Shippings is FREE for USA Orders Over $500!

Reproductive biology and ecological strategies of three species of medicinal Hirudo leeches

Journal of Natural History

Vol. 45, Nos. 11–12, March 2011, 737–747

Reproductive biology and ecological strategies of three species of medicinal leeches (genus Hirudo)

Laima Petrauskiene ̇a, Olga Utevskaand Serge Utevskyc*

aInstitute of Ecology, Nature Research Centre, Vilnius, Lithuania; bDepartment of Genetics and Cytology, V.N. Karazin Kharkiv National University, Kharkiv, Ukraine; cDepartment of Zoology and Animal Ecology, V.N. Karazin Kharkiv National University, Kharkiv, Ukraine

(Received 9 May 2010; final version received 26 October 2010; printed 4 February 2011)

An analysis of the reproduction of three species of medicinal leeches (genus Hirudo) was carried out. The highest fecundity (34.34 ± 3.72 hatchlings per leech) was observed in H. verbana, the lowest in H. medicinalis (11.10 ± 2.56). The heav- iest hatchlings were found in H. medicinalis (0.046 ± 0.0005 g), the lightest in H. verbana (0.032 ± 0.0003 g). Hirudo orientalis had an intermediate fecundity (21.63 ± 3.39 hatchlings per leech) and its hatchlings were of intermediate weight (0.038 ± 0.0005 g). The species differ in their growth rate and mortality: Hirudo ver- bana had the smallest hatchlings and the lowest survivorship but its growth rate was highest among the three species. The results agree with sister relationships between H. medicinalis and H. orientalis and the basal position of H. verbana with respect to them. The reproductive traits of the three species suggest that H. verbana is an r-strategist that inhabits unstable environments, such as temporary ponds in steppe landscapes, whereas its congeners are subject to K-selection and occur in more stable and predictable habitats.

Keywords: Hirudo medicinalisHirudo verbanaHirudo orientalis; breeding; repro- ductive traits.


The medicinal leeches (Hirudo spp.) have been applied in medicine for centuries (Phillips and Siddall 2009) and used as model organisms in various fields of the life sciences (Utevsky and Trontelj 2005; Petrauskiene ̇ 2008). During this long history, medicinal leeches were usually assigned to the species Hirudo medicinalis Linnaeus, 1758, but recently their taxonomic status has been revised dramatically. To date, it is generally accepted that four distinct species of the genus Hirudo – H. medicinalis Linnaeus, 1758, H. orientalis Utevsky and Trontelj 2005, H. verbana Carena, 1820 and H. troctina Johnston, 1810 – range through the western Palaearctic. Species sta- tus for these taxa has been corroborated by morphological differences (Nesemann and Neubert 1999; Utevsky and Trontelj 2005), random amplified polymorphic DNA molecular markers (Trontelj et al. 2004), nuclear and mitochondrial gene sequences (Trontelj and Utevsky 2005; Utevsky and Trontelj 2005) and both mitochondrial gene sequences and nuclear microsatellites (Siddall et al. 2007). These species also differ in their vicariant geographical distribution (Nesemann and Neubert 1999; Utevsky et al. 2010), biochemical composition of saliva (Baskova et al. 2008) and chromosome

*Corresponding author. Email:

ISSN 0022-2933 print/ISSN 1464-5262 online © 2011 Taylor & Francis
DOI: 10.1080/00222933.2010.535918

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

738 L. Petrauskiene ̇ et al.

numbers (Utevsky et al. 2009). The western Palaearctic species of Hirudo can inter- breed in aquaculture, but substantial reproductive barriers have been detected between them (Petrauskiene ̇ et al. 2009).

More recent investigations have revealed preferences of the four species of medic- inal leeches for different landscape types (Utevsky et al. 2010). Hirudo medicinalis occupies the deciduous arboreal zone of Europe. The range of H. verbana largely corresponds to the Mediterranean and sub-boreal steppe zones in Europe, Anatolia and Central Asia, whereas H. orientalis is associated with mountainous areas in the sub-boreal eremial zone and occurs in Transcaucasian countries, Iran and Central Asia. Hirudo troctina has been found in northwestern Africa and Spain in the Mediterranean zone.

