and oc - cognized by AC and MP separately as actually curring in any type of water bodies, including gar - being chrysostigma . It is to be stressed that light den ponds. However Sciberras (2008) reports it as stripes on the thorax, considered by the authors as declining, possibly also because of progressive in - diagnostic, are obvious only in females and not vasion of biotopes by O. trinacria . Some observa - fully mature males when thorax is not yet covered tions exist of O. cancellatum caught and devoured with pruinescence. Old males have dark, densely by O. trinacria . blue-pruinose thorax, with obsolete, unconspicuous stripes, therefore quite resembling anceps , and Orthetrum trinacria (Selys, 1841) when occurring in low numbers among large popu - lations of the latter may remain overlooked and un - Pavesi & Utzeri (1995) report a single record for noticed. As said above, because of partly lost Pantelleria, a sighting by Lohmann (1989) on 14- material, it is at present impossible to check some 15.V.1984, presumably a vagrant. The species was of previous Maltese anceps records, maybe partly collected on Lampedusa, Cala Croce and Spiaggia to actually refer to chrysostigma . dei Conigli, 14.X.2010 (AS), no doubt upon mi - grant individuals. O. trinacria is a strong flier, well Orthetrum cancellatum (Linnaeus, 1758) known as a migratory species (cfr. e.g. Fraser, 1936). No further records for any of the Italian is - Pavesi & Utzeri (1995) report a single historical lands of the Sicilian Channel exist. observation for Pantelleria by Ragusa (1875), refer - In Maltese archipelago large breeding popula - ring to “hundreds” of individuals seen at the Spec - tions occur on Malta and Gozo, first recorded in chio di Venere, as well as to large numbers of dead 2003 (Ebejer et al., 2008). Today the species is wi - dragonfly larvae, supposedly belonging to this spe - despread and common especially on Gozo, where cies and to equally abundant Sympetrum fonscolom - because of its highly aggressive territoriality and bii , on the bottom of the lake. The species was never predatory attitude, it has locally overcome most found again there, and the same authors presumed other Odonata (Balzan, 2008; Sciberras, 2008; it to be locally extinct, or at least to have become Sammut, personal data), systematically chasing extremely rare. Historical data about other groups away, when not hunting, any other dragonfly; cases of insects lead to suppose at that time a much lesser were reported of O. trinacria preying upon O. can - salinity of the lake, likely related to greater rainfall cellatum . Sympetrum fonscolombii , as well as the 472 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

more occasional S. striolatum , were reported to di - possibly only species able to breed in man-made re - sappear, following O. trinacria settlement (Sciber - servoirs on Lampedusa. Known for Pantelleria as ras, l.c.). The only smaller dragonflies able to early as the 19th century (Ragusa, 1875). It is in - coexist with large O. trinacria numbers seem to be deed quite common on all circum-Sicilian islands Crocothemis erythraea and Trithemis annulata , the (Terzani & Lo Cascio, 1997; Corso, personal data), latter another recent coloniser, whose populations including those of the Sicilian Channel. Because of had over last years a spectacular increase (Balzan, its attitude to regularly migrate, it can very often be l.c.; Sammut, personal data). seen in large numbers even on small islands com - pletely devoid of any water body. We have regularly Crocothemis erythraea (Brullé, 1832) observed egg-laying tandems on Pantelleria and Lampedusa. On the former island, exuviae and dead This is one of the commonest and most wide - adults are often found in the Roman cisterns at the spread odonates all over Italy, including Sicily San Marco acropolis (AC & OJ), and it can be seen (Corso, personal data), also known for its migratory at various localities along the coast, especially near attitude. It was not reported for Pantelleria or the the locality of Arenella, where a semi-permanent Pelagie by Pavesi & Utzeri (1995), who neverthe - rain-fed pool is present, but also quite far from less anticipated possible future records. On the for - water, e.g. at Scauri, Punta Spadillo, Kamma, Ghir - mer island, man-made water reservoirs, such as landa (AC & V. Penna). On Lampedusa we have those mentioned