Developing a species identification system in a forensic context for Diprotodontia Linzi Wilson-Wilde1,3,4, Janette Norman3, James Robertson2, Stephen Sarre1 and Arthur Georges1

Introduction Wildlife crime is a deliberate and purposeful illegal activity involving animals and plants, alive or dead (including their parts, products and derivatives) for which purposeful gain is the principle motive (McDowell 1997). The Australian marsupial order Diprotodontia which includes the kangaroos, possums and gliders are purportedly targets of wildlife crime, being illegally exported overseas as part of the pet trade or for collections, as well as illegally hunted for sport, meat or skins. A number of Diprotodontid species are linked in the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) appendices I and II listed and are therefore subject to strict exportation/importation restrictions. Illegal hunting of macropods, possums and koalas has been the subject of the Australian Senate Select Committee on the Commercial Utilization of Native Australian Wildlife (Anon 1998). Fundamental to the regulation of trade and harvesting of these species is an appropriate identification method.

Using the current classification as developed by Meredith et al. 2009 there are currently 11 extant families containing approximately 125 living species (Table 1). Diprotodontids derive their name from the two procumbent lower medial incisors, which are a synapomorphy of the order (Kirsch 1977, Aplin and Archer 1987). However there continues to be much debate regarding the classification and phylogeny of the Diprotodontia and many species are known only from the fossil record. Mitochondrial DNA research has been focussed on the ND2 gene as detailed in Osborne and Christidis 2002. However this work was mainly centred on Petauroidea with limited macropod species and therefore did not provide a robust phylogeny of the entire order.

Wilson and Reader 2005 Meredith et al. 2009 Meredith et al. 2009 produced an extensive phylogenetic analysis of the Diprotodonts using five nuclear genes: ApoB, BRCA1, IRBP, Rag1 and vWF and all genera. However Meredith et al. found a basal split into two suborders between Vombatiforms and Phalangerida, Order Diprotodontia Order Diprotodontia which is inconsistent with the last major phylogenetic analysis of Diprotontids of Wilson and Reader 2005 who found three monophyletic Suborder Vombatiformes Suborder Vombatiformes suborders (Vombatiforms, Phalangeriforms and Macropodiforms) (Table 1). Meredith et al. noted that Phalanderoidea grouped with Family Phascolarctidae Family Phascolarctidae Macropodiformes and this group was a sister group to Petauroidea. Family Vombatidae Family Vombatidae Suborder Phalangeriformes Suborder Phalangerida A key issue in wildlife forensics is defining the biological (taxanomic) units of concern, whether they are species or populations, or Superfamily Phalangeroidea Superfamily Petauroidea some other variant. A species may not always be easily defined by genetic means, particularly if the baseline knowledge of genetic Family Phalangeridae Family Petauridae variation within and among species is not well characterised. The DNA analysis of two closely related populations may result in most Family Burramyidae Family Pseudocheiridae of the specimens being grouped into one of the two populations, however there will always be outliers in the ‘grey area’ between the Superfamily Petauroidea Family Acrobatidae two populations that cannot be readily assigned to one group or the other (Mallet 1995). Family Petauridae Family Tarsipedidae Family Pseudocheiridae Superfamily Australoplagiaulacoida Species identification processes are underpinned by robust phylogenetic information regarding the relevant species. Traditionally Family Acrobatidae Clade Phalangeroidea Family Tarsipedidae Family Phalangeridae species identification and phylogenetic reconstruction is performed using morphological (traditional taxonomy) or mitochondrial DNA Suborder Macropodiformes Family Burramyidae data. While each source of data has its strengths, they can produce conflicting results (Hillis 1987, Shaffer et al. 1991). Morphological Family Potoroidae Clade Macropodiformes data integrates genomic differences but is subject to problems of moderation by the environment in which genes are expressed Family Hypsiprymnodontidae Family Potoroidae (phenotypic plasticity) and a correlated evolution of functionally linked traits (Georges et al. 1998). Family Macropodidae Family Hypsiprymnodontidae Family Macropodidae Mitochondrial DNA markers (mtDNA) are from a single maternally inherited unit as such may mislead through independent linage sorting and undetected instances of hybridisation and introgression (Brower 2006) and these factors may affect the ability of the Table 1: Recent Diprotodontid Taxonomic Classifications of Wilson and Reader 2005 and mitochondrial marker to correctly attribute the sample to the correct species. This study combines mitochondrial and nuclear marker Meredith et al. 2009 . information to create a more robust phylogeny to underpin a species identification method.

GAPD only displayed strong bootstrap values for the groupings of Pseudocheirus occidentalis and Pseudocheirus Materials and Methods peregrinus and Phalangeridae and as such, displayed limited phylogenetic resolution. The phylogenetic performance of COI appeared lower than the other markers. The macropods were split into two clades and the 55 Samples were obtained from vouchered and non vouchered specimens, two sub orders were not resolved. Also Macropus rufus did not group with the other macropods (Figure 4). While usually tissue (heart, liver), from various collections representing 10 families, ND2 did not resolve at the deeper nodes (it also did not resolve the two sub orders), it did group the shallow 27 genera and 62 species of Diprotodontia, plus an additional 2 species nodes as expected (Figure 5). used as an outgroup.

Taxonomic designations follow Meredith et al. (2009). All tissue samples were extracted by either the standard phenol chloroform procedure or the Qiagen DNeasy® Blood and Tissue Kit (Qiagen part # 69506), using the animal tissue bench protocol. Negative controls were used in all extractions.

Standard PCR reactions were performed for three nuclear markers: APOB, IRBP and GAPD. Mitochondrial markers targeted were COI and ND2. PCR products were visualised on a 1.2% agarose gel. All sequencing was performed in both directions by the Australian Genome Research Facility using M13 primers.

Run data was edited using SEQUENCHER version 4.1.4 (GeneCodes Corporation 1991-2002). Consensus sequences were aligned using the CLUSTAL module in MEGA version 4 with manual adjustments. Maximum parsimony analysis was performed using PAUP 4.0b10 (Swofford 2002) where gaps were coded as missing data.

Parsimony analysis utilised the heuristic search option and TBR branch swapping and employed bootstrap analysis incorporating 1,000 replications. Maximum likelihood analysis of concatenated data was performed using PHYML (Guindon and Gascuel 2003) on the Montpellier bioinformatics platform (http://www.atgc-montpellier.fr/phyml).

Results and Discussion After removing ambiguous data the following sequences were obtained for each marker: APOB 59 taxon, 724 basepairs, IRBP 56 taxon, 560 basepairs, GAPD 52 taxon, 478 basepairs, COI 61 taxon, 699 basepairs, ND2 63 taxon, 871 basepairs. The concatenated data set contained 64 taxon and 3332 basepairs.

Maximum parsimony analysis of individual markers displayed similar Figure 4: Phylogenetic analysis of Diprotodonts Figure 5: Phylogenetic analysis of Diprotodonts topologies. All trees were condensed to display only those nodes with more for the COI mitochondrial marker. Figures for the ND2 mitochondrial marker. Figures than 50% bootstrap support and manually rooted in Isoodon (apart from represent node support for maximum parsimony represent node support for maximum parsimony analysis based on 1,000 replications. analysis based on 1,000 replications. NO2 and the concatenated data set). All markers, except for COI were able to resolve Macropodiformes, however IRBP grouped Petaraus breviceps within this Clade. All markers were also able to resolve Phalangeridae with The likelihood analysis of the concatenated data combing the nuclear and mitochondrial markers displayed strong bootstrap supports. APOB performed well as a diagnostic marker expected grouping of species within clades (Figure 6). and grouped all species as expected (Figure 1). IRBP displayed some variation to expected with Burramys not grouping with Cercartetus. Both IRBP and GAPD showed lower resolution at the deeper nodes (Figures 2 and 3).

