Data on the occurrence of H. axyridis in Spain were collected by a thorough review of the literature and also by direct mentions from authors and colleagues.

In addition to the monitoring of reported aggregations during autumn and winter, from the end of the overwintering period until August, periodically (every 2-3 weeks) a sampling was done in several locations of Catalonia and the Basque Country on different types of vegetation. Sampled locations in Catalonia were La Seu d’Urgell [Lleida province; alfalfa and maize crop fields and plants in field margins; sweep net and Vortis® suction sampler (Burkard Manufacturing Co. Ltd., Ricksmanworth, UK) in alfalfa and visual sampling in maize and field margins], Alpens and Borredà (Barcelona province; margin vegetation in roads and gravel roads; visual sampling), Barcelona city (urban parks, visual sampling and yellow sticky traps), Lleida city (botanical garden and urban parks; visual sampling and yellow sticky traps), Lleida’s city surroundings and Gimenells (Lleida province; maize fields, visual sampling and yellow sticky traps), Cabrils (Barcelona province; margin vegetation in roads and in parks; visual sampling), and Torroella de Montgrí (Girona province; vegetation of the banks of river Ter; visual sampling). Scouting in the Basque Country was visually carried on in vegetable gardens and urban parks of Zarautz (Gipuzkoa province). Occasionally, some localities of northern Aragon (Jaca) and Navarra (Ochagavía, Elizondo, Pamplona) were also visually sampled by authors in forest and urban parks. The presence of the insect in Catalonia was also communicated by scientists and collaborators (colleagues from universities and research centres, urban landscape managers, farm advisers, etc.). Information to several colleagues and technical advisers in Central Spain, Valencia, Andalusia and Asturias was required for reporting current or existing observations of the recent presence of the ladybird in their geographical areas and for scouting.

Firstly, we characterized the sites where aggregations occurred: geographic coordinates, altitude, surrounding landscape, building characteristics, etc. Among the four aggregation sites reported, we selected the population of Les Llosses, since it was the largest. Characterization of this aggregation included natural mortality, parasitism rate, and phenology of the overwintering beetles, elytra colour pattern, sex ratio, weight, size, and female ovariole developmental stage before and after the overwintering period.

To determine the overwintering phenology, the aggregation was periodically monitored from November to May and the changes in colonies settle in visible places recorded. Yellow sticky traps and pollen bait traps were also placed inside and outside the Les Llosses overwintering site at the end of February and weekly monitored to detect individuals leaving the site. Along the overwintering period, individuals were collected and brought to the laboratory in order to determine sex, weight, size, ovariole development and parasitism. The adults were maintained in a refrigerator at 5 °C for 1-2 days before being individually photographed under stereoscopic microscope (Leica MZ8) on a millimetre paper. Then, specimens were individually preserved in ethanol 70°. The body length and width was measured from photos using the software programme ImageJ (Rasband W.S., US National Institutes of Health, Bethesda, MD, USA, http://imagej.nih.gov/ij/, 1997-2014). Alcohol preserved individuals were sexed under the microscope and then dissected in Beadle’s solution in order to establish ovariole developmental stage and potential parasitism. Weight and size were compared at the beginning and after overwintering by an ANOVA. Sex ratio was compared to 1:1 proportion by chi-square analyses.

By the end of the overwintering period, beetles started to be more active. Their distribution within the house (level and orientation) was periodically recorded (from 14 March to 2 May).

Table 1 and Fig. 1 show the historical records of H. axyridis in Spain, with the new mentions. In addition to the findings of the ladybird in Spain described in the Introduction section, and those reported from Carbonell & Sesma (2013) and Sesma (2015) through photographic records in the website http://www.biodiversidadvirtual.org, we have recorded overwintering aggregations in several new sites of central and north Catalonia and the presence in spring and summer of individuals (adult, larvae or eggs) in some others close or far away localities from the overwintering sites. Most of the records are concentrated in Catalonia, the only region where overwintering aggregations have been reported.