The different environmental conditions suggest substantial eco-physiological dif- ferences between the species. Some autecological features of the medicinal leeches have been found. In her pioneering research, Zapkuviene ̇ (1972a) revealed distinct temper- ature optima for cocoon deposition in H. medicinalis and H. verbana (H. medicinalis f. serpentina and H. medicinalis f. officinalis respectively in the original publication). Hirudo medicinalis starts laying cocoons at temperatures 2lower than in H. verbana (Zapkuviene ̇ 1972a). Such a difference may lead to different phenologies in nature, e.g. an earlier start to reproduction of H. medicinalis in habitats with a sympatric occur- rence of H. medicinalis and H. verbana. When living in the same habitat (see Utevsky et al. 2010), the species may tend to be reproductively isolated by the different temper- ature optima in a kind of ecological isolation (see Coyne and Orr 2004). The length of the season favourable to reproduction and the type of water bodies (temporary or sta- tionary) are conditions that, at least in part, determine fecundity and juvenile growth rate. Zapkuviene ̇ (1972a) previously found some differences in fecundities and growth rates of H. medicinalis and H. verbana. Obviously, different environmental conditions lead to divergent biological properties of related species.

This research aims to reveal differences in fecundity, growth rate and survival rate in H. medicinalisH. orientalis and H. verbana and to correlate these with their ecological preferences.

Materials and methods


Fifty individuals of H. medicinalis, 40 individuals of H. verbana and 40 individu- als of H. orientalis were used in breeding experiments. All leeches were more than 14 months old, and were therefore mature (Zapkuviene ̇ 1972b). The H. medicinalis originated from a commercial aquaculture facility in Moscow (Russian Federation) and from ponds in the vicinity of Vasyshcheve (Kharkiv Region, Ukraine). The H. ori- entalis were purchased in a pharmacy in Kharkiv, although it is known that they came from Azerbaijan. Hirudo verbana were obtained from the commercial aqua- culture facility in Moscow; they were probably descended from leeches collected in the Krasnodar Territory (Russian Federation), which is a traditional area for the commercial harvesting of medicinal leeches (see Lukin 1976).

Other factors that can influence fecundity were excluded. It has long been known that the number of hatchlings per cocoon depends on the size of the maternal individ- ual (Sineva 1949). In our experiment, the average weights were 9.81 ± 0.62 g (range 3.05–19.36 g) for H. verbana, 10.33 ± 0.51 g (6.52–13.72 g) for H. medicinalis and

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

Journal of Natural History 739

9.26 ± 0.83 g (4.00–18.72 g) for H. orientalis. An analysis of variance showed that there were no significant differences between the weights of the leeches in these three groups and as a result the impact of the size of maternal individual on fecundity was excluded.


To explore fecundities and other biological traits of H. verbanaH. medicinalis and H. orientalis, breeding experiments were carried out. The leeches were kept and bred according to methods already used by several workers (Zapkuviene ̇ 1972a; Wilkin and Scofield 1991; Davies and McLoughlin 1996; Utevskaya 1998; Petrauskiene ̇ et al. 2009). All adult leeches were kept singly in 1-litre pots with dechlorinated water for 1 month before mating. After that, leeches were fed with bovine or porcine blood in sausage casings at 37C, and placed in pairs for copulation for 1 month more. The temperature was increased to 25C to stimulate copulation.

After the 1-month period for copulation, the leeches were kept individually in con- tainers with wet peat for another month for cocoon deposition. Cocoons were then collected from the peat. The cocoons were counted and the percentage of leeches that laid cocoons was calculated. Some leeches died before placing cocoons in the peat, some died in the peat, and so the percentage of fertile leeches was calculated from the number of leeches that survived in the peat. Every cocoon was placed into its own separate container, and containers were checked daily for 1 month to determine when hatchlings left cocoons. Hatchlings were counted for all mating pairs. If the young leeches did not escape from cocoons after 1 month, cocoons were opened and any live leeches found in these cocoons were counted as hatchlings as well.


