Portal de Periódicos da CAPES

Tree-hole breeding mosquitoes in Israel

2012; Wiley; Volume: 37; Issue: 1 Linguagem: Inglês

10.1111/j.1948-7134.2012.00206.x

1948-7134

Günter C. Müller, Vasiliy D. Kravchenko, Amy Junnila, Yosef Schlein,

Viral Infections and Vectors

A survey was conducted to evaluate the number of tree-hole breeding mosquito species and their distribution in the six principal woodland types in Israel. Out of approximately 3,000 mature trees examined, only 38 contained holes that retained water for extended periods of time, and breeding mosquitoes were observed in 27 of them. Two specialized tree-hole breeders, Aedes pulchritarsis Rondani and Aedes geniculatus Oliver, were found breeding at several sites in northern Israel, always at locations 500 m above sea level (a.s.l) and with high annual precipitation. Aedes albopictus Skuse which, in Israel, is known as an opportunistic container breeder, was found in this study to have adapted remarkably well to breeding in tree holes and was found in most forest types investigated and in most tree species which had adequate tree holes. Two other species, Culiseta annulata Schrank and Culex pipiens Linnaeus instars, were found in one of the tree holes, but did not survive to reach maturity. Water-filled tree holes provide a unique habitat to which only a few highly specialized invertebrate communities are adapted. This type of microhabitat is typically restricted to mature, often deciduous trees (Jenkins and Carpenter 1946). Due to a long history of human habitation, most of the woodlands in the Mediterranean are synanthropic; mature forests are rare, and degraded shrub lands are often the rule. As such, only six stenobiontic tree-hole breeding mosquito species, and three eurybiontic species, which only occasionally breed in tree holes (Table 1), have been identified in the Eastern Mediterranean (Ramsdale et al. 2000, Samanidou-Voyadjoglou and Darsie 1993, Shannon and Hadjinicolaou 1937). Though some studies on the ecology of these species were conducted in western Europe, (Bradshaw and Holzapfel 1986, 1992, Marshall 1938, Service 1971, Yates 1979), northern Europe (Wesenber-Lund 1920, Natvig 1948), and central Europe, little is known from the Levant, which has been documented as the natural southern distribution border of tree-hole breeders in the eastern Mediterranean (Knio et al. 2005, Margalit and Tahori 1970). This survey was conducted to evaluate the distribution and abundance of tree-hole breeding mosquitoes in the principal woodland areas of Israel. Israel can be roughly divided into three longitudinal landscape units: the Coastal Plain, the Central Mountain Ridge, and the Rift Valley (Orni and Efrat 1980). This division, along with north-south and east-west temperature and precipitation gradients, creates a diverse range of microhabitats over a relatively small area (Figure 1). The northern part of Israel includes Mt. Hermon (2,200 m a.s.l.), which receives annual snow and contains typical tragacanth vegetation, whereas the Dead Sea area is about 400 m below sea level (b.s.l.) and contains pockets rich in Afrotropical fauna and flora (Zohary and Orshansky 1949, Bytinski-Salz 1961). The north and center of the country is temperate (Mediterranean), while the southern and eastern parts are semi-arid (Irano-Turanian grassland) and arid (desert) (Danin 1992). In terms of precipitation, the short winter accounts for 70% of the annual rainfall that occurs between November and February. Rain from May to September is negligible; the dry season lasts from June to August. Larval tree hole sites of Ae. geniculatus, Ae. pulchritarsis, and Ae. albopictus in conjunction with the major forest types and annual precipitation. The principal forest types in Israel are: Oro-Mediterranean (Montane) forest, Maquis forest, Oak Park forest, Carob Park forest, Savannoid Mediterranean woodlands, and Synanthropic woodlands. Oro-Mediterranean forests are located on Mt. Hermon (above 1,300 m). The dominant trees are deciduous oaks (Quercus boissieri Reut., Q. libani Olivier Fagaceae) and maples (Acer microphyllum Hort. Aceraceae). These forests are scattered on slopes facing south, but in canyons they are contracted and occasionally closed. Evergreen sclerophyllous Maquis covers most of the northern central mountain range and in areas with high precipitation the area is dominated by dense thickets of evergreen oak (Quercus calliprinos Webb), and its companions. In the Judean Mountains, the number of mesophytic components decreases gradually, allowing for woodlands dominated by Pinus halepensis Mill Pinaceae and Arbutus andrachne L. Ericaceae. In drier areas, only small isolated woods and hedges can be found, occasionally accompanied by Q. calliprinos, Pistacia palaestina (Boiss.), and P. atlantica (Desf. Anacardiaceae). North-facing slopes and narrow canyons, especially in the northern Galilee where humidity is relatively high throughout the year, are usually inhabited by