as breeding sites for Anax impera - observed territorial behaviour, mating pairs and tan - tor or Sympetrum fonscolombii , may well prove sui - dems from June to November in various localities table for this species. In September 2005, it was (e.g. Albero Sole, Ponente, Cala Pulcino, Cala Pi - observed and photographed on Lampedusa for the sana, Cala Madonna, Cala Francese, Valle Imbria - first time by M. Romano. In this island, we have cola). Egg-laying has regularly been observed at only seen it in July 2009, September 2010 and last Ponente, in a fairly large, temporary rain pool in the week of October 2012 (AC, AS). During the last si - western clearing of the pine grove within the is - ghting, 26-28.X.2012, not less than 4 pairs were ob - land’s nature reserve (the latest recorded on served mating and egg-laying in a small sewer 25.XI.2012, MP), and also in a water-filled, shallow outlet in the old harbour of Lampedusa town (AC gravel pit at Albero Sole (see above, under Orthe - & I. Maiorano). No occurrence was noticed there trum cancellatum ). As reported by Pavesi & Utzeri on 23-24.XI.2012 (MP). Its occurrence on Lampe - (1995), large numbers of this species arrive, toge - dusa is likely to be only occasional, due to the lack ther with Anax parthenope (Selys, 1839), during of large enough, permanent water bodies; the actual summer (especially August), on southern winds; we successful breeding in the above reported, heavily found this to happen not only in summer, but also polluted outlet remains in need of confirmation. in spring (April-May) and autumn (September-Oc - Also noticed on Linosa, Monte Nero and Cratere, tober). At these times, hundreds and sometimes 18.IX.2010 (AS). thousands of individuals can be seen arriving off the Recorded for the Maltese islands by Ebejer et sea, as in many other European coastal sites (cfr. al. (2008), Sciberras (2008), Boudot et al. (2009), Owen, 1958; Heyne, 1989). These influxes are as the most abundant and widespread dragonfly all often accompanied by numerous migratory birds, over the Maltese archipelago, and one of the very as noted for other species of migratory dragonflies few small species surviving habitat invasion by Or - (Anderson, 2009). Indeed, it seems that this species, thetrum trinacria . together with Anax parthenope and A. ephippiger (see above), is a key food source for many migra - Sympetrum fonscolombii (Selys, 1840) tory birds, especially raptors, as well as for breeding Eleonora’s Falcons Falco eleonorae (Corso, 2011). Reported for Pantelleria, Lampedusa and Linosa Numerous specimens, including late records (De - by Pavesi & Utzeri (1995), who mentioned nume - cember), have been c ollected, held in AC and MP rous records, including exuviae from Lampedusa, collections. Pavesi & Utzeri (1995) report the si - Cala Pisana, and deemed it to be one of the com - ghting of two unidentified Sympetrum on Pantelleria, monest odonates on these islands, as well as the 3.XII.1992, tentatively attributed to fonscolombii as Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 473

the most likely species. Although identification is North African coast, as well as the local existence not confirmed, the late date indicate a winter acti - of suitable biotopes with large breeding popula - vity for Sympetrum sp. already in the early 90’s, like tions, it can be supposed that migrant individuals regularly observed nowadays in both Sicily and Si - are a minority compared to whose locally emerged, cilian channel islands (Corso, personal data). and that most of individuals found on Maltese is - In the Maltese islands formerly considered very lands originate from locally breeding populations. common and widespread, it has drastically declined Influxes from Sicily, if any, are supposed to be nu - in recent years and locally disappeared, because of merically insignificant. predatory pressure from the increasingly abundant Evidences exist that exposure of larvae to dif - Orthetrum trinacria . It is still quite common in ferent average temperatures prior to emergence sites, mostly in coastal areas, not suitable to support may result both in reduced marking pattern and in breeding populations of O. trinacria . decreased size of adults. Waiting for a confirmation In nearly all specimens from the Pelagie exami - from rearing experience, we speculate that occa - ned by us, the yellow area at the base of hind