Remote collection of animal DNA and its applications in conservation management and understanding the population biology of rare and cryptic species

Please download to get full document.

View again

of 13
11 views
PDF
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
Remote collection of animal DNA and its applications in conservation management and understanding the population biology of rare and cryptic species
Document Share
Document Tags
Document Transcript
  CSIRO  PUBLISHING © CSIRO 200310.1071/WR020771035-3712/03/010001www.publish.csiro.au/journals/wr   Wildlife Research , 2003, 30 , 1–13 Remote collection of animal DNA and its applications in conservationmanagement and understanding the population biology of rareand cryptic species  Maxine    P  .  Piggott  A   and     Andrea   C  . Taylor  A,B A School of Biological Sciences, Monash University, Victoria, 3800, Australia B To whom correspondence should be addressed. Email: Andrea.Taylor@sci.monash.edu.au  Abstract. Obtaining useful information about elusive or endangered species can be logistically difficult, particularly if relying entirely on field signs such as hair, feathers or faeces. However, recent developments inmolecular technology add substantially to the utility of such ‘non-invasive’ samples, which provide a source of DNAthat can be used to identify not only species but also individuals and their gender. This provides great potential toimprove the accuracy of abundance estimates and determine behavioural parameters, such as home-range size,individual habitat and dietary preferences, and even some forms of social interaction. Non-invasive samples can also be a useful alternative to blood or tissue samples (the collection of which traditionally has required trapping of animals) as genetic material for applications such as relatedness, population genetic and phylogenetic analyses.Despite the huge potential of non-invasive genetic sampling, the current technology does have limitations. The lowquantity and quality of DNA often obtained from such sources results in an increased risk of genotyping errors,which may lead to incorrect inferences, particularly false identification of individuals. Appropriate precautions and  pilot studies are required to minimise these risks, and in some cases it may be wise to employ traditional methodswhen they are adequate. WR02077 Non-invasive DNA samplingM. P.Piggottand A.C.Taylor Introduction Wildlife field study methods are many and varied, but their utility is limited in applications involving species that arerare, endangered or cryptic. Determining the abundance of such species with any degree of accuracy can be difficult,although such information is vital for developingmanagement strategies. In addition, in-depth studies toexamine the population biology of these species via geneticmethods may be logistically impossible, particularly if dependent on opportunistic sampling (Kohn and Wayne1997). For rare and sensitive species, invasive study methodssuch as trapping (which may also be accompanied by tissueor blood sampling) and fitting of radio-transmitters may beneither feasible nor appropriate (Morin and Woodruff 1996;Taberlet et    al.  1999). This is particularly so if the species of interest is difficult to trap, or the risk of injury or death is toohigh for animals already under pressure from predators,disease, shrinking habitat and other factors (Greenwood 1996). In any case, invasive study methods may unavoidablyalter the dynamics of a population in ways that are difficultto predict (Harrison et    al.  1991), and possibly change a behaviour or characteristic being studied (Morin et    al. 1994 b ; Morell 1995). Field collection of samples such as hair or faeces can provide some useful information, such as presence/absenceand possibly a rough estimation of abundance. However,unequivocal verification that such samples came from thespecies of interest may not be possible on the basis of morphology alone. By utilising non-invasively collected samples as sources of DNA for molecular genetic marker analysis, many of these shortcomings can be surmounted. It is useful to distinguish between two types of situation inwhich collection and use of non-invasive samples for geneticanalysis may be desirable. The first is when capture or observation of animals is impossible or inefficient, yetmorphological analysis of remotely collected samples suchas hair and scats provides only limited information (Kohnand Wayne 1997). Appropriate DNA analyses can greatlyextend the utility of such samples by unequivocallydetermining the species, identity and gender of theindividuals from which they came (Kohn and Wayne 1997;Taberlet et    al.  1999). In the case of faecal samples,identification of plants and animals that make up their dietmay be enhanced (Hoss et    al.  1992; Hofreiter et    al.  2000;Symondson 2002). The second major use