pockets of winter deciduous woodlands dominated by Q. boissieri. Some of the streamlets in the northern Mediterranean zone, like Nahal Skziv, are interspersed with small patches of riparian forests dominated by sycamore (Platanus orientalis L. Platanaceae), poplar trees (Populus euphratica Oliv.), and willows (Salix L. species) Salicaceae are often present with a rich undergrowth of herbaceous plants. There are two types of park forest in Israel: Oak Park forest (Quercus ithaburensis Decne.), and Carob Park forest (Ceratonia siliqua L. Park Forest). The Oak Park forests are xerotherm, yet occur exclusively in the northern parts of the country. Depending on the stand, Q. ithaburensis is accompanied by a large variety of other trees. Scattered forests usually have rich undergrowth composed of shrubs, semi-shrubs, grasses, and many herbaceous species. The Carob Park forests are found in hot and dry areas from 0 to 300 m a.s.l. They are found on all the limestone hills at the foot of the Central Mountain Range of the Mediterranean zone in Judea, Samaria, Carmel, Gilboa, and the Galilee. Depending on the stand, these communities are often accompanied by olive trees (Olea europaea L. Oleaceae) and Q. calliprinos trees and bushes. The undergrowth of this scattered woodland is not as rich as in Q. ithaburensis Park forests. The Savannoid Mediterranean shrub and woodland is restricted to the lower regions of the southeastern Galilee, especially by the Sea of Galilee and the adjacent southern Jordan Valley. This dry, forested grassland is dominated by Ziziphus lotus (L.) Lam., Z. spina-christi (L.) Desf. Rhamnaceae, and occasionally Faidherbia (Acacia) albida (Delile) A. Chev. Mimosaceae. Synanthropic woodland, intensively managed or planted by humans, replaced many natural forests in Israel. Typical elements on the Coastal Plain are citrus, mangos (Mangifera indica L. Anacaridiaceae), and to a smaller extent hickory orchards (Carya illinoinensis (Wangenh.) Koch Juglandaceae). On the central mountain range, olive and almond trees (Amygdalus communis L. Rosaceae) are most abundant, whereas in the Northern Galilee, apple and cherry plantations are more common. Pine forests, mainly P. halepensis, and to a smaller extent Pinus pinea L. and Cupressus sempervirens L. Cupressaceae, were planted throughout Israel as far south as the northern and western Negev Desert. Eucalyptus groves, (several species, but especially Eucalyptus camaldulensis Dehn. Myrtaceae), are abundant along the coastal plain, the larger inland valleys, along streamlets, springs, around settlements, and along roads. Numerous exotic species of trees can be found in parklands and gardens in most towns in Israel. (Danin 1992, Eig 1926, Zohary 1973). All principal woodland types found in Israel were surveyed during mid-January to mid-February, 2006. In each forest type, about 500 mature trees were checked for holes. Potential areas were investigated by direct sight and also using a round mirror (15 cm diameter) mounted on a telescopic stick which could stretch up to 3 m long. For preliminary verification of water presence and depth, holes were examined using a flexible wire probe with paper wrapped around the distal end (50 cm). The following parameters were recorded for each hole: forest type, sun exposure (north or south-facing slope), tree species, height above ground, opening size, and whether it was a rot or a pan hole. All tree holes with water were then visited every ten days from February 20, 2006 until May 10, 2006 and then again before the first rains: from the first of November until the end of April, 2007. Later, during the dry summer months, the holes were visited only once a month. On each visit, if water was present, the entire water content of the holes was siphoned, its volume measured, the larvae defined and counted, and they were returned with all the water back into the tree hole (Bradshaw and Holzapfel 1986). Tree-holes in which water was observed at any time during the survey (65) were sampled during August, 2007, after all the holes had dried up. Samples of 50 g of detritus were scratched from the bottom and the side walls of the holes and incubated in the laboratory, under standard insectary conditions, in beakers with one liter of distilled water for up to one month. Hatching was observed during the following three months. During the survey, almost 3,000 mature trees were checked, from which about 10% had some kind of holes, with most either permanently dry or filled with detritus. In about 2% of the inspected trees, water was found at some point in the survey. Most of the holes with water were found in broad-leafed trees, including oaks, eucalyptus, hickory, olive, sycamore, pistachio, mango, poplar, and willow (Table 1). Very few were found in conifers, including C. sempervirens, P. pinea, and P. halepensis. Most of the holes with water were rot holes, but a few were pan holes found in the partially flooded forest of Tel Dan Nature Reserve, in