OUP user menu

Effects of organophosphate and synthetic pyrethroid sheep dip formulations on protozoan survival and bacterial survival and growth

Tatiana K. Boucard , Jackie Parry , Keith Jones , Kirk T. Semple
DOI: http://dx.doi.org/10.1016/S0168-6496(03)00253-8 121-127 First published online: 1 January 2004

Abstract

Sheep dipping with organophosphate or synthetic pyrethroid-based formulations is still widely used by farmers in the UK to control ectoparasites and results in 175–220 million litres of spent sheep dip produced each year. Spent sheep dip may be diluted in animal slurry or water prior to disposal onto land. However, the effects of this practice on the microbial ecology of animal slurries, soil and aquatic systems are still relatively unknown. This paper investigated the effect of Bayticol (synthetic pyrethroid sheep dip) and Ectomort (organophosphate sheep dip) concentrations on (i) the survival of 15 protozoan species, (ii) the recovery of the four species of amoebae, and (iii) bacterial survival and growth. This investigation found that overall Bayticol was less toxic to protozoa than Ectomort, with minimum inhibitory concentrations ranging from 0.01 to 0.03% (v/v) and 0.005 to 0.06% (v/v), respectively. Amoebic cysts remained viable and emerged from dormancy, thereby pointing to the potential for recovery of protozoan communities in contaminated environments. The presence of sheep dips did not affect bacterial survival and growth on agar; however, the five test bacteria were not able to utilise the sheep dips as sole carbon sources. These findings have implications for the contamination of animal slurries, soil and aquatic systems, in that there is the potential for significant increases in microbial numbers, containing putative pathogens due to the diminution of bacteriophagous protozoan populations.

Keywords
  • Sheep dip formulation
  • Protozoon
  • Bacterium
  • Toxicity
  • Risk assessment

1 Introduction

Sheep suffer from infestation by a variety of ectoparasites (scab mites, blowfly, ticks and lice) and may be treated by ‘dipping’, i.e. immersion in a mixture of water and an insecticide formulation [1]. Although no longer compulsory in the UK since 1992, sheep dipping is still widely used by farmers for economic, cosmetic and welfare reasons [2].

There are two major classes of sheep dip: organophosphates (OPs), containing the active ingredient diazinon, and synthetic pyrethroids (SPs), containing the active ingredient cypermethrin or flumethrin [2]. Previously, the OPs propetamphos and chlorfenvinphos were also used. SPs were introduced following increasing concerns over the potential effects of OPs on the health of farmers from acute and chronic exposure [3]. However, OPs are still widely used, with 50.2 tonnes sold in the UK in 1998 compared to an estimated 5.8 tonnes of SP sheep dips sold in the same year [4].

Disposal of spent sheep dip remains a contentious issue. Although the quantity of waste sheep dip produced by each operation is relatively small [5], the toxic nature of the compounds, together with an overall figure of 175–220 million litres of waste sheep dip produced each year [6], highlight the need for clear guidelines and safe practices. The high cost and lack of facilities for disposal of waste sheep dip by chemical treatment, land filling and incineration has led the Environment Agency [5] to identify disposal onto land as the most practical option. Thus, waste sheep dip may be spread undiluted at a rate of no more than 5000 l ha−1, or diluted threefold in water or slurry to compensate for the limitations of slurry tankers, which cannot operate at rates lower than 20 000 l ha−1[5]. However, concerns have arisen, as both slurry and sheep dip insecticides contribute to contamination of aquatic systems [1,2,7]. In 1990, the OPs diazinon and propetamphos were found in each of the 20 catchments sampled by the Tweed River Purification Board, in Scotland, with concentration ranges of 14–124 ng l−1 for diazinon and 72–366 ng l−1 for propetamphos [1]. In Wales, a 1998 survey of 107 sites of water quality sampling revealed that the presence of sheep dip insecticides was widespread, with 75% of the sites scoring positive. Diazinon was present at 52% of the sites, whilst propetamphos and the SPs cypermethrin and flumethrin were found at 34%, 33%, and 6% of the sites, respectively [2]. In 2000, concentrations of sheep dip detected in surface freshwater in England and Wales were in the ranges of 1–11 738 000 ng l−1 for propetamphos, 1–85 100 ng l−1 for flumethrin [4]. Sewage treatment facilities receiving effluents from wool processing and livestock markets were also identified as potential point sources of sheep dip insecticides [2,5].