for non-invasive samples is as analternative to blood or tissue samples (also referred to asnon-destructive sampling, see Woodruff 2003) taken fromtrapped animals, as a source of DNA for relatedness, population genetic and phylogenetic analyses (Hoss et    al.  2  Wildlife Research M. P. Piggott and A. C. Taylor    1992; Kohn et    al.  1995; Fernando et    al.  2000). While wedraw attention to some studies in which non-invasivesamples have been used for these purposes becauseconventional sampling was not feasible, a summary of theimportant contributions made by genetic analyses in thestudy of animal biology is beyond the scope of this review(instead, see Sunnucks 2000). Although there is intense interest in non-invasive geneticsampling, and numerous publications on the great potentialof this new technology, only a relatively small number of comprehensive studies on wild populations have used non-invasive sampling (e.g. Fernando et    al.  2000; Garnier et    al. 2001; Vigilant et    al.  2001; Utami et    al.  2002). This number is likely to increase, as more researchers become aware of thevast potential for molecular analysis in gleaning useful biological data from non-invasive samples. However, thetechnology currently has significant drawbacks that meanmore traditional methods should perhaps be employed whenthey are adequate. In this review we aim to present anoverview of the tempting range of potential applications for this technology, while warning of their limitations. Weexplain how these methods work and discuss how they mayimprove on traditional approaches to the same questions and how they have been put to good use in wildlife-related  projects to date. There is no doubt this technology is beginning to revolutionise many areas of the study of elusiveand endangered species. The basis of non-invasive genetic typing  How   it    works DNA sequencing of appropriate genes can unequivocallydetermine the species from which a non-invasive sample wasobtained. However, recent developments in molecular technology extend the usefulness of such samples far beyond determination of a species’ presence and rough estimation of its abundance (Taberlet et    al.  1999; Sunnucks 2000). Thegreatest contribution to this breakthrough was thedevelopment of the polymerase chain reaction (PCR) (Saiki et    al.  1985; Arnheim et    al.  1990), a method of enzymaticallyamplifying informative DNA sequences using short pieces of DNA that act as ‘primers’ for DNA extension. Coupled withDNA sequencing and its surrogates (such as analysis of single-stranded conformation polymorphisms, or SSCP:Sunnucks 2000) this enables the detection of species-, population-, individual-specific and sex-specific DNA signa-tures from a sample. Even very degraded and low-quantitytarget DNA, such as that recoverable from many kinds of non-invasively collected samples, can act as PCR template. The second crucial development, from the point of viewof individual identification and other applications requiringhigh resolution, was the discovery of DNA sequencescomprising short repetitive arrays embedded in uniquesequences (Tautz 1989). These markers, called microsatellites, exhibit a high degree of variability within populations and can provide individuals with unique DNA profiles when a number of these markers are used incombination. Although development of microsatellite primers is time-consuming they can be used in other closely-related species (usually within the same family) so that the process does not have to be repeated for every species (e.g.Coote and Bruford 1996; Engel et    al.  1996; Primmer et    al. 1996; Slate et    al.  1998; Zhang et    al.  2001; Zenger et    al. 2002). Importantly, because any given set of microsatellitePCR primers typically only amplify DNA from closelyrelated species, contaminating DNA (either from bacteria or dietary components in the case of faeces) is unlikely tointerfere with interpretation. The availability of DNA sequence information from thesex chromosomes of mammals and other vertebratesfacilitates gender identification of an animal from a non-invasive sample. For example, genetic sex identification inmammalian samples can proceed via PCRs employing primers specific for both Y-chromosomal DNA and anautosomal or X-linked marker (Griffiths and Tiwari 1993;Kohn and Wayne 1997; Sloane et    al.  2000). Samplesyielding products from both markers are deemed to be maleand those with only one, female. Alternatively, the presenceof a length polymorphism in the amelogenin gene, which hashomologues on both the X and Y chromosomes, can beexploited to distinguish between the sexes, as has beendemonstrated in great apes (Bradley et    al.  2001) and black  bears (Yamamoto et    al.  2002). Although sex chromosomeDNA has been sequenced in only a small number of animalspecies, its apparent high level of conservation has enabled the design