the bottom of the Nahal Sksiv canyon and on the coastal plain east of Tel Aviv. Water capacity of pan-holes never exceeded 500 ml, which was typically maintained for less then two weeks after rainfall, and larvae or pupae were never observed in them. The height of rot holes ranged from 10 to 235 cm above the ground, the size of their apertures from 30 to 4,200 cm2, and the maximum encountered water capacity during the study period in these holes was from 100 to 8,500 ml. Only about two-thirds of these rot holes (38) held water longer than one month, and larvae/ pupae of mosquitoes were observed in only 27 of them. These tree holes retained water from the beginning of the rainy season in early November, 2007, and were completely dry four weeks after the last strong rains, in mid-April. With the exception of a single tree hole in an oak tree, all the active holes were found in mesophile forests located in shady areas, such as north-facing slopes, river canyons, and in one specific case, the shade of a church. Evidence of larvae and/or pupae was discovered in 27 out of 38 suitable tree holes containing water for more than one month. These 27 tree holes were distributed among nine tree species found in the six principal forest types of Israel (Table 2). Five mosquito species were breeding in their tree holes. They were Aedes geniculatus, Aedes pulchritarsis, Ae. albopictus, Cs. annulata, and Cx. pipiens. Four larval sites of Ae. geniculatus were found, all in the northern-most part of Israel. Three were located in deciduous Maquis and one in the nearby Oro-Mediterranean forests on Mt. Hermon. The first site was in a narrow, shady canyon in the upper part of Nahal Sksiv, about 650 m a.s.l. Ae. geniculatus breeding was observed in the large rot hole of a partially burnt stump of a P. orientalis. The second site was on the north-facing slope of Mt. Meron, about 950 m a.s.l. in a dense shady oak stand. Larvae of Ae. genculatus were observed in a deep hole with a narrow opening in a mature specimen of Q. boissieri. The third site was found on the north-facing slope of Mas'ada, a broad-leaved deciduous forest in the Golan Heights (1,000 m a.s.l). In the montane forest of Mt. Hermon, one Ae. geniculatus larval site was found in a severely damaged Q. boissieri at the bottom of a canyon (1,650 m a.s.l). First and second instar larvae of Ae. geniculatus were collected throughout the entire period in which water was present. Later instars were only observed from March until the holes dried up and again in small amounts several weeks after the first rainfall from mid-November to early December. Although water contents of Q. calliprinos tree holes were consistently less than those in the P. orientalis tree holes, the former were preferred, containing several times more larvae then the latter. Breeding of Ae. pulchritarsis was mostly confined to northern Israel; five sites were in the Galilee, one was on Mt. Hermon, one was in the Golan Heights, and one was on the Carmel Mountain Ridge; all of these are located in mountainous areas. Half of the breeding sites were in Maquis, about a third in Q. ithaburensis Park Forests and only one in Synanthropic Parkland (Table 2). Most breeding sites were in shady locations like north-facing slopes and canyons. The exception is one site in a scattered Q. ithaburensis park forest on a south-facing slope. In the montane forests of Mt. Hermon, breeding of Ae. pulchritarsis was observed in the same tree hole utilized by Ae. geniculatus, described above. The only true Synanthropic breeding site was in urban Beit Shalla, east of Jerusalem, in a mature eucalyptus tree. First and second instar larvae were observed in the tree holes throughout the entire period in which water was present, although more were observed during the spring. Older instars were seen from March until the holes dried up. An exceptionally prolific Ae. pulchritarsis site was found close by in Mas'ada village. Hundreds of larvae of all instars were observed in two ornamental wooden buckets that were situated in a shady area of a garden with several mature olive and oak trees. The water content of the buckets fluctuated between 1.5 to 6.0 liters between mid-July to the end of August, which was the period the instars were observed. First and second instar larvae, a single egg raft of Cx. pipiens, and a few second instar larvae of Cs. annulata, together with Ae. pulchritarsis, were discovered once in February in the biggest rot hole with the largest water capacity in a Q. ithaburensis tree. The site was located in a Q. ithaburensis Park Forest, on a south-facing slope. On consecutive visits, only early stages of Ae. pulchritarsis were found; Cx. pipiens and Cs. annulata were absent. During the survey, no active Ae. albopictus breeding habitats were observed above 500 m, at least in natural tree holes. Ae. albopictus was by far the most common and versatile species. It was found mainly in Synanthropic woodland like mature mango and