The effects of sheep dip disposal on animal slurries, soil and aquatic systems are still relatively unknown. A study of the indigenous microflora of animal slurries containing Bayticol, an SP-based sheep dip with 6% flumethrin, reported an increase in faecal coliforms, heterotrophic and protolytic bacteria as well as putative Salmonella sp. of up to four orders of magnitude compared to control slurries [8]. Two hypotheses were proposed for the observed increase in bacterial abundance. Firstly, the bacteria may be using the SP sheep dip as a nutrient source. Secondly, the SP sheep dip may be affecting the major bacterial predators in the system, i.e. protozoa. To investigate these hypotheses, this study aimed (i) to identify the effects of SP and OP sheep dip concentrations on the survival of 15 protozoan species, (ii) to test the recovery of the four species of amoebae, that is their ability to excyst, and (iii) to characterise bacterial survival and growth on solid media containing SP or OP sheep dip formulations.

2 Materials and methods

2.1 Protozoa and culture conditions

Protozoan cultures were used to test the toxicity of each of the sheep dip formulations. Table 1 lists the protozoan species tested and their culturing conditions (growth medium and incubation times) prior to use in experiments. Bacterial suspensions were used as food source. Amoebae were cultured on Escherichia coli K12, Tetrahymena pyriformis, Jakoba libera, Bodo saltans, Paraphysomonas vestita and Paraphysomonas imperforata were grown on B1, a Gram-negative rod-shaped bacterium, possibly Alteromonas [9]. The remaining species were grown on mixed bacterial assemblages, each isolated from the original protozoan culture.

View this table:
1

List of the protozoan species used in MIC tests and their culture conditions

Protozoan speciesSourceaMediumbIncubation
Time (days)Temperature (°C)
CiliatesColpoda inflataCCAP 1615/2SPL520
Cyclidium glaucomaCCAP 1616/1Ck515
Tetrahymena pyriformisCCAP 1630/1WCk320
Vorticella similisCCAP 1690/2Ck520
FlagellatesEntosyphon sulcatumCCAP 1220/1BCk1420
Jakoba liberaD.J. PattersonASW320
Bodo saltansCCAP 1907/2Ck315
Cercomonas sp.H.L.J. JonesDiatom520
Paraphysomonas imperforataB.S.C. LeadbeaterASW320
Paraphysomonas vestitaCCAP 935/14Ck520
Spumella elongataCCAP 955/1Ck515
AmoebaeValkampfia avaraCCAP 1588/1AAS420
Tetramitus rostratusCCAP 1581/1AS420
Hartmannella cantabrigiensisCCAP 1534/8AS420
Saccamoeba limaxCCAP 1257/3AS420
  • aCCAP, Culture Collection of Algae and Protozoa, CEH Windermere, Ambleside, Cumbria, UK.

  • bSPL, Sigma cereal leaf Prescott liquid; Ck, chalkleys; ASW, modified artificial seawater; AS, amoeba saline (Culture Collection of Algae and Protozoa, Catalogue of Strains, 1995, NERC, UK).

2.2 Sheep dip formulations

The sheep dip formulations chosen were an OP, Ectomort (Young's Animal Health, UK), containing 8% propetamphos as the active ingredient, and an SP, Bayticol (Bayer, Germany) containing 6% flumethrin as the active ingredient. The sheep dip formulations were purchased as ready-prepared solutions, and therefore comprised a mixture of unknown constituents in addition to the active ingredients.

2.3 Minimum inhibitory concentrations (MICs)

MICs were performed in triplicate wells in 96-well microtitre plates. Standard solutions of the OP and SP dips were prepared by mixing each of the formulation with sterile distilled water to achieve final well concentrations ranging from 0.00001% to 1% (dip formulation:total volume). Aliquots (8 μl) of each standard were added to each of the wells, followed by 192 μl of protozoan suspension. Controls consisted of protozoan suspensions (192 μl) to which 8 μl distilled water was added. The wells were incubated at 20°C and examined by phase contrast inverted microscopy after 24 h. A qualitative scale was used to assess the viability of the cultures and thus their sensitivity to the sheep dip formulations. No specific observations were made regarding changes in protozoan activity.