of ‘universal’ primers for sex identification in agreat diversity of animals (e.g. mammals: Aasen and Medrano 1990; Griffiths and Tiwari 1993; cetaceans:Berube and Palsboll 1996; birds: Griffiths et    al.  1998;marsupials: Watson et    al.  1998). The ability to identify individuals and their sex from non-invasively collected samples offers unprecedented potentialto improve the accuracy of abundance estimates and determine behavioural parameters for individuals, such astheir home-range size, habitat and dietary preferences, and even some forms of social interaction. Applications for thistechnology are theoretically restricted only by the ingenuityof the researcher. For example, because both nuclear and mitochondrial DNA (mtDNA) was amplified successfullyfrom rodent and insectivore skulls found in regurgitated barnand tawny owl pellets, Taberlet and Fumagalli (1996)suggested that such samples might be a good source of material for studying population genetics of these smallanimals. Of course, in many cases a traditional trappingapproach may be more efficient for straightforward  population genetic analyses, but as Symondson (2002) points out, molecular analysis of owl prey would revealselective predation on certain genotypes or sexes.   Non-invasive DNA sampling  Wildlife Research 3  Potential     DNA    sources   Non-invasive sources of DNA that have proven useful to dateinclude shed or plucked hairs from various primates,marmots, wombats and bears (Morin et    al.  1994 b ; Taberlet et al.  1997; Field et    al.  1998; Goossens et    al.  1998 a ; Woods et al.  1999; Sloane et    al.  2000; Constable et    al.  2001; Banks et al.  2002 b , 2003). DNA analysis of plucked hairs has beensuccessful in providing a variety of data on wombats. Their  burrowing nature makes hair collecting relatively easy by the placement of double-sided tape across the burrow entrance,removing hairs from animals as they move in and out of their  burrows (Sloane et    al.  2000; Banks et    al.  2002 b , 2003). Hair samples have been collected from Capuchin monkeys( Cebus   olivaceus ) by shooting a tape-covered syringe froman air-powered dart pistol (Valderrama et    al.  1999), and fromfree-ranging black bears ( Ursus   americanus ) and brown bears ( U  . arctos ) by means of hair traps consisting of barbed wire attached to a tree encircling a scent lure (Woods et    al. 1999). Methods for other species include wrapping bait toforce animals to handle it and thus leave a hair sample, and making a corral of tape so animals squeeze their bodies between tape rails to reach bait (Valderrama et    al.  1999).These novel methods can be adapted to target specificindividuals, and will ‘pluck’ fresh hair, which has higher-quality DNA than shed hair (Valderrama et    al.  1999; Morin et    al.  2001). Another source of DNA is epithelial cells shed from theintestinal lining and deposited in, and on the surface of,faeces (Hoss et    al.  1992). Such DNA has successfully beenanalysed from a variety of animals including primates(Constable et    al.  1995; Gerloff et    al.  1999; Utami et    al. 2002), mountain lions (Ernest et    al.  2000), coyotes (Kohn et al.  1999), bears (Taberlet et    al.  1997), ungulates (Flagstad et al.  1999), dolphins (Parsons et    al.  1999), bats (Vege and McCracken 2001), common wombats (Banks et    al.  2002 a )and black rhinos (Garnier et    al.  2001). More unusual sourcesof DNA have been wolf urine in snow (Valiere and Taberlet2000), chimpanzee buccal cells from chewed food remnants(wadges) (Sugiyama et    al.  1993; Takenaka et    al.  1993;Hashimoto et    al.  1996; Morin and Woodruff 1996), sloughed skin from cetaceans (Bricker et    al.  1996; Valsecchi et    al. 1998) and for birds, nest materials, feathers, eggshells and urine (Morin et    al.  1994 a ; Pearce et    al.  1997; Nota and Takenaka 1999). However, most comprehensive studies haverelied on DNA extracted from either hair or faeces, which inmany cases is abundant and relatively simple to collect. Molecular exploitation of ‘field signs’ for species identification Scats, hairs or feathers collected in the field are traditionallysubjected to a variety of morphological analyses in order todetermine which species they are from (Putman 1984).However, there are situations in which such samples may not be reliably identified to species level on the basis of morphology alone (Bulinski and McArthur 2000). Hair morphology, in particular, is often indistinguishable or  problematic between closely related taxa (Brunner and Coman 1974; Friend 1978; Valente and Woolley 1982;Taylor 1985). For example, there is no current macroscopictechnique that can reliably distinguish black from brown bear hairs (Woods et    al.  1999). Even highly skilled specialists canmisidentify species. In a blind test carried out to determinethe accuracy of results from microscopic examination