hickory orchards, and in Synanthropic parkland (Table 2). In natural habitats, it was found in Maquis, Quercus ithaburensis, and C. siliqua/P. lentiscus park forests (Table 2). All instars could be found in the tree holes one month after the hole filled with water and they remained until the holes dried up. The productivity of the holes was not connected to their size, water capacity, or tree species. Actually, Ae. albopictus was able to utilize the smallest holes, often with very small water capacities (Table 2). Wherever Ae. albopictus was found, no larvae of any other species were ever observed. During the survey, this species was frequently observed in natural forests breeding in various types of disposed containers such as plastic bags, beverage cans, buckets, and tires. In synanthropic woodlands and gardens, it was found breeding in ornamental and disposed containers, as well as tiny irrigation puddles. Detritus samples were collected from all 65 tree holes in which water was observed and brought to the laboratory for incubation. No hatching of mosquito larvae was observed from the ten pan hole samples. From 26 out of 27 active rot holes, viable eggs of all three Aedes species were recovered and successfully hatched in the laboratory. The bulk of Ae. geniculatus larvae hatched quickly; 87% hatched after five days, 96% after ten days, and the remainder during the following five days. On the other hand, Ae. pulchritarsis eggs hatched much later; 45% hatched after five days, 63% after ten days, 86% after 20 days, and 96% after 30 days. Ae. albopictus eggs hatched in a time frame between the other Aedes species; 61% hatched after five days, 75% after 10 days, 94% after 20 days, and the rest in the next ten days. In one sample taken from an Ae. geniculatus site, no larva were observed following incubation. Generally, the newly hatched larvae were from the same species that were observed occupying the site; the exception was an Ae. geniculatus and a Ae. pulchritarsis site where only Ae. albopictus were hatching. From another Ae. pulchritarsis site (500 m a.s.l.), about 80% of the hatching larvae were Ae. albopictus and the rest were Ae. pulchritarsis. From the 11 tree holes in which water was present for more than one month, but without any record, two samples with Ae. albopictus were recovered. From the 17 tree holes with water retained less than one month and no breeding record, three samples with hatching Ae. albopictus were obtained. These five sites were on the coastal plain/foothills along the Central and Carmel Mountain Ridges. The number of hatching eggs varied among species: out of 50 g of detritus, an average of 35 Ae. geniculatus larvae hatched, while from similar samples an average of 59 Ae. pulchritarsis larvae and 253 Ae. albopictus larvae were counted. To date, about 30 mosquito species have been documented in Israel (Orshan et al. 2008, Margalit and Tahori 1970). Of these, only three species Ae. geniculatus, Ae. pulchritarsis and Ae. albopictus were observed regularly in tree holes during our survey. Ae. albopictus, an invasive species, has been in Israel since 2002 and since then, it has continuously spread along the coastal plain (Pener et al. 2003). This species is a known to opportunistically occupy containers, rather than obligately (Tandon and Ray 2000). A fourth native container/tree-hole breeder, Ae. aegypti, was widespread in the lowlands of the temperate zone in northern and central Israel until the 1930s. Despite a brief reoccurrence in the mid 1970s along the coastal plain, it most likely became extinct shortly thereafter (Pener-Salomon and Vardi 1975). In previous surveys which utilized either CDC light traps (incandescent light) or CO2-baited traps, Ae. pulcritarsalis was rarely collected and Ae. geniculatus was recorded only once (Margalit and Tahori 1970). However, a recent faunistic survey of Lepidoptera and Coleoptera with miniature UV traps yielded fairly good catches of both species (Müller et al. 2005, 2006). In Israel, Ae. pulchritarsis feeds readily on people, while Ae. geniculatus was never observed to do so, although this is usual behavior in its northern distribution area (Yates 1979). The distribution of tree-hole breeding mosquitoes depends on the distribution of mature deciduous trees, the precipitation frequency, and the evaporation rate due to solar radiation. Accordingly, the distribution center of the two palearctic species Ae. geniculatus and Ae. pulchritarsis is in the forest belt of temperate Europe and Asia. Many animal and plant species found in Israel reach their most marginal point of geographic distribution, and this is especially true for woodland species (Bodenheimer 1935, Furth 1975). Generally, both plant and animal species gradually become infrequent at the periphery of their geographic distribution before they finally vanish (Hengefeld and Haeck 1982, Brown 1984). This also seems to be the case with the two Aedes species. Though Ae. geniculatus and