The weight of hatchlings was determined immediately after they left the cocoons and after each feeding. Five feedings were performed (at 2, 5, 11, 20 and 36 weeks after hatching). Hatchling weights were measured 2–3 weeks after feeding to provide leeches with an opportunity to remove water and salt ions associated with the blood meal. Medicinal leeches begin to concentrate ingested blood immediately after feed- ing (Boroffka 1968; Zerbst-Boroffka 1973; Wenning 1996); during the 2 weeks after the feeding period they lose 40–50% of their weight, but their weight subsequently stabilizes (Büsing et al. 1953; Zebe et al. 1986; Petrauskiene ̇ 2001, 2004). The leeches were fed with fresh bovine or porcine blood as described above. All groups received the same kind of blood at one time.

Dead hatchlings were counted every day and mortality was expressed as a percentage of the initial number of hatchlings.

Statistical methods

Differences between various experimental groups were determined using Mann– Whitney U-tests (central tendency comparisons) and F-tests (percentage compar- isons). As three groups were compared, the significance level for multiple comparisons (p) was 0.017 after Bonferroni adjustment (p′ p/0.05/0.017, where is the significance level for a single comparison, p′ is the significance level for multiple

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

740 L. Petrauskiene ̇ et al.
comparisons, and is the number of pair comparisons). This method was used for

all traits.



The results of breeding experiments are summarized in Table 1. The highest fecundity was found in H. verbana, the lowest in H. medicinalis.

The percentage of H. verbana that laid cocoons was 91.43%, which did not differ significantly from the percentage, 82.35%, found in H. orientalis. In H. medic- inalis this value was 37.50%, less than half as much as in the other two species and these differences (between H. medicinalis and the other two species) were significant (p0.017).

Other indices of reproductive ability – number of cocoons per leech that survived in the peat and number of hatchlings per cocoon – were also highest in H. verbana and lowest in H. medicinalis with significant (p0.017) differences between them. Hirudo orientalis and H. medicinalis were significantly (p0.017) different only in the number of hatchlings per cocoon.

The total number of hatchlings per leech that survived in the peat was highest in H. verbana and lowest in H. medicinalis, while H. orientalis had an intermediate value and all the differences were significant (p0.017). However, differences for number of cocoons and number of hatchlings calculated from leeches that deposited cocoons were insignificant.

Weight and growth

The weight of hatchlings was highest in H. medicinalis (0.046 ± 0.0005 g) and lowest in H. verbana (0.032 ± 0.0003 g). The weight of H. orientalis hatchlings was intermediate (0.038 ± 0.0005 g). The differences between all groups were significant (p0.017).

The three species had different types of hatchling weight distribution (Figure 1). The distribution of H. verbana had the maximum frequency towards the lower weights with a tail towards high values, whereas the distributions of H. medicinalis and H. orientalis were more symmetrical.

Juveniles received five blood meals over 9 months. Their weight increased and reached 2.47 ± 0.11 g in H. medicinalis, 2.55 ± 0.08 g in H. verbana and 2.74 ± 0.09 g in H. orientalis. Although mean hatchling weight of the three species varied significantly immediately after hatching, these differences were not significant after nine months. Small juveniles of H. verbana were similar in weight to those of H. medicinalis and H. orientalis.

The growth curves (Figure 2) demonstrate that H. medicinalis and H. orientalis had similar patterns of growth, whereas H. verbana initially lagged behind the other two species although its growth rate subsequently increased and caught up. After five blood meals, the body weight of H. medicinalis went up by a factor of 54, H. orientalis by 72 and H. verbana by 79.

Survival rate

Eight months after hatching, 11.06% of H. verbana leeches had died, 2.45% of H. ori- entalis had died and 7.71% of H. medicinalis had died. The survival curves of the three

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

Journal of Natural History 741

Table 1. Reproductive traits of three Hirudo species.