of hairs from 37 mammal species occurring in south-easternAustralia the accuracy and consistency of speciesidentification varied considerably among taxa (Lobert et    al. 2001): 19 species were reliably identified but the remaining18 (including Sminthopsis  spp.,  Antechinus  spp.,  Petaurus spp., Trichosurus  spp., Gymnobelideus   leadbeateri  and   Rattus   rattus ) were subject to some degree of identificationerror (Lobert et    al.  2001). Most errors were due to intra-taxon variation in hair characteristics. Misidentification of species from scats is probablycommon. It has been estimated that faeces are assigned to thecorrect species in only 50–66% of North American cases(Halfpenny and Biesot 1986). In Australia there are manyopportunities to confuse the faeces of sympatric macropod species, such as those of eastern grey kangaroos (  Macropus giganteus ) with red-necked wallabies (  M  . rufogriseus )(Johnson and Jarman 1987) and those of Bennett’s wallabies(  M  . rufogriseus ) with red-necked pademelons ( Thylogalebillardierii ) (Bulinski and McArthur 2000). A recent studyto determine the presence of quokkas ( Setonix   brachyurus ) atsites in Western Australia has utilised mtDNA analysis todistinguish quokka scats from those of other macropodsknown to be present (Alacs et    al.  2003). Determining the presence and abundance of endangered species and carnivores from morphological analysis of field signs is particularly difficult (Palomares et    al.  2002).Abundance estimates and feeding ecology of the endangered  pine marten (  Martes   martes ) is traditionally carried out bymorphological identification of faeces in the field by expertsurveyors (Davison et    al.  2002). However, DNA analysis of morphologically identified pine marten faeces showed thatthe surveyors often failed to distinguish them from those of red foxes ( Vulpes   vulpes ) and failed completely when pinemartens were at very low densities (Davison et    al.  2002). Amulti-evidence approach incorporating DNA analysis of faeces is recommended for the management of this species,which may be extinct in England and Wales (Davison et    al. 2002). A similar problem is experienced with the endangered Iberian lynx (  Lynx    pardinus ), as faeces may be confused withthose of foxes as well as of wild and domestic cats (  Felis silvestris  and  F  . catus ) and domestic dogs ( Canis    familiaris )(Palomares et    al.  2002). The endangered San Joaquin kit fox( Vulpes   macrotis   mutica ) is sympatrically distributed withfour other canids with similar scat morphology, causing  4  Wildlife Research M. P. Piggott and A. C. Taylor    identification problems, particularly at low populationdensities (Paxinos et    al.  1997). By contrast, mtDNA analysisallowed unequivocal species identification from field-collected scats, in surveys for both Iberian lynx and SanJoaquin kit foxes (Paxinos et    al.  1997; Palomares et    al. 2002). Identification by molecular means can thus greatlyenhance the utility of non-invasive samples for indicating presence or absence of a species. Indeed, as molecular technology becomes cheaper and more routine it may be themethod of choice even in situations where traditionalmorphological analysis is definitive for species identi-fication, particularly as the latter is typically dependent onthe availability of experienced practitioners. Abundance estimates aided by individual identification of ‘field signs’  Limitations   of    traditional    methods Traditional methods of estimating animal abundance are based on direct observational counts of individuals (either free-ranging or following capture), or on indirect signs suchas footprints and faeces (e.g. Grigione et    al.  1999). Whiledirect approaches are effective for many animals, they areinadequate for species that are elusive and/or difficult to trap,and for endangered species for which such methods may betoo disruptive. For example, trapping of the highlyendangered northern hairy-nosed wombat (  Lasiorhinuskrefftii ) affects both its health and behaviour. Analysis of trapping records showed that wombats lost an average of 0.5kg between first and second captures separated by intervalsof up to six months (Hoyle et    al.  1995). In addition, areastrapped twice in succession had lower population-sizeestimates for the second trapping period, suggesting thatanimals may have temporarily left areas disturbed bytrapping (Hoyle et    al.  1995). Other species, such as tammar and parma wallabies, can exhibit strong trap avoidance(Vujcich 1979) and repeated trapping drives in an area mayincrease trap wariness (Lentle et    al.  1997). Animals withlarge home ranges and mobility, such as large carnivores and elephants, are difficult to observe or capture (Grigione et    al. 1999; Kohn et    al.  1999; Woods et    al.  1999; Ernest et    al. 2000). In any case, the latter may pose unacceptable