Ae. pulcritarsalis are generally woodland species, they are specific in their habitat preferences. Both species were only found in forested mountainous areas (≥500 m a.s.l.) with high annual precipitation (≥700 mm). Both species were most common in the north, and Ae. geniculatus was not observed south of the Nehring Line (Northern Galilee – Golan Heights). Both Ae. pulchritarsis and Ae. geniculatus were found in mesophilic woodlands with mature trees, the latter species exclusively. Although Ae. pulchritarsis penetrated into xerotherm forests, breeding was still restricted to shady areas with high precipitation. In areas with low precipitation and along the coastal plain, both species were absent despite the abundance of suitable tree holes which retained water for ample durations. Ae. geniculatus has thus far only been found in natural habitats, whereas Ae. pulchritarsis has been collected in both natural woodlands and synanthropic woodlands. At some sites, Ae. pulchritarsis were found in wooden buckets, although suitable tree holes were present in the immediate vicinity. This occurred throughout the year, including the dry season (unpublished data). The same species was observed in Greek villages without any mature trees and far from forests (Shannon and Madjinicolaou 1937). It is not clear how important these secondary habitats are for Ae. pulchritarsis, but they might partly explain the numerous records of adults from urban Jerusalem and its surrounding villages (unpublished data). During the dry season, viable eggs were recovered from both of the local Aedes species and from almost all of the sites with evidence of previous breeding. In a small riverine forest (500 m a.s.l.), close to the coastal plain, the two species were possibly replaced by the invasive Ae. albopictus. It is important to point out that no hatching larvae were observed from any other samples taken from the water-filled holes. It seems that both species are k-strategists with a low rate of dispersal from their breeding habitats (McNeely 1994). Though some adult females of Ae. pulchritarsis and Ae. geniculatus were regularly caught in UV CDC traps and were observed resting during the daytime in tree holes, it remains unclear if these specimens were old aestivating specimens, or if there was some continuous breeding activity in unidentified tree holes or even in other secondary habitats. Nevertheless, in Israel the bulk of the population of Ae. pulchritarsis and Ae. geniculatus seem to pass the dry season in the oval stage. The differential hatching time of the two species might also explain their different distribution patterns. Delayed hatching of Ae. pulchritarsis might be the reason why this species is more widespread in Israel then Ae. geniculatus, especially towards the drier south. Half of the eggs hatch within a very short period following introduction of water and develop rapidly, while the other half hatch gradually. This mechanism might ensure survival of viable eggs for the next rainfall, if water availability is sporadic due to low precipitation or high evaporation rates. The few eggs of Ae. geniculatus hatch faster, which might explain its restriction to places with high and reliable precipitation; dispersion would be more of a risk than an advantage. Ae. albopictus is mainly an opportunistic container breeder (Comiskey et al. 1999, Knudsen 1995). In our survey, this is the first time this species has been found breeding in natural tree-hole habitats in Israel. Moreover, it was the most versatile tree-hole breeder found in all types of woodlands, from sea level up to 500 m a.s.l., excluding the Montane forests on Mt. Hermon. From the nine tree species that were occupied by tree-hole breeders (Table 2), Ae. albopictus was found in eight. However, in Israel its main breeding site is in garbage and ornamental containers as well as in irrigation puddles. Garbage in Israel is often deposited in and around forests and might have been a stepping stone to tree holes. Most locations where Ae. albopictus was found would probably not have been suitable for the other Aedes species to breed; Ae. albopictus is invading the natural habitats of the two local species. Along the coastal plain, Ae. albopictus might replace the two species in the long run. It appears that Ae. albopictus is also invading habitats of stenobiontic tree-hole breeding mosquitoes in the Americas (Barrera 1996, Juliano et al. 2004, Sota et al. 1992). Tree holes are unique breeding habitats, and only a few mosquito species are fully adapted to this specialized ecological niche (Bradshaw and Holzapfel 1986, Woodward et al. 1988). It is therefore not surprising that Cx. pipiens and Cs. annulata were observed for the first time in this habitat in Israel. They apparently did not manage to breed there successfully. The large opening of this particular hole, the exposure to sunlight, or the relative scarcity of breeding sites in general might have encouraged them to oviposit there.