2.4 Survival of amoebic cysts following MIC testing

Following MIC testing, encysted amoeba species were washed and plated in order to assess their viability. An E. coli suspension (200 μl) was spread onto the surface of amoeba saline agar plates (AS broth plus 2% agar No. 2; Culture Collection of Algae and Protozoa, Catalogue of Strains, 1995, NERC, UK), in order to provide a food source for excysting amoebae. Wells containing encysted Saccamoeba limax, Valkampfia avara, Tetramitus rostratus and Hartmannella cantabrigiensis were pooled by species and sheep dip concentration in microcentrifuge tubes. AS broth was added to each tube, to a final volume of 1 ml, and centrifuged at 1600×g for 30 min. After the supernatant had been discarded, the cyst pellets were washed three times and resuspended in 150 μl AS broth. Triplicate aliquots (20 μl) of protozoan suspensions were subsequently spot-plated on duplicates of the AS agar plates, seeded with E. coli, and incubated at 20°C. After 15 and 30 days, plates were examined microscopically for the presence of trophozoites (active cells) and/or their trails on the bacterial lawn.

2.5 Survival and growth of environmental bacterial isolates

The survival of 13 environmental bacterial isolates was tested on sodium lauryl sulfate (SLS, Oxoid, UK) agar plates (SLS broth plus 1.5% agar No. 2) containing 0.01, 0.1 and 1% OP or SP sheep dip formulations (the bacteria had previously been isolated and maintained on SLS). The dip formulations were added unfiltered to sterile (autoclaved) medium, which had cooled to 45–55°C. The medium was mixed and immediately poured into Petri dishes. Positive controls consisted of SLS agar alone, whilst negative controls contained tetracycline (100 μg ml−1). Suspensions of E. coli 51407, 53443, 55007, 71007, 71403, 71407, 71507, 73407, 75403, Proteus vulgaris 20223, Klebsiella ozaenae 73410, Salmonella paratyphi A 51000 and Serratia liquefaciens 67110 were prepared in sterile distilled water and inoculated on triplicate plates with a multipoint inoculator (MAST laboratories, Bootle, UK). The plates were incubated at 37°C and checked for bacterial growth after 24 and 48 h.

The capacity of five bacterial strains to utilise the dip formulations for growth was assessed. All glassware was heated to 450°C for 48 h to remove carbon traces. E. coli 75403, P. vulgaris 20223, K. ozaenae 73410, S. liquefaciens 67110 and S. paratyphi A 51000 were grown on nutrient agar (Lab M, International Diagnostics Group, UK) for 48 h at 25°C. Bacterial cells were then washed 10 times in C-free M63 medium (KH2PO4 13.6 g l−1, (NH4)2SO4 2 g l−1, MgSO4.7H2O 0.277 g l−1, thiamine 0.01 g l−1, FeSO4·7H2O 0.0005 g l−1, plus 40 μg l−1 of 18 amino acids: arginine, histidine, lysine, proline, serine, alanine, threonine, isoleucine, valine, phenylalanine, glutamine, leucine, asparagine, tryptophan, glutamic acid, tyrosine, aspartic acid, cystine; medium adjusted to pH 7 with KOH), by centrifugation at 1500×g for 15 min and resuspension in fresh C-free M63. M63 agar plates (M63 broth plus 2% agar No. 2) were prepared in triplicate for the following treatments: C-free, glucose (2 mg ml−1), 0.01, 0.1 and 1% SP or OP dip formulations. Aliquots (10 μl) of the washed cell suspensions were spot-plated onto each treatment. Plates were incubated at 25°C for 14 days. Growth was assessed qualitatively relative to the glucose controls. Colonies from each bacterial species and each treatment were subsequently sampled using a sterile loop, streaked onto nutrient agar and incubated, as above, to assess viability.

3 Results and discussion

3.1 MICs for 15 species of protozoa exposed to SP or OP sheep dip formulations

MICs, or the concentrations of sheep dip formulations at which protozoa encysted, ceased to be motile or disappeared from the suspension altogether, varied with species and concentration of the formulations. The MICs for flagellates and amoebae in the presence of the OP were highly variable, ranging from 0.005 to 0.06%, whereas ciliates were more consistently affected, with a MIC range of 0.007–0.01% (Table 2). The MICs of protozoa in the presence of the SP were relatively consistent, ranging from 0.01 to 0.03% (Table 3).