Figure 6: Phylogenetic analysis of Diprotodonts for the concatenated data set for nuclear markers APOB, IRBP and GAPD and mitochondrial markers COI and ND2. Maximum likelihood analysis was performed using PHYML. The results in this study do not support the three suborder phylogeny of Wilson and Reader 2005. The concatenated data displayed support for the two suborders of Vombatiformes and Phalangerida as per Meredith et al. 2009. However, Burramyidae grouped with the Macropodiformes and importantly Phalangerida did not. Therefore it appears that the current published phylogeny of Meredith in regards to the grouping of the families within this order may not be accurate. In a forensic context, APOB and ND2 in particular would be useful for species identification within this order and given the degraded nature encountered within wildlife forensics, it is recommended that ND2 be used as a diagnostic tool in the first instance. This is due to the correct grouping of all species of the shallow nodes. Conclusion Figure 1: Phylogenetic analysis of Diprotodonts Figure 2: Phylogenetic analysis of Diprotodonts Figure 3: Phylogenetic analysis of Diprotodonts for the APOB nuclear marker. Figures represent for the IRBP nuclear marker. Figures represent for the GAPD nuclear marker. Figures represent This study shows the benefits of combining nuclear and mitochondrial data to produce a more robust phylogeny node support for maximum parsimony analysis node support for maximum parsimony analysis node support for maximum parsimony analysis based on 1,000 replications. based on 1,000 replications. based on 1,000 replications. to underpin a species identification test. Of the markers tested, ND2 provided the greatest level of diagnostic accuracy and could be used as a species identification tool for Diprotodonts.

References Affiliations Anon (1998) Australian Senate Select Committee on the Commercial Utilization of Native Australian Wildlife: www.aph.gov.au/Senate/committee/rrat_ctte/completed_inquiries/1996-99/wild/report/contents.htm. Aplin KP, Archer M (1987) Recent advances in marsupial systematics with a new syncretic classification. In: Archer, M. (Ed.),Possums and Opossums: Studies in Evolution. Surrey Beatty & Sons, Chipping Norton, NSW, pp. Xv–Lxxii. 1 2 Brower, A.V.Z. (2006) Problems with DNA barcodes for species delimitation: ‘ten species’ of Astraptes fulgerator reas¬sessed (Lepidoptera: Hesperiidae). Systematics and Biodiversity, 4, 127–132. Institute for Applied Ecology, University of Canberra, ACT Australia, Australian Federal Police, Forensic and Data Georges A., Birrell J., Saint K., McCord W.P. and Donnellan S. (1998). A phylogeny for side-necked turtles (Chelonia: Pleurodira) based on mitochondrial and nuclear gene sequence variation. Biological Journal of the Linnean Society, 3 4 London 67:213-246. Centres, Weston ACT Australia, Museum Victoria, Carlton Gardens VIC Australia, ANZPAA National Institute of Guindon S, Gascuel O. (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology, 52(5):696-704. Kirsch, J A W, Lapointe F J, Springer M S (1997) DNA-hybridisation studies of marsupials and their implications for metatherian classification.Australian Journal of Zoology 45:211-280. Forensic Science, Melbourne, VIC Australia. Hillis, DM. (1987) Molecular Versus Morphological Approaches to Systematics. Annual Review of Ecology and Systematics Vol. 18: 23-42. Mallet J. (1995) A species definition for the modern synthesis.Tree 10:294-299. McDowell D. (1997) Wildlife Crime Policy and the Law. Canberra: Australian Government Publishing Service. Meredith RW, Westerman M, Springer M S (2009) A phylogeny of Diprotodontia (Marsupialia) based on sequences for five nuclear genes.Molecular Phylogenetics and Evolution 51(3):554-71. Osborne ML, Christidis L, Norman JA. (2002) Moelcular Phylogenetics of the Diprotodontia (Kangaroos, wombats, koala, possums and allies). Molecular Phylohentics and Evolution 25:219-228. Shaffer, H.B.,. Clark J.M, and Kraus F. (1991) When molecules and morphology clash: A phylogenetic analysis of the North American ambystomatid (Caudata:Ambystomatidae) salamanders. Systematic Zoology 40(3): 284-303. Swofford DL. (2002) PAUP*. Phylogenetic analysis using parsimon (*and other methods). Sinaur Associates, Sunderland MA. www.anzpaa.org.au Wilson DE, Reeder DM (eds.) (2005) Mammal species of the world: a taxonomic and geographic reference. third edition. The Johns Hopkins University Press, Baltimore.