In spring and summer, we found adults and larvae of H. axyridis in some herbaceous plants (Rumex sp., Urtica sp.), on giant reed (Arundo donax L.) and on some trees (Liriodendron tulipifera L.) infested by aphids (Table 1). Other coccinellid species were also collected in these sites: Adalia decempunctata L., Coccinella septempunctata L., Hippodamia variegata Goeze, Propylea quatuordecimpunctata L., and Scymnus sp.

The aggregation from Les Llosses occurred in a isolated three-storey house of 135 m² floor and 7.90 m height, at 795 m altitude in the south-east slope of a 915 m altitude forest hill dominated by pines (Pinus sylvestris L.) and oaks (Quercus rubor L.). The house is surrounded by a meadow field and a tiny garden of vegetables and with some pear, plum and quince trees. Fifty meters slope down of the building there is a water stream with bank river vegetation.

The aggregation from Borredà was observed within a room of the east side of a several attached buildings of an isolated four-storey farm house. The house was located on the SE slope of a hill of 970 m, at 800 m altitude. The building was surrounded by garden vegetation, meadows and cereal fields and separated from a pine and oak forest (around 100 m) and from a water stream with bank river vegetation (150 m).

The aggregation of Alpens was observed inside a room of the east side of a three storey farm-house, 150 m to the East from the end of the village at 870 m altitude. The house was located in an area dominated by pine and oak forest and uncultivated lands. Near the house there is a road and beyond it a cultivated area with fields of winter cereals and grasses which ends 400 m slope down at 810 m altitude.

The aggregation of Capmany occurred inside a tool-hut and in a wood pile besides that building. The site was in a plane area at 100 m altitude. The hut was surrounded by orchards with vegetables and some scattered fruit trees, an almond tree field (0.5 ha) and a vineyard of nearly 1 ha.

Overwintering phenology. The house owner realized the presence of the coccinellids in middle October when she observed that in several days a large number of a new “type” of ladybird stayed on the outer East and South walls of her house. More than 500 adults stayed on that walls for some days and later insects were punctually observed within the house. On 4 November 2014, we recorded more than 200 individuals in several quiet groups, mainly on the East inner wall of the house. Some few adults were also observed on herbaceous plants and grasses around the building. During winter, overwintering individuals were mostly hidden, but several clusters remained quiets in visible places. An adult of Harmonia quadripunctata Pontoppidan was also identified within the aggregation. On 5 March, we detected movement of individuals and from then to the end of April active adults in the house were recorded. These individuals abandoned progressively the house and dispersed. On 2 May, no adults remained inside the house. No captures of adults were obtained from yellow sticky traps or pollen bait traps.

The end of the overwintering period started at the same time for the populations from Borredà and Alpens but occurred a little bit earlier in Capmany (end of February).

Elytra colour pattern. Forms succinea, conspicua and spectabilis (according Osawa & Nishida, 1992; Mitchie et al., 2010) were found over the sampling period. The form succinea was the prevalent over the whole period and represented between 73 to 81% of the individuals (Fig. 2). Among melanic forms, spectabilis was lightly more abundant than conspicua. There was no difference in the elytral colour pattern between sexes and between individuals collected before and after overwintering.

Sex. The sex ratio was not statistically different from 1:1 for any insect form and sampling date, although more females than males were recorded after overwintering (Table 2).

Weight and size. The weight of adult individuals was not affected by colour forms (F=1.93, p=0.1473; df=2, 271) but influenced by sex (F=4.18, p=0.0419, df=1, 271), females weighting more than males at each sampling date (Table 3). Also, individual weight changed according to the sampling date (F=4.52, p=0.0118, df=2, 271). Individuals collected before overwintering (4 November) being heavier (30.8±0.7 mg) than individuals collected after overwintering (14 and 28 March) (25.0±1.5 mg) (F=8.71; p=0.0034; df=1, 277). This difference was also significant according to beetle sexes (F=5.35; p=0.0214, df=1, 277) with a weight loss of 15% in females and of 28% in males.

The length and width did not vary with the elytra pattern form, but again changes with sex. Body size of females was significantly greater than males at the three sampling dates (Table 3).