View this table:
2

Viabilitya of 15 species of protozoa after 24-h exposure to a range of concentrations (%) of the OP formulation

SpeciesControl0.000010.000050.00010.00050.0010.0030.0050.0070.010.0150.030.060.11
C. inflata++++++++++++++++++++++++
C. glaucoma++++++++++++++++++++++
T. pyriformis++++++++++++++++++++++++++
V. similis++++++++++++++++++++++++
E. sulcatum++++++++++++++++++N/AbN/AN/AN/A
J. libera++++++++++++++++++N/AN/AN/AN/A
B. saltans+++++++++++++++++++++++++++++++
Cercomonas sp.+++++++++++++++++++++++++++++++
P. imperforata+++++++++++++++++++
P. vestita+++++++++++++++++++++++++++++++
S. elongata+++++++++++++++++++++++++++
V. avara+++++++++++++++++++++++CCCCCCC
T. rostratus+++++++++++++++++++CCCCCCCC
H. cantabrigiensis+++++++++++++++++++C+CCCCCCCC
S. limax++++++++++++++++++++++++++++++C+CCCC
  • aWells were evaluated relative to the controls. +++, Fully motile and active (controls); ++, motile, numbers depleted relative to controls; +, motile, number depleted by over 80% relative to controls; −, non-motile or no visible organisms; +C, mostly cysts although some trophozoites may be present in very small numbers; C, only cysts visible.

  • bNot assessed.

View this table:
3

Viabilitya of 15 species of protozoa after 24-h exposure to a range of concentrations (%) of the synthetic pyrethroid (SP) formulation

SpeciesControl0.000010.000050.00010.00050.0010.0030.0050.0070.010.0150.030.060.11
C. inflata++++++++++++++++++++++++++++
C. glaucoma+++++++++++++++++++++++++
T. pyriformis+++++++++++++++++++++++++++++++
V. similis++++++++++++++++++++++++++++
E. sulcatum++++++++++++++++++N/Ab++N/AN/AN/A
J. libera++++++++++++++++++N/A+++N/AN/AN/A
B. saltans+++++++++++++++++++++++++++++++
Cercomonas sp.+++++++++++++++++++++++++++
P. imperforata++++++++++++++++++++++++++
P. vestita+++++++++++++++++++++++++++++++
S. elongata+++++++++++++++++++++++++
V. avara++++++++++++++++++++++++++CCCCCC
T. rostratus+++++++++++++++++++++++++CCCCCC
H. cantabrigiensis+++++++++++++++++++++++++++CCCCCC
S. limax++++++++++++++++++++++++++++++++CCCC
  • aWells were evaluated relative to the controls. +++, Fully motile and active (controls); ++, motile, numbers depleted relative to controls; +, motile, number depleted by over 80% relative to controls; −, non-motile or no visible organisms; +C, mostly cysts although some trophozoites may be present in very small numbers; C, only cysts visible.

  • bNot assessed.

Overall, the SP formulation appeared to be less toxic to protozoa than the OP formulation (Tables 2 and 3). This was the case for all of the ciliates and amoebae tested, with the exception of S. limax, which seemed particularly resistant to the OP formulations with an MIC of 0.06%. Colpoda inflata, Vorticella similis, Entosyphon sulcatum, J. libera, P. imperforata, V. avara, T. rostratus and H. cantabrigiensis were twice as sensitive to the OP compared to the SP. Four flagellates, namely B. saltans, Cercomonas sp., P. vestita and Spumella elongata, did not follow this trend: B. saltans and P. vestita behaved similarly for both SP and OP with an MIC of 0.03%, whilst Cercomonas sp. was more resistant to OP with a MIC of 0.03% compared to 0.015%, when exposed to SP. However, it must be noted that B. saltans, Cercomonas sp. and P. vestita were all severely depleted at 0.015% OP. The protozoa that were least sensitive to the SP were T. pyriformis, B. saltans, P. vestita and S. limax, all with MICs of 0.03%. However, T. pyriformis, B. saltans and P. vestita were all severely depleted at 0.015% SP. Of all the protozoa tested, Cyclidium glaucoma, E. sulcatum, J. libera, P. imperforata, S. elongata and T. rostratus were more sensitive to SP (0.01%). E. sulcatum, J. libera, P. imperforata and T. rostratus were also more sensitive to OP (0.005%).