safetyrisks to both humans and animals (e.g. mountain lions:McCrown et    al.  1990; bears: Woods et    al.  1999; elephants:Eggert et    al.  2002). There are also a variety of analytical disadvantages totrapping-based abundance estimation. One is that mosttrapping techniques are unable to provide ‘snap-shot’estimates of population size for many species because theyrequire many months or even years to obtain sufficientsample sizes (Kohn et    al.  1999). Deceased and migratingindividuals may thus be mistakenly included in such population estimates, resulting in overestimates (Kohn et    al. 1999). On the other hand, poor trapping success may lead tounderestimation of population size. For example, non-invasive genotyping of coyote faeces showed that more thantwo-thirds of the current population may have been missed  by long-term ecological surveys, perhaps due in part to a lowoverall trapping efficiency of only one animal per 58 trap-nights (Kohn et    al.  1999). Population estimation of red foxes( V  . vulpes ) in Australia is also hampered by low trappingefficiency, and trapped samples may be strongly male- biased, with important ramifications for estimating population parameters (Kay et    al.  2000). Similarly, trappingdata for northern hairy-nosed wombats indicates an excess of males, which in the absence of other evidence, might beinterpreted as trapping bias because females are known to beharder to recapture and are more mobile (Alan Horsup,Queensland Parks and Wildlife Service, unpublished).However, a recent census using microsatellite analysis of remotely collected hair samples confirmed the male bias(Banks et    al.  2003). Finally, misidentification can occur using any tagging system, due to lost ear tags, distorted tattoos, changes in appearance, or lost or malfunctioningradio-transmitters (Woods et    al.  1999), and any of these mayimpinge on the accuracy of abundance estimates.A frequently used indirect   method of estimating animalabundance is to count faecal pellets. This is of most use if fixed plots are employed to examine trends in abundance(Jarman and Capararo 1997; Bulinski and McArthur 2000).Relying on scat counts alone may lead to importantoverestimates in population size, as exemplified by theestimate that a small colony of brush-tailed rock-wallabies(  Petrogale    penicillata    penicillata ) was surviving in theGrampians, Victoria, when, in fact, only a single animalremained (J. Reside, Victorian Brush-tailed Rock-wallabyRecovery Group, unpublished). Examination of footprintsleft in sand plots is employed for some elusive species (e.g.mountain lions: Smallwood and Fitzhugh 1995). Footprintsare only marginally informative, as it is rare to observe a perfectly formed one, and usually very difficult todistinguish between the tracks of closely related species(Triggs 1992). In any case, methods such as these onlyindicate population trends, as extrapolating absoluteabundance from them is problematic (Jarman and Capararo1997; Bulinski and McArthur 2000). Ultimately, traditionalanalyses of field signs provide little information on absolute population size and understanding of the demographic, behavioural and life-history strategies of individuals and  populations. What    can   molecular     genetic   analysis   add?  New molecular methods can help to overcome some of theabove limitations by providing accurate identification of remotely collected samples to both species and individuallevel, allowing direct and relatively unbiased enumeration(Kohn and Wayne 1997; Taberlet et    al.  1999). At the very   Non-invasive DNA sampling  Wildlife Research 5 least, this approach can provide an estimate of the minimumnumber of animals of a particular species in a given area. For example, in one of the first studies of this type, extensivenon-destructive (skin biopsies) and non-invasive (sloughed skin) sampling across the North Atlantic Ocean detected 7698 humpback whales and found an unexpected male sex bias in the population (Palsboll et    al.  1997). Microsatellite profiling and sexing of faeces (in combination with footprintmeasurements to estimate age) collected from the dwindlingPyrenean brown bear ( Ursus   arctos ) population gave anestimated population size of at least one yearling (male),three adult males and one adult female (Taberlet et    al.  1997).A combination of faecal DNA typing and trapping provided a minimum estimate of 16 mountain lions (seven by captureand nine by faecal DNA analysis) living in, or travellingthrough, the Yosemite Valley during an 18-month period (Ernest et    al.  2000).Collection of non-invasive samples followed bymicrosatellite genotyping is highly amenable to samplingdesigns appropriate for mark–recapture estimates of  population size, as applied to trap-based estimates. Thisapproach was used to estimate the abundance of coyotes fromfaeces collected from transects