No. of leeches

% of leeches deposited cocoons

No. of cocoons per leech (all leeches)

No. of hatchlings per cocoon

No. of cocoons per leech (deposited cocoons)

No. of hatchlings per leech (all leeches)

No. of hatchlings per leech (deposited cocoons)

Hirudo verbana Hirudo orientalis Hirudo medicinalis

35 34 48

91.43 ± 4.73 82.35 ± 6.54 37.50 ± 6.99

3.29 ± 0.277 2.53 ± 0.356 1.65 ± 0.366

10.45 ± 0.710 8.55 ± 0.592 6.73 ± 0.434

3.59 ± 0.291 3.07 ± 0.393 4.39 ± 0.599

34.34 ± 3.72 21.63 ± 3.39 11.10 ± 2.56

36.25 ± 3.56 26.29 ± 4.14 31.29 ± 3.63

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

742 L. Petrauskiene ̇ et al.

H. verbana


Weight, mg

H. orientails


Weight, mg

H. medicinalis


Weight, mg

Figure 1. The hatchling weight distribution of three Hirudo species. (A) Hirudo verbana; (B) Hirudo orientalis; (C) Hirudo medicinalis.

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

Percent of individuals Percent of individuals Percent of individuals

Journal of Natural History 743

Number of meals

Figure 2. Growth curves of three Hirudo species demonstrating that H. medicinalis and H. orientalis had similar patterns of growth, whereas H. verbana initially lagged behind the other two species but its growth rate subsequently increased and caught up.

Hirudo species (Figure 3) demonstrate similar survival patterns in H. orientalis and H. medicinalis. The juveniles of H. verbana faced a high mortality for the first months after hatching.

The different rates of survival resulted in the highest total number of adult off- spring per leech that survived in the peat in H. verbana (30.54 ± 3.31) and the lowest number in H. medicinalis (10.24 ± 2.36), while H. orientalis had an intermediate value (21.10 ± 3.31). All differences were significant (p0.017). The numbers of adult off- spring per fertile leech were 33.41 ± 3.46, 27.33 ± 3.85 and 25.61 ± 3.65 in H. verbanaH. medicinalis and H. orientalis, respectively. These values did not differ significantly.


Differences and similarities among Hirudo species

The breeding experiment demonstrated that Hirudo species differ in their reproduc- tive characteristics. Differences were found in many fecundity traits: the proportion of

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

Weight, g

744 L. Petrauskiene ̇ et al.

Months after hatching

Figure 3. Survival rates of Hirudo species demonstrating similar survival patterns in H. orien- talis and H. medicinalis and a higher mortality in H. verbana.

leeches that laid cocoons, the number of cocoons, and the number of hatchlings cal- culated from individuals that survived in the peat. The highest fecundity was observed in H. verbana, the lowest in H. medicinalis. On the other hand, the heaviest hatch- lings were found in H. medicinalis, whereas H. verbana produced the smallest juveniles. Hirudo orientalis had intermediate fecundity, and its hatchlings were intermediate in size. The species also differed in their growth rate and mortality. Hirudo verbana had the smallest hatchlings and the lowest survivorship, but its growth rate was highest among the three species.

Two of the species, H. medicinalis and H. orientalis, bear a great resemblance to each other in a number of traits. They demonstrated similarities in distribution of hatchling weight and in growth and survival rates. This result agrees with the sister relationship of H. medicinalis and H. orientalis and the basal position of H. verbana

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

Survival, %

Journal of Natural History 745

in relation to them as revealed by a phylogenetic analysis (Trontelj and Utevsky 2005; cf. Phillips and Siddall 2009). The same pattern of relationships was deduced from external morphology (Utevsky and Trontelj 2005) and biochemical composition of saliva (Baskova et al. 2008).

Life strategies of Hirudo species

The results demonstrate that H. verbana has the highest fecundity and juvenile mor- tality and small juvenile body size, whereas H. medicinalis and H. orientalis have the opposite characteristics. Though numbers of cocoons, hatchlings and adult off- spring calculated per fertile leech do not differ significantly, the percentage of fertile individuals is significantly lower in H. medicinalis in comparison with its congeners. This suggests greater selection for mate choice and lower tolerance to aquaculture conditions in H. medicinalis. The two distinct patterns of reproduction and post- embryonic development should be assigned to r-strategies and K-strategies, respec- tively. Hirudo verbana may be characterized as an r-strategist that inhabits unstable environments, such as temporary ponds in arid steppe landscapes (Utevsky et al. 2008, 2010). Its high fecundity and small juvenile body size are traits that have been shaped by r-selection. Conversely, H. medicinalis produces fewer juveniles with larger body sizes and lower mortality rates. Obviously, this species is subject to K-selection. Hirudo orientalis occupies an intermediate position between the two strategies or verges towards H. medicinalis. Both live in more stable and predictable environments including static ponds and lakes of the arboreal zone in Europe and mountainous landscapes of the Caucasus and Central Asia.

Different life strategies of the three Hirudo species may be responsible for asym- metric patterns of the reproductive isolation between them (see Petrauskiene ̇ et al. 2009). Hirudo verbana and H. orientalis were more fertile in interspecific crosses than H. medicinalis. Being fertilized by other species, H. verbana and H. orientalis showed a moderate decrease in fecundity, which was caused by the decrease of hybrid offspring fitness: low weight, slow growth and high mortality. The third species, H. medicinalis, had the lowest fecundity in the hybrid crosses. In crosses with H. verbanaH. medic- inalis did not yield adult offspring and in matings with H. orientalis the number of adult offspring was fewer than in intraspecific crosses by a factor of five.

We suppose that the asymmetric pattern of reproductive isolation may be caused by different life strategies of these species. As a K-strategist, H. medicinalis is more discriminating in mating and a possible loss in fitness is more substantial in the case of a wrong choice compared with species that produce more offspring. That explains why natural selection prevents the hybridization of H. medicinalis with individuals of other species.

On the other hand, H. verbana, as an r-strategist, has more numerous offspring. A possible loss in fitness would not be so critical in this species when mating with indi- viduals of another species compared with the K-selected H. medicinalis. Consequently, in H. verbana isolating barriers are not very strict.

The r-selected features of H. verbana suggest that this species is most suitable for breeding in aquaculture compared with other medicinal leeches, which are adapted to the specific conditions of their natural habitats. Indeed, recently it has been found that leeches marketed as H. medicinalis are actually H. verbana (Siddall et al. 2007). In spite of the fact that viable populations of H. medicinalis still persist in Europe, it

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

746 L. Petrauskiene ̇ et al.

is H. verbana that is sold by commercial facilities and widely used in medicine as an effective remedy and in neurobiology as a model organism (Siddall et al. 2007) because of its ability to tolerate artificial conditions.


This research was supported by INTAS grant no. 05-1000008-8147 “The medicinal leech (Hirudo spp.), famous and unknown: taxonomy, conservation, and medical applications.”


Baskova IP, Kostrjukova ES, Vlasova MA, Kharitonova OV, Levitskiy S, Zavalova LL, Moshkovskii SA, Lazarev VN. 2008. Proteins and peptides of the salivary gland secre- tion of medicinal leeches Hirudo verbanaH. medicinalis, and H. orientalis. Biochemistry (Moscow). 73:315–320.

Boroffka I. 1968. Osmo- und Volumenregulation bei Hirudo medicinalis. Z Vergl Physiol. 57:348–375.

Büsing KH, Döll W, Freytag K. 1953. Die Bakterienflora der medizinischen Blutegel. Arch Microbiol. 19:52–86.

Coyne JA, Orr HA. 2004. Speciation. Sunderland (MA): Sinauer Associates, Inc.
Davies RW, McLoughlin NJ. 1996. The effects of feeding regime on the growth and reproduc-

tion of the medicinal leech Hirudo medicinalis. Freshwater Biol. 36:563–568.
Lukin EI. 1976. Piyavki presnykh i solonovatovodnykh vodoemov [Leeches of fresh and

brackish waters]. Leningrad: Nauka.
Nesemann, H., Neubert, E., 1999. Annelida: Clitellata: Branchiobdellida, Acanthobdellea,

Hirudinea. In: Süβwasserfauna von Mitteleuropa, 6/2. Heidelberg, Berlin: Spektrum

Akademischer Verlag.
Petrauskiene ̇ L. 2001. Water toxicity assessment using medicinal leeches. Aquat Ecosyst Health.

Petrauskiene ̇ L. 2004. Medicinal leech as a convenient tool for water toxicity assessment.