These results confirm the hypothesis that the presence of sheep dip formulations in slurries may reduce both protozoan numbers and diversity. Concentrations of sheep dip formulation in animal slurry may range between 0.02 and 0.04% (v/v), if spent sheep dips are diluted threefold, as advised by the Environment Agency [5]. This falls within the MICs of 11 of the 15 protozoa tested for both SP and OP. However, some degree of protection may be present in the animal slurry from the suspended organic colloids. Binding of the lipophilic active ingredients to organic matter is also likely to occur [10]. Tests are currently under way to assess sheep dip toxicity to autochthonous protozoan communities in slurry systems.

3.2 Survival of amoebic cysts following MIC testing

Cysts were present in all amoeboid suspensions (Tables 2 and 3), but became the dominant form at varying dip concentrations depending on species and formulation types. T. rostratus no longer showed trophozoites at 0.005% OP and 0.01% SP, and was followed by V. avara and H. cantabrigiensis at 0.007% OP and 0.015% SP. S. limax was most resistant to both formulations. Encysting may be considered a defence mechanism necessary for survival under unfavourable conditions resulting from direct toxicity of the sheep dip formulations and/or changes in environmental factors, e.g. changes in water chemistry due to the presence of the sheep dip in solution. The pH of the sheep dip suspensions did not change significantly at the boundaries of the MIC results (data not shown); neither encysting nor death could thus be attributed to changes in pH. Studies are therefore needed to determine the physiological processes by which the sheep dip formulations affect protozoan survival. The encysting process was reversible, and the cysts remained viable, even at the highest concentrations tested (Table 4). Following washing and plating on a readily available food source (E. coli), cysts were able to break their dormancy, as the presence of trophozoites on AS plates showed. Only V. avara did not consistently excyst. Cysts of various species have been shown to withstand extreme conditions, including desiccation and strong acid [11]. Species possessing resistant cysts are thus more likely to survive the presence of sheep dip formulations. In a study by Fernandez-Leborans and Novillo [12], the effects of cadmium on a freshwater protozoan community were investigated and it was shown that, although some pollutants can facilitate the rise of undesirable protozoan species, which adversely affect food chain dynamics, communities can recover. However, in a study by Petz and Foissner [13], lindane was shown to have a profound impact on the concentration and community structure of soil ciliates. Although our results point towards the potential for community recovery within the amoebic community, the effects of sheep dip formulations on whole protozoan community composition and dynamics are unknown.

View this table:
4

Presence of trophozoites on amoeba saline plates seeded with E. coli following MIC tests in OP and SP formulations

OP concentration (%)SP concentration (%)
00.0050.0070.010.0150.030.060.1100.0050.0070.010.0150.030.060.11
V. avara+n/aa++++++n/an/an/a++++
T. rostratus++++++++++n/an/a++++++
H. cantabrigiensis+n/a++++++++n/an/an/a+++++
S. limax+n/an/an/an/an/a++++n/an/an/an/a++++
  • aNot plated as trophozoites still visible in MIC suspension.

3.3 Survival and growth of bacterial environmental isolates

The presence of SP or OP formulations did not affect bacterial survival and growth on SLS medium containing 0.01, 0.1 and 1% sheep dip concentrations. All the species formed recognisable colonies at all treatments, including the positive controls (agar alone) within 48 h. Moreover, no colonies were visible on plates containing tetracycline (100 μg ml−1).

The growth of five test bacteria (E. coli 75403, P. vulgaris 20223, K. ozaenae 73410, S. liquefaciens 67110 and S. paratyphi A 51000) on the various M63 agar treatments, and thus their ability to use the sheep dip formulations as sole carbon sources, is shown in Table 5. After 14 days, all bacteria plated on glucose M63 exhibited large colonies over a thick (opaque) yellow growth. In addition, P. vulgaris 20223 swarmed over the surface of the agar plates. Cryptic growth occurred to some extent on the C-free plates with small white colonies forming on the dried (transparent) aliquot drop. Similarly, small colonies of P. vulgaris appeared at the inoculation sites, but did not extend over the plates; no swarming was evident. In contrast, growth on M63 containing dip formulations appeared inhibited, regardless of concentrations and chemical types. The aliquot areas did not change over the incubation period and remained a transparent white. White colonies were visible after 14 days, but were smaller than their counterparts growing on C-free plates. Bacteria appeared unable to utilise SP and OP formulations as sole carbon sources during the 14-day incubation. However, they remained viable and subsequently grew on nutrient agar within 48 h.

View this table:
5

Growth of five bacterial species on M63 agar containing no C source or three concentrations of OP and SP formulations, relative to the M63 glucose controls

GlucoseC-freeOP concentration (%)SP concentration (%)
0.010.110.010.11
E. coli 75403++++++++++++
P. vulgaris 20223++++++++++++
K. ozaenae 73410++++++++++++
S. paratyphi A 51000++++++++++++
S. liquefaciens 67110++++++++++++

The survival of bacterial isolates was unaffected by the presence of sheep dip, but these organisms were not able to use either of the dip formulations as carbon source. Maloney et al. [14] found that pure cultures of Bacillus cereus, Pseudomonas fluorescens and Achromobacter sp. were able to use five pyrethroid insecticides only during cometabolism with Tween 80. The development of cometabolism between the various compounds present in the dip formulations, including the active ingredients, may be possible. However, bacteria may have been unable to access nutrients in sufficient amount from the agar plate for such cometabolism to occur. Similarly, the potential for development of catabolic activity after long or repeated exposures to the chemicals, and the use of nutrients contained in the slurries, cannot be excluded.

3.4 Implication for animal slurry, soil and freshwater systems

Run-off from slurry-applied fields is known to decrease water quality, through additions of high levels of nutrients, organic matter, particulate, and biological contamination [7,15,16]. The presence of sheep dip in the slurry is likely to exacerbate this problem. The relationship between some populations of protozoa and bacteria is of a predator/prey nature, with protozoa in turn stimulating and regulating bacterial populations [17]. Sheep dip formulations present in animal slurries may affect this trophic interaction, thus leading to an increase in bacterial populations. Similar feedbacks were reported by Day et al. [18] and Thirup et al. [19]. In the former [18], the presence of fenvalerate, an SP insecticide, caused a reduction in numbers of cladocerans (filter-feeding zooplankton), thereby inducing a significant increase in phytoplankton. In Thirup et al. [19], the inhibition of protozoa by Corbel®, a formulation containing the fungicide fenpropimorph, resulted in an increase in early-colony-forming bacteria. In animal slurries, bacterial populations may grow unchecked, as the exponential increase of faecal coliforms and faecal streptococcus populations in slurries containing fresh SP dip has shown [8]. The scale of the contamination is dependent upon microbial number and survival in the soil and water environments. The presence of sheep dip formulations potentially impacts on both these factors by depleting protozoan populations. Thus, Davies et al. [20] showed that faecal coliforms were able to grow in freshwater sediments in the absence of protozoa. Conversely, the presence of protozoa resulted in a net decrease in bacterial numbers. Indeed, studies have shown that protozoan predation is one of the main factors controlling bacterial density and the removal of allochthonous bacteria from aquatic systems [21,22]. Although coliform numbers may increase in the slurries, their populations may be regulated by soil and freshwater protozoa following spreading and run-off. However, this is a contentious issue, as sheep dipformulations present in the slurry may affect soil protozoa and active ingredients of the sheep dip formulations have been detected in surface freshwater of England and Wales [4].

It must be noted that our results cannot be directly extrapolated to slurry systems. In Semple et al. [8], a concentration of 0.02% SP in animal slurry showed no effect on bacterial populations, despite being at or above the MICs of most protozoa tested in this study. At 1% SP, however, the dramatic increase in faecal coliform numbers correlates with the significant protozoan mortality found here. Nevertheless, this study emphasises the need to include surveys of microbial and protozoan communities within slurries, soil and aquatic systems, in ecological risk assessment aimed at evaluating the effects of sheep dip disposal onto land.

Acknowledgments

The authors would like to thank Joanna English for providing the bacterial isolates. The authors would also like to thank Janice Drinkall, Karen Heaton and Joanna English for the help and technical advice they kindly provided throughout this study.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
View Abstract