Ovariole development. Dissection of females showed that before overwintering the ovarioles were empty of eggs. However, after overwintering there was a progressive increase of the proportion of ovarioles with eggs, reaching more than 60% by the end of March (Table 4). No copulating individuals were observed.

Mortality and parasitism. No old dead corpse from previous years was found. Three hundred and fifty-eight adults (males and females) were dissected and no parasitized individuals were found. Neither ectoparasitic fungi, nor ectoparasite mites were recorded. During the monitoring of the aggregation, no mortality was recorded.

Distribution of active ladybeetles in the overwintering house. A total of 761 observations of active individuals after overwintering were effectively done. Individuals were frequently found in groups, especially in the corners of the windows, or below window frames on the wall, or in the corners of ornamental devices on shelves on the wall. Most of the first active individuals were observed at the first floor, probably because it was the warmer level of the house. Later on the season, although active ladybeetles were observed in all house storeys, most of them were at the ground floor (Fig. 3A). In relation to the orientation, the insects preferred the East side of the house (Fig. 3B).

Two hypotheses can be formulated about the potential origin of the population: first, the Spanish origin (colonization derived from previously released ladybeetle in Almeria, south of Spain) and second, the European origin. Logically, the Almeria hypothesis should be discarded, since the release was considered unestablished (Jacas et al., 2006; Brown et al., 2011a), and no individual has been reported from the south of Spain since this time. Considering the European hypothesis, Lombaert et al. (2010, 2014) reported three routes of invasion of the European populations of the beetle, one coming from its native area released for biocontrol 1982 in France (named, European Biocontrol Strain), and another two, each one coming from eastern or western USA. In absence of genetic analyses, it is impossible to determine without any doubt the origin of the Spanish populations. However, the relative proportion of elytral color forms can be quite different among populations from different countries. For example, in USA and Eastern Canada, melanic forms are virtually absent (Koch, 2003; Lucas, pers. obs.). In the Spanish aggregation about 20% of the individuals have melanic forms and the relative proportion of elytral color forms (Fig. 2) is similar to other European populations (Adriaens et al., 2008; Burgio et al., 2008; Brown et al., 2008; Stankovic et al., 2011). Brown et al. (2008) suggest that the elytral color similarity in European populations is linked to limited points of origin in Europe, as Lombaert et al. (2014) has recently stated. Finally, all the reported mentions of H. axyridis in the present paper come from sites at less than 150 km from the French border. The entry may then have occurred from France trough the Pyrenees either by natural dispersion of the insect, or via human transport. The finding of the insect in Andorra (Sesma, 2015) supports this hypothesis.

Considering the timing of the invasion, the Spanish overwintering aggregation of Les Llosses is probably a new invasive population from 2013, since no cadavers were found from previous years and since the owner of the house who is living there had never seen the invader before. This aggregation successfully overwinters in 2013-14 and the phenology of the overwintering coincides with the described general pattern in the invaded countries of Europe and the Northern Hemisphere (Iperti & Bertrand, 2001; Koch, 2003; Raak-van den Berg et al., 2012).

Overwintering started with the gathering of individuals in mid October, the hibernation by the end of October or the beginning of November, and the end of the overwintering occurred at the beginning of March. Adults preferred mainly the East side of the house for overwintering as is suggested by the records during the aggregation period and by the distribution of active insects within the house in March and April.

The presence of undeveloped ovaries in females and the weight decrease in male and females by loss of fat body content have been reported as signals of hibernating individuals (Sakurai et al., 1992; Iperti & Bertand, 2001; Raak-van der Berg et al., 2012). Females of the Les Llosses’ aggregation had their ovaries empty of eggs until middle of March, when they started developing ovarioles and, by the end of March, nearly two of each three females had eggs in their ovarioles. Furthermore, males and females lose weight, although loss was higher in males than in females. The results on female ovariole development and adult weight suggest that the aggregation of Les Llosses has suffered a real hibernation. Ceryngier et al. (2004) showed that females of some coccinellid species may have inactive ovaries in early dormancy but sperm in their spermatheca, facilitating an early fertilization process in overwintering sites or nearby them. The lower weight loss in females than in males in the Les Llosses’s aggregation may be an indicator that females started their reproductive processes. Several studies reports a short diapause followed by a quiescence period allowing insects to be active if weather conditions become favourable (Iperti & Bertrand, 2001; Raak-van den Berg et al., 2012; Awad et al., 2013). The population from Campmany, a northern locality close to France border, finished the overwintering period before that of Les Llosses. This may be due to a milder weather in Capmany (due to lower altitude) according to a closer proximity to the sea: 14.4 to 8.3 °C from October to March in Les Llosses and 18.3 to 11.9 °C in Capmany, with minimum 3.7 and 8.2 °C in Les Llosses and Capmany, respectively (Servei Meteorològic de Catalunya, http://www.meteocat.cat). Further studies should be conducted in order to precise overwintering features of the Spanish populations of H. axyridis according to temperature, altitude, latitude and distance from the sea.

Some Asiatic populations are female biased due to the presence of a male-killing bacterium from genus Spiroplasma sp. This male killer has not been reported from H. axyridis in its introduced range (Kenis et al., 2008). In spite of that, Awad et al. (2013) found in Czech populations a lower survival rate during winter of males than females resulting in a low proportion of males in spring. We did not record significant differences to 1:1 sex ratio between sampling dates, although a lower proportion of males was recorded after overwintering.

After overwintering, insects from the aggregation of Les Llosses abandoned the house refugia and dispersed without evidence of their destination, although a few individuals were found in June relatively close to the different overwintering sites. The mentions of individuals in spring and summer far of the overwintering sites may be an evidence of the movement of the populations or could be due to the presence of other existing overwintering aggregations. H. axyridis is often considered as a good flyer (Hodek et al., 1993) and a rate of spread up to 200 km per year in Europe has been evaluated (Brown et al., 2011a), but spreading could also be favored by involuntary human transport (one mention reports an adult ladybeetle in Lleida within a car coming from northern Catalonia). It would be very important to follow the colonization of the Iberian Peninsula by systematic samplings of individuals during spring-summer and of overwintering populations during winter.

Beside the spreading capacity, individuals collected in the present study show that H. axyridis is able to occupy very different habitats (urban landscapes, reed beds, riparian or ruderal zones, etc.) on trees, shrubs and herbaceous plants and that it seems that there is an expansion from northern to central areas of Catalonia.

The impacts of the invasion by the Asiatic ladybeetle have been thoroughly reported in the invaded regions, with positive but mainly negative effects related to its establishment [see Majerus et al. (2006) as a revision]. Since the establishment of the beetle in Spain is now demonstrated, impacts of H. axyridis should be expected, and negative ones should be similar to those described elsewhere. Firstly, domestic problems would occur with high densities of the insect invading human houses for overwintering, as it was the case in the present study.

Secondly, impact on native or already established ladybeetle species may also occur. The arrival of H. axyridis is associated to drastic changes in coccinellid assemblages all around the word (Colunga-Garcia & Gage, 1998; Michaud, 2002; Lucas et al., 2007a,b; Pell et al., 2008; Bélanger & Lucas, 2011; Roy et al., 2012). Catalonia has a diverse agriculture dominated by winter cereals, maize, alfalfa, vegetables in open air and greenhouse, fruit orchards and vineyards. H. axyridis has been reported in all of these crops over Europe and America, as well as in urban landscapes. For example, in Belgium, Vandereycken et al. (2013) reported that maize is one of the crops most affected. Some of the most abundant coccinellids in Spanish arable crops (that is, C. septempunctata, H. variegata and P. quatuordecimpunctata) move between alfalfa, winter cereals and maize to follow aphid populations and demonstrate clear patterns of relative abundance (Pons & Eizaguirre, 2009; di Lascio et al., unpublished data). In the present study, no H. axyridis has been recorded in maize and alfalfa (two crops where coccinellids are abundant) and this suggests that the invasion of the Asian ladybeetle is recent. However, the invasion of crop fields should be promptly observed, as has happened in other countries, and all these three native coccinellids species are showed to be in disadvantage intraguild predation experiments with H. axyridis (Ware & Majerus, 2008; Lucas, 2012; Katsanis et al., 2013).

Spring and summer records from the present study show that H. axyridis has already colonized urban landscapes in Spain. In these environments, Oenopia conglobata L., Adalia bipunctata L., A. decempunctata, H. variegata and Scymnus sp. are the predominant ladybeetle species (Lumbierres et al., 2005; Pons & Lumbierres, 2013). Since asymmetric intraguild predation in favor of the invader has been reported for all the native ladybeetle species (Ware & Majerus, 2008; Brown et al., 2011b; Lucas, 2012; Katsanis et al., 2013), negative impact on them should be expected also in urban landscapes. In Eastern Canada, A. bipunctata has disappeared from urban landscapes since the arrival of H. axyridis (Lucas, pers. obs.).

The Asian lady beetle would also probably invade forestal and more-open natural environments. For example, fifteen years after its arrival in Quebec (Canada), the multicoloured H. axyridis was clearly the dominant species in undisturbed wild meadow ecosystems demonstrating that these habitats will not constitute a natural refuge from the invader for native species (Bélanger & Lucas, 2011).

To improve their survival during winter, H. axyridis feeds on sugar-containing fruits like apples, pears or grapes in late autumn (Koch, 2003; Lucas et al., 2007b; Galvan et al., 2008). In the case of grapevines, insects could be harvested with the fruit and during the process beetles release haemolymph containing substances that can alter the smell and taste of wine (see Pickering et al., 2004). The invasion by H. axyridis could then constitute a threat for the fruit and wine industry in Spain.

No parasitized individuals were recorded from the aggregation of Les Llosses in samples taken at the beginning and after overwintering. Furthermore, no parasitized individuals were collected in spring or summer. Dinocampus coccinellae Schrank (Hymenoptera: Ichneumonidae) and Hesperomyces viriscens Thaxt. (Ascomycota; Laboubeniales) are common parasites of coccinellids recorded in Catalonia on several coccinellid species as C. septempunctata, H. variegata, O. conglobata and A. decempunctata (Lumbierres, pers. obs.), so that H. axyridis is susceptible to be parasitized by these parasites. However, a recent study shows that natural enemies are still unable to control populations of H. axyridis in Europe (Raak-van den Berg et al., 2014). In America also, despite the fact that D. coccinellae readily attacks H. axyridis, the success of the parasitism is very low and lower than on other coccinellid species, due to host behavioral defenses and unsuitability (Firlej et al., 2005, 2010).

Van Lenteren et al. (2008) defined H. axyridis as a potentially risky species for Northwest Europe and Brown et al. (2008) predicted than the expansion of the invader in Europe would occur mainly to the northern and eastern countries and that the spread southward is uncertain (see also Kontodimas et al., 2008; Soares et al., 2008). However, the reports of Kulijer (2010) in Bosnia-Herzegovina, Stankovic et al. (2011) in Croatia and the present study, demonstrate the high potential of invasion in southern Mediterranean countries. The Asiatic ladybeetle has a huge phenotypical plasticity. The insect is euryphagous, eurytopic and has successfully colonized drastically different countries and landscapes (Brown et al., 2008). In spite of the different climatic conditions of Spain (oceanic in north and west, continental in center, and Mediterranean in northeast, east and south) we may expect the invasion process to continue until colonization of the complete country, even the sub-tropical area of southwest Almería, Granada, Málaga) may be colonized since H. axyridis was recently reported in Kenya (Nedved et al., 2011). Furthermore, the relatively warm Spanish climate would probably allow a higher survival rate during winter, and would not act as a barrier in absence of human houses such as in Canada (Labrie et al., 2008).

In conclusion, the results of this paper show that H. axyridis is already established in Spain and, although until now is restricted to Catalonia, the invading process is in progress.