Environ Toxicol. 19:336–341.
Petrauskiene ̇ L. 2008. The use of the medicinal leech (
Hirudo sp.) in ecotoxicological and other

scientific research—a short review. Lauterbornia 65:163–175.
Petrauskiene ̇ L, Utevska O, Utevsky S. 2009. Can different species of medicinal leeches (

spp.) interbreed? Invertebr Biol. 128:324–331.
Phillips AJ, Siddall ME. 2009. Poly-paraphyly of Hirudinidae: many lineages of medicinal

leeches. BMC Evol. Biol. 9: 246.
Siddall ME, Trontelj P, Utevsky SY, Nkamany M, Macdonald KS III. 2007. Diverse molecular

data demonstrate that commercially available medicinal leeches are not Hirudo medicinalis.

Proc. R. Soc. Lond. Ser. B. 274:1481–1487.
Sineva MB. 1949. Biologicheskie nablyudeniya nad razmnozheniem meditsinskoy piyavki

[Biological observations on propagation of the medicinal leech]. Zool Zh. 28:213–224. Trontelj P, Sotler M, Verovnik R. 2004. Genetic differentiation between two species of the medicinal leech, Hirudo medicinalis the neglected H. verbana, based on random amplified

polymorphic DNA. Parasitol Res. 94:118–124.
Trontelj P, Utevsky SY. 2005. Celebrity with a neglected taxonomy: molecular systematics of

the medicinal leech (genus Hirudo). Mol Phylogenet Evol. 34:616–624.
Utevskaya OM. 1998. Analysis of reproductive ability of the medicinal leech (
Hirudo medici-

nalis) bred under laboratory conditions. Vestn Zool. 32(1–2):119–122.
Utevsky S, Kovalenko N, Doroshenko K, Petrauskiene ̇ L, Klymenko V. 2009. Chromosome

numbers for three species of medicinal leeches (Hirudo spp.). Syst Parasitol. 74:95–102.

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

Journal of Natural History 747

Utevsky SY, Trontelj P. 2005. A new species of the medicinal leech (Oligochaeta, Hirudinida, Hirudo) from Transcaucasia and an identification key for the genus Hirudo. Parasitol Res. 98:61–66.

Utevsky S, Zagmajster M, Atemasov A, Zinenko O, Utevska O, Utevsky A, Trontelj P. 2010. Distribution and status of medicinal leeches (genus Hirudo) in the Western Palaearctic: anthropogenic, ecological, or historical effects? Aquat Conserv. 20:198–210.

Utevsky SY, Zinenko AI, Atemasov AA, Huseynov M.A, Utevska OM, Utevsky AY. 2008. New information on the distribution of the medicinal leech (genus Hirudo) in the Iberian Peninsula, the Caucasus and Central Asia. Lauterbornia 65:119–130.

Wenning A.1996. Managing high salt loads: from neuron to urine in the leech. Physiol Zool. 69:719–745.

Wilkin PJ, Scofield AM. 1991. Growth of the medicinal leech, Hirudo medicinalis, under natural and laboratory conditions. Freshwater Biol. 25:547–553.

Zapkuviene ̇ D. 1972a. Razvedenie i vyrashchivanie meditsinskikh piyavok v laboratornykh usloviyakh (1. Razvedenie Hirudo medicinalis f. serpentina H. medicinalis f. officinalis). [Breeding and growing of medical leeches under laboratory conditions (1. Breeding of Hirudo medicinalis f. serpentina and H. medicinalis f. officinalis)]. Lietuvos TSR Mokslu ̨ akademijos darbai. C serija. 3(59):71–76.

Zapkuviene ̇ D. 1972b. Razvedenie i vyrashchivanie meditsinskikh piyavok v laboratornykh usloviyakh (2. Vyrashchivanie Hirudo medicinalis f. serpentina). [Breeding and grow- ing of medical leeches under laboratory conditions (1. Growing of Hirudo medicinalis f. serpentina)]. Lietuvos TSR Mokslu ̨ akademijos darbai. C serija. 3(59):77–84.

Zebe E, Roters FJ, Kaiping B. 1986. Metabolic changes in the medicinal leech Hirudo medicinalis following feeding. Comp Biochem Phys A. 84:49–55.

Zerbst-Boroffka I. 1973. Osmo- und Volumenregulation bei Hirudo medicinalis nach Nahrungsausfnahme. J Comp Physiol A. 84:185–204.

Downloaded By: [Colorado State University] At: 15:16 27 May 2011

Share this post