wings sional individuals with more or less large yellow is highly reduced to hardly visible, on fore wings pattern found on Lampedusa may result from local almost always absent. Also body size is somewhat breeding, while the majority, with reduced pattern, diminutive. North African populations have an may be of North African origin. The Maltese is - identical pattern (Corso pers. obs.; C. Mancìn pers. lands do not have an obviously different climate, com). However a series of specimens collected on when compared with the Pelagie, yet unlike the lat - Ponente, 23.XI.2012 (MP) mostly have indeed on ter ones they harbour large viable populations (see hind wings the yellow area reduced, reaching up to above), which may account for having Maltese in - about mid-length between base of wing and base of dividuals usually larger yellow pattern than those triangle, not invading the median space; yet in two from the Pelagie. the yellow reaches up to the triangle base, with me - dian space lightly tinged, and in another the yellow Sympetrum striolatum (Charpentier, 1840) encloses the entire triangle and also fore wings are distinctly tinged at base. In Sicilian populations, the Not reported for the Pelagie or Pantelleria by Pa - extent of yellow area is generally larger, and the vesi & Utzeri (1995), nor by other authors, it ac - same proved for Maltese ones, the yellow on hind tually occurs on all these islands (not yet reported wings often reaching about the base of triangle, for Lampione), as well as on other circum-Sicilian upon an overview (MP) of photos, on the website islands, although erratically and in much smaller Forum Natura Mediterraneo, by Albert Floridia number than the previous species (Corso, personal (“xilpa”) and Charles Gauci (“Selys”). data). No spotting was noticed in autumn 2012 (AC, Since large numbers of S. fonscolombii regularly MP). It occurs in spring (March-May) and again in migrate across the Mediterranean, the existence of autumn (October-November). However, from seve - genetically separate and morphologically different ral specimens collected and held in AC collection, populations on their north and south sides is highly all spring individuals, as well as for late winter to unlikely. It is to be stressed that most of the indivi - early spring S. sinaiticum (see below) are quite aged duals occurring on the Pelagie or Pantelleria very according to general appearance and worn-out likely originate from Africa; although this species wings, which obviously indicates they have over - is definitely recorded to breed e.g. on Lampedusa, wintered, either on the Pelagie or in North Africa. there is little doubt that no viable resident popula - Absence of February records, contrary to S. sinai - tion actually does exist, the locally emerged stocks ticum , may be simply coincidental and only result likely being inadequate in numbers to support a from insufficient observations. S. striolatum may such population, also because of local aquatic bio - breed at least on Lampedusa, at Ponente, where we topes becoming completely dried-up all over sum - have occasionally seen pairs in tandem; yet no evi - mer; and on the other hand, locally emerged stocks dence from exuviae or newly emerged individuals being totally overcome by large, ceaseless influx does exist. from North Africa. The situation on Maltese islands In the Maltese Islands already in the past it was is quite different; given the greater distance from found to be scarce; only in a single instance, in 474 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

2007, a fairly good number of individuals was no - bruary 2009-2011 (AC, personal data). Emergences ticed (Sciberras, 2008). Strongly declined, if not to - take place in late spring, but the newly hatched in - tally disappeared in recent years (the last record on dividuals estivate, often travelling far from the 18.VI.2009), most likely because of its breeding emergence site, and do not breed until late autumn sites being progressively invaded by Orthetrum tri - or early winter. Therefore during years with high nacria (Sciberras, 2008). breeding success, some individuals disperse nor - thwards, and under favourable conditions may Sympetrum meridionale (Selys, 1841) reach the southernmost islands of the Sicilian Chan - nel, such as Lampedusa, more often in February- The hitherto only record for Sicilian Channel is - April. Although a single record on September exists lands is a male, caught and subsequently released (see above), it should be noted that throughout Oc - on Lampedusa, Cala Croce, 14.IX.2010 (AS), no tober/November 2012, when strong southern wind doubt a straggler. The species, widespread in North periods resulted in outstanding records of North Africa, is known to migrate over long distances. African dragonflies (see also under Trithemis kir - byi ), not a single sighting was noticed (AC, MP). It Sympetrum sinaiticum Dumont, 1977 should also be stressed that only more or less aged individuals were observed on Lampedusa; moreo - Not recorded previously for the islands of the ver no mating pairs or ovipositing tandems were Sicilian Channel, or anywhere else in Italy. On mid- ever noticed, despite observations period falls into March 2009, at various localities on Lampedusa, reproductive season of the species. Therefore at pre - we observed several unfamiliar Sympetrum indivi - sent no evidence exists even to suppose that the spe - duals, quite pale, the wings without any trace of cies may breed on Lampedusa, despite its basal yellow spot, obviously quite aged according seemingly quite regular occurrence from late winter to general appearance and worn-out wings; particu - to early spring. larly male features, such as pattern of thorax and On the other hand, S. sinaiticum breeds in de - abdominal segments, matched none of the locally sertic environments, in temporary, summer-dry likely species we are familiar with, such as S. strio - water bodies, and European populations, wide - latum , S. meridionale or S. fonscolombii , conver - spread in Mediterranean Spain (Dijkstra & Lewin - sely these argued for the Saharan and Near-Eastern gton, 2006; the same authors suggest it may be Sympetrum sinaiticum Dumont, 1977. The females, overlooked elsewhere), are reported to breed also without close-up view or in-hand examination, re - in concrete water reservoirs, where exuviae were mained somewhat puzzling. An individual of the found. Therefore further investigations are needed, same species was seen on 14.IX.2010 (AS). In Fe - in order to assess the actual status in the Pelagie, as bruary 2011, several individuals, of which 4 males, well as likely winter occurrence, or even possible were observed again in various localities on Lam - breeding, on Pantelleria. Never reported for Maltese pedusa (AC); also on Linosa individuals, possibly islands, possibly because overlooked. belonging to the same species, were observed later in March 2011, yet no close enough to enable defi - Brachythemis impartita (Karsch, 1890) nite identification. Finally, on 12-15.IV.2012, of 6 males seen on Lampedusa at Albero Sole, 4 mature Only recently (Dijkstra & Matushkina, 2009) re - to aged could be collected (K. Bacon leg.), enabling cognized as separate species, a long time confused definite identification by AC; now housed in AC (2) with the supposedly unmistakable B. leucosticta and MP (2) collections. All these likely reached (Burmeister, 1839). Therefore, all papers cited in Lampedusa from North Africa supported by sou - the following text refer to the species as to leuco - thern winds. sticta . The latter is not known to occur north of Sa - In North Africa S. sinaiticum breeds in winter hara (Dijkstra & Matushkina, 2009). (Jödicke, 2003; Boudot et al., 2009), after which Previously unrecorded for the Pelagie or Pantel - many individuals disperse, often reaching the coast leria; on Linosa 1 male found on 20.X.2009 (AC), of Tunisia. For instance, numerous individuals were and again spotted on Monte Vulcano, 7.IV.2010, 1 seen on the island of Djerba (Gulf of Gabés) in Fe - female according to entirely colourless wings (AS). Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 475

Figures 13-20. Sympetrum sinaiticum , male, Lampedusa, Albero Sole, 15.IV.2012, the first confirmed record for Italy of this species, widespread in North Africa; some distinctive features. Figures 13-14. Two specimens (MP coll.). Figure 15. Head, frontal view. Figure 16. Idem, ventral view. Figure 17. Thorax, lateral view. Figure 18. Accessory genitalia, lateral view. Figure 19. Terminal appendages, lateral view. Figure 20. Idem, ventral view. Note highly reduced dark markings, pale wing venation, hind wings totally devoid of basal amber spot, peculiar accessory genitalia. Photos 13, 14, 17, 18 by M. Zilioli.

Healthy po pulations of this species are found in cen - the flight season extends into October (Dijkstra & tral and Northern Tunisia and throughout southern Lewington, 2006). Indeed, much of the Pelagie fauna Sicily (Galletti et al., 1987; Boudot et al., 2009; originates from North Africa rather than from Sicily, Corso, personal data) and whole Sardinia. As such, both with regards to (cf. Massa, 1995) and considering the ability of the species to breed also in birds (Corso, 2005). On the other hand, B. impartita water reservoirs, colonisation of the islands of the Si - is a well-known migrant, since all Italian established cilian Channel would not be unexpected, as sugge - populations result from recent colonisation from sted already by Pavesi & Utzeri (1995). The Linosa North Africa (Galletti et al., 1987; Pavesi & Utzeri, records came, the former on a rather late date, the lat - 1995), as well as South European ones; previous re - ter on a quite early one, compared to the main flight cords for North Mediterranean countries were refer - period of this species in mainland Sicily (Corso, per - red to vagrant individuals not breeding in the area. sonal data), arguing for a North African origin, where Not recorded for the Maltese islands. 476 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

Trithemis annulata (Palisot de Beavois, 1807) where on the Pelagie or Pantelleria. The above re - cord shows that T. kirbyi , at least with favourable Only reported for Pantelleria and the Pelagie conditions of southern winds, is able to migrate through the sighting of 2 males and 1 female on over long distances. Indeed, the simultaneous oc - Lampedusa, Cala Calandra, 14.IX.2010 (AS); ad - currence on the Pelagie, during a period of sustai - ditional records are to be expected, given the recent ned “libeccio” (south-western wind), of two colonisation and very rapid population increase in African, never locally recorded species such as T. the nearby Maltese islands, where it was first recor - kirbyi and Pantala flavescens may hardly be regar - ded in 2005. From 2007 forward, regular records ded as coincidental. Such wind periods are also par - followed (Sciberras et al., 2007; Balzan, 2008; Ebe - ticularly fruitful for observation of several bird jer et al., 2008). In a very short time, the observa - species migrating across the Sicilian Channel; the tions became more frequent and the population above-said one also resulted in several North Afri - increased rapidly. The first exuviae were found on can bird species records in the Mediterranean basin Malta, at Chadwick Lakes and the Chinese Gardens (Corso, unpublished data). The minimum distance in Santa Lucia. In recent years, exuviae have also between southern Sardinia and Tunisia is about 180 been found at il–Qattara (Gozo). Today it is the se - km, no doubt within the reach of even a less strong cond most abundant anisopteran species in the Mal - flier, when supported by wind. Given the complete tese islands, superseded only by Crocothemis lack of further records, in subsequent years and de - erythraea . It seems to be among the few small dra - spite intensive researches, in the concerned locality gonflies to thrive in territories dominated by Orthe - or elsewhere in Sardinia, the actual occurrence of trum trinacria (MS, personal data). T. annulata is breeding Italian populations of T. kirbyi clearly indeed extremely aggressive, also towards other needs confirmation. species; in Greece it was seen attacking and chasing away considerably larger dragonflies such as O. Selysiothemis nigra (Vander Linden, 1825) cancellatum or even Lindenia tetraphylla (Vander Linden, 1825) (MP, unpubl.). Some cases are Not recorded for Pantelleria or the Pelagie. known in which habitat colonisation by T. annulata Given the status and positive trend in the Maltese resulted in local decline of the formerly abundant islands, possible future records are to be expected. C. erythraea (see Balzan, 2008; MP, unpubl.). The species is well known also for mass migrations (Fraser, 1936); on the other hand, Compte-Sart Trithemis kirbyi Selys, 1891 (1960) reports for Mallorca (Balearic islands) large populations breeding in concrete water reservoirs. Not previously reported for Pantelleria and the First recorded for the Maltese islands by Valletta Pelagie, nor for Maltese archipelago. On 28.X.2012 (1957), with 2 specimens collected in 1952; no fur - morning 1 male and 1 female were very well and ther records until 1996, when a specimen was col - long observed at Capo Grecale, Lampedusa (AC & lected (AS). In 2007, a single specimen was R. Finati), yet they proved impossible to catch; al - collected in July, at Ramla Bay, Gozo, and 5 fema - ready in the early afternoon of the same day they les were observed in August in a burnt field at Tas- were no longer to be seen, nor any further sighting Sellun, Xaghra, Gozo. During the same period, a occurred later. This constitutes also the first record permanent population was discovered in two artifi - for Sicily and the first Italian record outside Sardi - cial pools in a valley at Marfa, Malta. In 2008 two nia, where it was recorded on 23.VI.2003 at Oridda further, rather large populations were found at L- stream, Villacidro (VS) (Holusa, 2008) but never Ahrax and Ghadira, Malta, where from 19.VII to noticed since, despite active researches in recent 22.VIII the species was observed daily with a ma - years (B. Kunz, pers. comm.). ximum of 15 observed at the same time. Although The record for Lampedusa raises some doubts an increased abundance of S. nigra in recent years about the actual status of T. kirbyi in Italy. Although is likely, it is suspected that in the past this species the species may breed also in water reservoirs, there was simply overlooked and may have been more is little doubt that the above individuals did not re - regular than believed. This is primarily due to its sult from local breeding on the island, nor else - elusive behaviour; its tendency to fly very low, Annotated checklist of the dragonflies (Insecta Odonata) of the islands of the Sicilian Channel 477

along with small size and unconspicuous colouring, visits to Pantelleria and partially to Lampedusa pri - make it more difficult to detect than other Odonata, marily for ornithological studies. Also AS greatly even more when the species is unexpected and the - thanks his family for their support: Esther, Romario refore not expressly searched for. Since 2009, it and Jeffrey Sciberras. MS thanks his many field has been regularly observed in small numbers companions, but especially Joseph Grech, Christo - even in other areas, including Buskett and Dingli pher Cachia Zammit, and Dominic Frendo and the (Malta), especially on August-September (MS). late John Azzopardi and for various discussions and On 17.VII.2011 e.g., not less than 13 individuals communications with Charles Gauci who shares a were observed at Cirkewwa (AS). high interest in Odonata. MP also thanks Charles Gauci for discussions about Maltese Odonata on the Pantala flavescens (Fabricius, 1798) website Forum Natura Mediterraneo. Finally many thanks to Michele Zilioli (Museo di Storia Naturale, First records for Italy, and among the very few Milano) for the beautiful photos of Sympetrum si - for Europe, on Lampedusa and Linosa, X.2012 (OJ, naiticum material. MV, AC, I.Maiorano & G. Soldato), no doubt in the same days also to be found on Pantelleria. No lon - ger recorded on 23-25.XI.2012 (MP). Because of REFERENCES their outstanding interest, they will be thoroughly dealt with in a separate paper. Never recorded for Ander son C.R., 2009. Do dragonflies migrate across the Maltese islands. western Indian Ocean? Journal of Tropical Ecology, 25: 347-358. Balzan M.V., 2008. The distribution of Orthetrum trina - SPECIES ERRONEOUSLY RECORDED TO cria (Selys, 1841) and Trithemis annulata (Palisot de DELETE FROM LOCAL FAUNAL REPORTS Beauvois, 1807) in the Maltese Islands (Odonata: Li - bellulidae). Bulletin of the Entomological Society of Malta, 1: 91-96. Trithemis arteriosa (Burmeister, 1839) Boudot J.-P., Kalkman V.J., Azpilicueta Amorín M., Bogdanović T., Cordero Rivera A., Degabriele G., Reported for Malta upon a single specimen, an Dommanget J.-L., Ferreira S., Garrigós B., Jović M., aged female, collected in 2002 (Ebejer et al., 2008). Kotarac M., Lopau W., Marinov M., Mihoković N., However, after re-examination the specimen was Riservato E., Samraoui B. & Schneider W., 2009. found to be T. annulata (Sciberras, 2008), determi - Atlas of the Odonata of the Mediterranean and North nation now confirmed (AC, MP). Although its oc - Africa. Libellula, 9 (Supplement): 1-256. currence may be not unlikely, at present T. arteriosa Carfì S. & Terzani F. 1993. Attuali conoscenze del popo - is to be deleted from the list of Maltese Odonata. lamento odonatologico della Sicilia e delle isole di - pendenti (Odonata). Memorie della Società entomologica italiana, 71: 427. Carchini G. & Di Domenico M., 1992. The larval stages ACKNOWLEDGEMENTS of Ischnura fountainei Morton (Zygoptera: Coena - grionidae). Odonatologica, 21: 473-479. We would like to express our heartfelt thanks to Compte-Sart A., 1960. Biografía de la Selysiothemis our longstanding travel and research companions nigra V.d.L. (Odonatos). Graellsia, 18: 73-115. and island lovers from our MISC group: Hans Lar - Consiglio C., 1960. Odonata. In: Zavattari E. [ed.]. Bio - sson, Igor Maiorano, and Lucio Maniscalco. For geografia delle Isole Pelagie. Rendiconti dell’Acca - their help in the field and unpublished data, we demia Nazionale dei XL, 11: 330. would like to thank Giacomo Assandri, Giovanni Corso A., 2005. Avifauna di Sicilia. L’Epos, Palermo, 323 pp. Soldato, Cosmic Mancìn, Kevin Bacon, Andrea Corso A., 2011. Migrating dragonflies as a food source Belvisi, Piero Ferrandes, Rosa Ana Sanchéz, Sonia for breeding Eleonora’s Falcons and migrating rap - Ferreira, Jean-Pierre Boudot, Bernd Kunz, Sönke tors. British Birds, 104: 671-672. Hardersen, Elisa Riservato, Verena Penna and Rai - Dijkstra K.-D.B. & Lewington R., 2006. Field guide to mondo Finati. Finally, we thank Marco Gustin, the dragonflies of Britain and Europe. British Wil - Elena d’Andrea, and LIPU, who funded most of AC dlife Publishing, Gillingham, 320 pp. 478 A. C ORSO , O. J ANNI , M. P AVESI , M. S AMMUT , A. S CIBERRAS & M. V IGANÒ

Dijkstra K.-D.B. & Matushkina N., 2009. Kindred spirits: (Selys) = O. anceps (Schneider) in Europa und die “ ”, Africa’s most familiar dra - Konsequenzen für deren taxonomischen Rang (Odo - gonfly, consists of two species (Odonata: Libellulidae). nata, Libellulidae). Deutsche Entomologische Zeit - International Journal of Odonatology, 12: 237-252. schrift, 41: 235-256. Dijkstra K.-D.B. & Kalkman V.J., 2012. Phylogeny, clas - Owen D.F., 1958. Dragonfly migration in south-west sification and taxonomy of European dragonflies and Portugal, autumn 1957. The Entomologist, 91: 91- (Odonata): a review. Organisms Diversity 95. & Evolution, 12: 209-227. Pavesi M. & Utzeri C., 1995. Odonata. In: Massa B. Dyatlova E.S., 2006. Orthetrum coerulescens anceps [ed.]. Arthropoda di Lampedusa, Linosa e Pantelleria (Odonata, Libellulidae) in Odessa and its vicinities (Canale di Sicilia, Mar Mediterraneo). Il Naturalista (Ukraine). Vestnik zoologii, 40: 275-278. siciliano, 19 (suppl.): 151-162. Ebejer M.J., Degabriele G. & Sciberras A., 2008. An An - Ragusa E., 1875. Gita entomologica all'isola di Pantel - notated Checklist of Odonata of the Maltese Islands, leria. Bullettino della Società Entomologica italiana, with evidence for a recent influx of species. Libellula, 7: 238-256. 27: 133-145. Fraser F.C., 1936. The fauna of British India, including Sciberras A., 2008. A contribution to the knowledge of Ceylon and Burma. Odonata 3. Taylor & Francis, Odonata in the Maltese Islands. The Central Medi - London, 1-461. terranean Naturalist, 4: 275-288. Forum Natura Mediterraneo. http://www.naturamediter - Sciberras A., 2011. First record of a successful breeding raneo.com/forum/ of Anax ephippiger (Burmeister, 1839) in the Mal - Galletti P.A., Pavesi M. & Romano F.P., 1987. Brachy - tese islands (Insecta Odonata). Il Naturalista sici - themis leucosticta (Burm.) e considerazioni su altri liano, 35: 157-162. Odonati nuovi per la Sicilia (Insecta, Odonata). Il Na - Sciberras A. & Sammut M., 2008. On the occurrence of turalista siciliano, 11: 27-46. Calopteryx virgo meridionalis (Selys, 1873) (Odo - Gauci C. & Sciberras A., 2010. First Records of Orthe - nata: Calopterygidae) in the Maltese Islands. The trum chrysostigma (Odonata, Libellulidae) Burmei - Central Mediterranean Naturalist, 4: 339-342. ster, 1839, from the Maltese Islands. The Central Sciberras A. & Sammut M., 2013. The occurrence of the Mediterranean Naturalist, 5: 78-80. Copper Emerald Calopteryx haemorrhoidalis (Van - Heyne K.-H., 1989. Massenhafte Wanderung der Frühen der Linden, 1825), records of rare species, changing Heidelibelle ( Sympetrum fonscolombei , Selys 1840) population trends of some hitherto common species in Portugal. Dendrocopos, 16: 126-127. and recent colonisers in the Maltese Islands. Journal Holuša O., 2008. Trithemis kirbyi auf Sardinien: Erstna - of the British Dragonfly Society, 28: 1-9. chweis für Europa (Odonata: Libellulidae). Libellula, Sciberras, A., Sciberras, J. & Magro, D. 2007. A Cele - 27: 111-115. bration of Dragonflies. The Malta Independent. No - Jödicke R., 2003. Mid-winter occurrence of dragonflies vember 19: 8-9. in southern Tunisia (Insecta: Odonata). Kaupia, Dar - Sciberras A., Sciberras J. & Kunz B., 2010. Orthetrum mstädter Beiträge zur Naturgeschichten 12: 119-128. Lohmann H., 1989. Ischnura fountainei Morton auf der nitidinerve new for the fauna of the Maltese Islands. Insel Pantelleria, Italien: Erst Nachweis für Europa Libellula, 29: 55-60. (Zygoptera: Coenagrionidae). Notulae odonatologica, Terzani F. & Lo Cascio P. 1997. Odonati delle Isole 3: 61. Eolie (Insecta, Odonata). Il Naturalista siciliano, 21: Massa B. (ed.), 1995. Arthropoda di Lampedusa, Linosa 39-43. e Pantelleria (Canale di Sicilia, Mar Mediterraneo). Valletta A., 1949. A preliminary list of the Odonata of Il Naturalista siciliano, 19 (suppl.), 909 pp. the Maltese Islands. The Entomologist, 82: 85-87. Mauersberger R., 1994. Zur wirklichen Verbreitung von Valletta A., 1957. Second contribution to the Odonata of Orthetrum coerulescens (Fabricius) und O. ramburi the Maltese Islands. The Entomologist, 90: 306- 307.