in the Santa MonicaMountains of California (Kohn et    al.  1999). Abundance of Canadian black bears and grizzly bears over a 64 ×  64-kmgrid was estimated using the hair-trapping method described earlier, followed by microsatellite analysis. More than 1750hair samples were collected and 1496 of them determined tospecies level using mtDNA analysis (Woods et    al.  1999). Of the 54 brown bears that contributed 303 hair samples, only 12had previously been captured and radio-collared. This wasthus a very effective and efficient way of censusing a speciesthat is difficult to observe, exists at low densities and haslarge home ranges (Mace et    al.  1994; Mace and Waller 1997).Applications involving individual identification have also been successful using both hairs and faeces of wombats,animals that are particularly difficult to enumerate bytraditional means (McIlroy 1977; Taylor et    al.  1998). Thesize of a population of common wombats ( Vombatus   ursinus )in suburban Melbourne parkland was estimated with verynarrow confidence limits from faecal DNA (Banks et    al. 2002 a ). In a similar analysis, but based on DNA profiling of hairs collected on double-sided tape at burrow entrances, thesole remaining population of the highly endangered northernhairy-nosed wombat was estimated to contain 113individuals (Banks et    al.  2003). This estimate substantiallyexceeded earlier (trap-based) ones, which, in combinationwith other indicators, suggests that the population mayrecently have increased in size following several years of favourable climatic conditions (Banks et    al.  2003). Other    applications   requiring    individual    identification Individual identification by non-invasive genotyping canalso be useful for tracking particular individuals in the wild and identifying dispersal events. For example, DNA profilesof captive-raised chimpanzees recorded prior to their releaseinto a native population will facilitate long-term tracking of these individuals (and their offspring), and thus assist ingauging the success of such translocation programmes(Goossens et    al.  2002). Analysis of 692 ‘recaptures’ of  Northern Atlantic humpback whales detected from non-invasive samples confirmed previous assumptions (based onidentification and tracking of natural markings) regardingindividual and migratory patterns, site-fidelity to summer feeding grounds and mixing in winter breeding grounds(Palsboll et    al.  1997). Dietary analysis The definitive species identification offered by molecular technology, in combination with traditional dietary analysisof scats has made it possible to elucidate the diets of a varietyof species whose faeces are not readily distinguished morphologically. For example, this approach allowed Reed et al.  (1997) to determine the relative impacts of grey(  Halichoerus    grypus ) and harbour (  Phoca   vitulina ) seals onfisheries, and Hansen and Jacobsen (1999) to better interpretthe feeding biology of mink (  Mustela   vison ), otters (  Lutralutra ) and polecats (  Mustela    putorius ). In another example,assignment of field-collected scats to each of four sympatricVenezuelan carnivore species using mtDNA analysis,indicated that scat size overlapped considerably amongspecies. This produced a much-altered profile of their dietary niches, which had previously been interpreted on the basis of erroneous assumptions about scat size (Farrell et    al. 2000).Another possible application of faecal DNA technology isin the identification of plants and animals consumed by thespecies of interest, although conventional methods may be better for analysing their actual abundance in the diet (Kohnand Wayne 1997). Nonetheless, in cases where foods arethoroughly digested or difficult to identify, including wherehard parts of similar species are difficult to distinguish,molecular identification will be invaluable. Analysis of bear droppings by DNA sequencing of amplification productsfrom the chloroplast rbc L gene identified the presence of DNA from the genus  Photinia  (Hoss et    al.  1992). Althoughexact identification to species level may have been possible by analysis of a more informative gene, the researchersinferred that the plant was most likely the common  P  . villosa ,and that its berries formed a significant component of the bears’ diet during late summer when the scats were collected (Hoss et    al.  1992). Molecular analysis of scats has even beenused to infer diet in a long-extinct sloth, whose faeces werediscovered in a cave in Arizona (Hofreiter et    al.  2000), and from an 11 700-year-old rodent midden (Kuch et    al.  2002). Molecular identification of prey remains in bird faeces isan alternative to invasive techniques and also a potentiallymore accurate approach (Sutherland 2000; Casement 2001).
Similar documents
View more...
Search Related
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks