Cladistic methods as a tool for terrane analysis: a New Zealand example

ABSTRACT

A cladistic analysis of New Zealand's basement terranes was conducted using 44 present/absent characters that were derived from the literature for a broad range of geological characters, including zircon age profiles, clastic source, metamorphic grade and igneous petrology. The analysis resulted in a classification showing that both the Eastern and Western provinces are natural groupings. Two monophyletic clades (natural historical groupings) were identified within the Eastern Province: a Central Arc clade of South Island terranes (Maitai/Dun Mountain, Murihiku and Brook Street), and the younger greywacke terranes plus the North Island Murihiku Terrane. The Rakaia and Kaweka terranes of the older Torlesse are basal to (less derived than) the Central Arc and Eastern Province younger greywacke terranes. Although cladistic methodologies have been developed for biological systems, these results demonstrate a proof of principle for their application to terrane analysis.

KEYWORDS:

• Terrane analysis
• Australide orogeny
• cladistics
• tectonostratigraphic terranes
• Eastern Province
• Central Arc terranes
• Western Province

Introduction

The recognition that orogenies could be initiated by the accretion of exotic terranes (island arcs, large igneous provinces, micro-continental fragments and sedimentary basins) to continental margins (Coney et al. 1980) has stimulated research in the Southern Hemisphere (Howell 1980; Bradshaw et al. 1981). The ‘Australide’ orogeny (Vaughan et al. 2005) represented an accretionary complex some 7000 km long and 1500 km wide along the Australian, New Zealand, West Antarctic and South American margin of Gondwana. Accretion was staggered during the orogenic event, but the entire complex was active from the beginning of the Paleozoic (Ross and Delamerian orogenies of Antarctica and Australia) through to the Cretaceous (Rangitata orogeny of New Zealand), a period of more than 400 million years.

New Zealand workers were instrumental in developing terrane analysis as a tool for understanding complex, long-lived orogenies in terms of terrane formation, dispersal and amalgamation (Bradshaw et al. 1981). Terrane analysis consists of three steps: identification of individual terranes that are ‘suspect’ because their boundaries are faulted; geological characterisation (often but not necessarily showing they are ‘exotic’ with respect to the continental margin or other terranes to which they are now adjacent); and description of their histories (origin, movement and timing of amalgamation). Many different approaches have been applied to elucidating terrane history including comparisons using petrology, isotope geochemistry, provenance studies, sediment/clast geochemistry and paleontology.

We suggest that this last aspect of terrane analysis is one of analysing metadata (data from many different sources). Although individual studies may suggest certain correlations between terranes, what is the overall picture and how are conflicting data resolved? Whenever three or more objects are compared the issue of conflicting evidence arises. Some data may suggest that A and B are more similar to each other than either is to C {((A,B)C)}, whereas others may suggest that A and C are more similar {((A,C)B)}. Comparative biology, in particular systematics, has a long history of developing methods for resolving such conflicts (reviewed in Ebach 2017). Cladistics consists of a suite of methods that can be implemented using computer programs. Although a description of the many different approaches that have been developed and implemented since the 1970s is beyond the scope of this paper, interested readers are referred to standard texts such as Biological Systematics (Schuh and Brower 2009) and lists of available computer programs for their implementation and accompanying documentation (e.g. http://evolution.genetics.washington.edu/phylip/software.html).

Cladistics was developed in biological systematics as a way of grouping objects (usually but not exclusively species) in such a way as to minimise conflicting evidence (Schuh and Brower 2009). The results of such analyses are usually presented as branching diagrams (‘trees’) in which objects are grouped according to shared characters (e.g. presence of feathers). Cladistics differs from phenetic methods (based on overall similarity) in that only derived characters are used to discover these groupings. Character polarity (i.e. whether the character is derived or primitive) can be determined in several ways, most commonly by using an outgroup. The character state of an outgroup becomes the primitive state for the ingroup. A node is formed at the junction of branches, and all objects connected to any node form a monophyletic group, sometimes referred to as a clade. The identification of monophyletic groups is the primary aim of cladistics because such groups are natural in the sense that they are not artefacts but the products of evolution.

The application of cladistic methods to terrane analysis was pioneered by the Australian paleontologist Gavin Young (Young 1995). A recent use of terrane analysis of Caribbean, Central America and northern South American terranes by Echeverry et al. (2012), concluded that cladistic methodologies were an important analytical tool. Our aim is to extend the application of cladistic methods to terrane analysis using data available from the literature for New Zealand basement terranes.

Case study: New Zealand basement terranes

Howell et al. (1985) defined terranes as ‘fault-bounded entities of regional extent, each characterized by a geological history distinct from that of neighbouring terranes’. The basement rocks of New Zealand (Figure 1) consist of a series of tectonostratigraphic terranes. The youngest and most easterly group comprises Mesozoic volcaniclastics deposited in offshore basins and transported to their present positions (Figure 1: 1–5). These terranes are juxtaposed against Upper Paleozoic to Lower Mesozoic ophiolites, island arc volcanics and associated sedimentary units (Figure 1: 6–8). The greywacke and arc volcanic terranes are grouped together into an Eastern Province (Mortimer et al. 2014), although Adams et al. (2002) informally recognised the volcanic terranes as a distinct component within the Eastern Province, which they referred to as the Central Arc terranes (CATs). The most inboard of the CATs – the Brook Street Terrane – is sutured to Western Province rocks along the Median Tectonic Zone/Median Batholith in Nelson and Fiordland. CATs are also found in the North Island in the form of Murihiku rocks exposed at Kawhia (Waterhouse and White 1994) and by a lateral equivalent of the Dun Mountain Ophiolite that is inferred to be present at depth because of a strong positive magnetic anomaly along the western margin of the North Island (Hunt 1978). The Western Province is composed of the Paleozoic Buller and Takaka terranes that represent Devonian to Ordovician continental-derived sediments, Cambrian volcanic arc rocks, and later passive margin sediments respectively. Western Province terranes are found in the Nelson region, the west coast of the South Island, Fiordland, and possibly Stewart Island (Turnbull and Allibone 2003) but are not exposed in the North Island of New Zealand.

Figure 1. Locality Map showing New Zealand basement terranes.

The basement rocks of New Zealand have been the subject of a considerable research effort that has resulted in clearer definition of terrane boundaries (still ongoing in the Eastern Province terranes of the North Island; Leverenz and Balance 2001; Kear and Mortimer 2003), identification of possible sedimentary sources and places of origin (Wandres et al. 2004a, 2004b; Adams et al. 2007), classification of tectonic context, and correlation between units.

Material and methods

Units of analysis

The largest of New Zealand's tectonostratigraphic terranes were used as units of analysis. Of the 11 terranes, eight were subdivided geographically into North Island, South Island, Fiordland or Nelson, resulting in a total of 18 ‘taxa’ for analysis (Table 1). The selected terranes have been summarised by Briggs et al. (2004), Mortimer et al. (2014) and Adams et al. (2015), and chosen because they are well documented and contain sufficient geological characteristics to undertake a detailed cladistic analysis. Smaller units, such as the Akatarawa, Aspiring, Bay of Islands or Bitterness terranes, are included as part of larger terranes, or were too small for analysis (i.e. missing most geological characteristics). All terranes are listed in the New Zealand Stratigraphic Lexicon (http://data.gns.cri.nz/stratlex/search.jsp) and in Mortimer et al. (2014).

Table 1. Names and descriptions of tectonostratigraphic terranes used in the analysis. Age and regions from New Zealand Stratigraphic Lexicon http://data.gns.cri.nz/stratlex/search.jsp (Accessed 14 September 2017).

Province (code)	Terrane name (code)	Age	Region/remarks	References
Eastern Province	Waipapa Terrane (1)	Tr-J	Northern and central North Island	Leverenz and Balance (2001), Kear and Mortimer (2003), Adams et al. (2007), Jiao et al. (2014)
	Kaweka Terrane (2)	J-K	Rotorua, Kaimanua, Wanganui	Adams et al. (2009), Jiao et al. 2014
	Pahau Terrane North (3n)		Hawkes Bay, Wairarapa
	Pahau Terrane South (3s)	C-J	South Is. (Younger Torlesse Terrane)	Wandres et al. (2004a)
	Rakaia Terrane North (4n)	C-J	North Is. (Older Torlesse Terrane)	Adams and Graham (1996), Jiao et al. (2014)
	Rakaia Terrane South (4s)	C-J	South Is. (Older Torlesse Terrane)	Adams (2003), Wandres et al. (2004b), Mortimer et al. (2012)
	Caples Terrane Nelson (5nel)	P-Tr	Nelson	Mortimer (1993)
	Caples Terrane South (5s)	P-Tr	Marlborough, Southland, Wanganui, Westland	Mortimer et al. (2012)
Central Are Terranes	Dun Mt-Maitai Terrane Nelson (6nel)	P	Nelson	Rattecnbury et al. (1998)
	Dun Mt-Maitai Terrane South (6s)	P	South Is.	Turnbull and Allibone (2003), Mortimer et al. (2012)
	Murihiki Terrane North (7n)	Mes	North Is.	Adams et al. (2002), Briggs et al. (2004)
	Murihiki Terrane South (7s)	Mes	South Is.	Adams et al. 2002, Turnbull and Allibone (2003), Briggs et al. (2004), Mortimer et al. (2012)
	Brook Street Terrane Nelson (8nel)	P-J	Nelson	Rattecnbury et al. (1998)
	Brook Street Terrane South (8s)	P-J	South Is.	Tulloch et al. (1999), Adams et al. (2002), Turnbull and Allibone (2003), Davey (2005)
Western Province	Takaka Terrane Nelson (9nel)	Cam-D	Northwest Nelson	Wysoczanski et al. (1997), Rattecnbury et al. (1998), Tulloch et al. (2009), Adams et al. (2015), Jongens (2006)
	Takaka Terrane Fiordland (9f)	Cam-D	Fiordland	Scott et al. (2009), Tulloch et al. (2009), Turnbull et al. (2010), Adams et al. (2015)
	Buller Terrane Nelson (10nel)	Ord	Northwest Nelson.	Wysoczanski et al. (1997), Rattecnbury et al. (1998), Tulloch et al. (2009), Adams et al. (2015)
	Buller Terrane Fiordland (10f)	Ord	Fiordland	Scott et al. (2009), Tulloch et al. (2009), Turnbull et al. (2010), Adams et al. (2015)

Note: Cam, Cambrian; Ord, Ordovician; D, Devonian; C, Carboniferous; P, Permian; Tr, Triassic; J, Jurassic; K, Cretaceous; Mes, Mesozoic.

Geological characters

We treated the New Zealand basement terranes in the same manner as biological taxa. Taxa are compared in systematics based on the characters they possess, which can be morphological, behavioural, physiological, biochemical or genetic. These characters are assumed to be semi-independent in the sense that a change in one character is not linked to change in another. Developmentally linked characters cause a type of character-inflation, artificially increasing the apparent quantity of data but are not considered further here. From a theoretical perspective, any geological attribute can be used as long as it is semi-independent. Characters have been coded as one of two states – presence (1) or absence (0) – because unlike biological character states that may be developmentally linked, no such linkage exists for geological characters. Continuously variable characters, such as metamorphic grade, have been scored as discrete characters by choosing states along a continuum (low, medium, high), and complex characters, such as zircon age profiles, by geological period allowing for the capture of subtle differences in age patterns between terranes.

There is a practical need to have as many characters as possible in the analysis, because the more (informative) characters there are the greater the potential for the resolution of relationships. The characters used in this study were extracted from published sources (Table 1). The coding of 44 geological characters is shown in Table 2. The characters are grouped into 14 sets that represent types of geological information (e.g. age, zircon profile, clastic source, tectonic setting, metamorphic age and grade, type of volcanics, amalgamation age, granite age and petrology). The characters cover a wide range of geological history to represent the whole terrane rather than, for example, recent depositional clastic or volcanic material.

Table 2. A list of characters and character-states used in the analysis.

Character		Character states
Age	1. Precambrian	Absent (0), present (1)
	2. Early Paleozoic (Cambrian–Silurian)	Absent (0), present (1)
	3. Late Paleozoic (Devonian–Permian)	Absent (0), present (1)
	4. Early Mesozoic (Triassic–Jurassic)	Absent (0), present (1)
	5. Late Mesozoic (Cretaceous)	Absent (0), present (1)
Zircon profile	6. Precambrian	Absent (0), present (1)
	7. Early Paleozoic (Cambrian–Silurian)	Absent (0), present (1)
	8. Late Paleozoic (Devonian–Permian)	Absent (0), present (1)
	9. Early Mesozoic (Triassic–Jurassic)	Absent (0), present (1)
	10. Late Mesozoic (Cretaceous)	Absent (0), present (1)
Clastic source	11. Island arc	Absent (0), present (1)
	12. Continental arc	Absent (0), present (1)
	13. Continental	Absent (0), present (1)
Tectonic setting	14. Oceanic arc	Absent (0), present (1)
	15. Accretionary complex	Absent (0), present (1)
	16. Continental	Absent (0), present (1)
Metamorphic age	17. Early Paleozoic (Cambrian–Silurian)	Absent (0), present (1)
	18. Late Paleozoic (Devonian–Permian)	Absent (0), present (1)
	19. Early Mesozoic (Triassic–Jurassic)	Absent (0), present (1)
	20. Late Mesozoic (Cretaceous)	Absent (0), present (1)
Metamorphic grade	21. Low	Absent (0), present (1)
	22. Medium	Absent (0), present (1)
	23. High	Absent (0), present (1)
Volcanics	24. Basic	Absent (0), present (1)
	25. Intermediate	Absent (0), present (1)
	26. Ophiolites/ultrabasics	Absent (0), present (1)
Amalgamation age	27. Early Paleozoic (Cambrian–Silurian)	Absent (0), present (1)
	28. Late Paleozoic (Devonian–Permian)	Absent (0), present (1)
	29. Early Mesozoic (Triassic–Jurassic)	Absent (0), present (1)
	30. Late Mesozoic (Cretaceous)	Absent (0), present (1)
Mélange boundary	31. Mélange boundary	Absent (0), present (1)
Granite age	32. Early Paleozoic (Cambrian–Silurian)	Absent (0), present (1)
	33. Devonian	Absent (0), present (1)
	34. Carboniferous	Absent (0), present (1)
	35. Cretaceous	Absent (0), present (1)
Devonian granite petrology	36. S-type	Absent (0), present (1)
	37. I-type	Absent (0), present (1)
Tahua orogeny folding type	38. F1	Absent (0), present (1)
	39. F2	Absent (0), present (1)
Carboniferous granite petrology	40. S-type	Absent (0), present (1)
	41. I-type	Absent (0), present (1)
	42. A-type	Absent (0), present (1)
Cretaceous granite petrology	43. I-type	Absent (0), present (1)
	44. A-type	Absent (0), present (1)

Cladistic analysis

As discussed previously, there are many different cladistics packages providing a range of analysis options, but an adequate description and commentary are beyond the scope of this study. We have run an efficient algorithm based on parsimony, namely on the shortest number of steps necessary to achieve a result. First, a binary absence/presence matrix was constructed in which the rows represent the tectonostratigraphic terranes and the columns the geological characters (Table 3). Presence of geological character is coded as ‘1’, absences as ‘0’, and missing or doubtful information as ‘?’. This binary data matrix was analysed with the cladistic software TNT v1.5 (Goloboff et al. 2008), using the New Technology Search (ratchet iterations = 1000; drift cycles = 100) and implied weighting (set to the K=10). A strict consensus tree is constructed from all the most parsimonious cladograms. This tree was rooted using an all zero outgroup because the absence of a character in geological taxa, unlike biological taxa, can never be a derived state.

Table 3. The parenthetical terrane/character data matrix consisting of 18 terranes (units) and 44 lithological characters.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	40	41	42	43	44
1	0	0	1	1	0	0	0	1	1	1	0	1	0	0	1	0	0	0	1	0	1	0	0	1	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
2	0	0	0	1	0	0	0	1	1	1	0	0	1	0	1	0	0	0	1	1	0	1	0	0	0	0	0	?	?	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
3n	0	0	0	1	1	1	0	1	1	1	0	1	0	0	1	0	0	0	1	0	1	0	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
3s	0	0	0	1	1	1	0	1	1	1	0	1	0	0	1	0	0	0	1	0	1	0	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
4n	0	0	0	1	0	1	0	0	1	0	0	0	1	0	1	0	0	0	1	1	1	0	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
4s	0	0	1	0	0	0	1	1	0	0	0	0	1	0	1	0	0	?	1	0	1	1	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
5nel	0	0	0	1	0	0	0	1	1	0	0	1	0	0	1	0	0	0	0	?	1	1	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
5s	0	0	1	1	0	0	1	1	1	0	0	1	0	0	1	0	0	0	?	1	1	1	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
6nel	0	0	1	1	0	0	0	1	1	0	1	0	0	1	0	0	0	?	?	0	1	0	0	1	0	1	0	1	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
6s	0	0	1	1	0	0	0	1	0	0	1	0	0	1	0	0	0	1	1	0	1	0	0	1	1	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
7n	0	0	0	1	1	0	0	0	1	0	0	1	0	0	0	1	0	0	1	1	1	0	0	0	0	0	0	0	?	?	0	0	0	0	0	0	0	0	0	0	0	0	0	0
7s	0	0	1	1	0	0	0	1	1	0	1	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
8nel	0	0	1	1	0	0	0	0	1	0	1	0	0	1	0	0	0	0	0	0	1	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
8s	0	0	1	1	0	0	0	0	1	0	1	0	0	1	0	0	0	0	1	0	0	1	0	1	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
9nel	0	0	1	0	0	1	1	0	0	0	1	0	0	1	0	0	1	0	0	1	0	0	1	1	0	1	0	1	0	0	0	0	1	1	1	0	1	1	1	1	1	1	1	0
9f	0	1	0	0	0	1	1	0	0	0	1	0	0	1	0	0	0	1	0	1	0	0	1	1	0	1	?	?	0	0	0	1	0	1	1	0	0	0	1	1	0	1	1	0
10nel	1	1	0	0	0	1	1	0	0	0	0	0	1	0	0	1	1	0	0	1	1	0	0	0	0	0	0	1	0	0	0	0	1	1	0	1	0	1	1	1	1	1	1	1
10f	0	1	0	0	0	1	1	0	0	0	0	0	1	0	0	1	1	1	0	1	0	0	1	0	0	0	?	?	0	0	0	0	0	1	1	1	0	0	1	1	0	0	1	0

Results and discussion

The strict consensus tree is shown in Figure 2 and synapomorphy distribution over the tree in Figure 3. Character-based trees show only relationships between taxa and do not imply a time axis. The analysis recovered both the Western and Eastern provinces as monophyletic clades and confirms these groupings are natural and not artefacts. Within the Eastern Province there are two other monophyletic groupings – that of the South Island CATs and the younger greywacke terranes plus the Murihiku Terrane of the North Island.

Figure 2. Classification scheme or cladogram of New Zealand Terranes using cladistic analysis. TNT found a single parsimonious tree (with implied weight set at K = 10). Terrane symbols as in Figure 1.

Figure 3. A cladogram showing synapomorphies (characters) listed in Table 2. Character-states with an asterisk are non-homologous.

Within the Western Province, the Takaka terranes of Nelson and Fiordland are sister taxa indicating they were once geographically contiguous. The analysis did not recover a similar relationship between the Buller terranes of Nelson and Fiordland, showing instead that the Buller Terrane is non-monophyletic. However, this is probably a result of the study's lack of resolving power and additional data may well show that the Buller Terrane was originally contiguous also. While the Eastern Province as a whole is a natural grouping, there are some interesting features within it, including the separation of the older from other greywacke terranes, the recovery of the CATs as a natural grouping, and the non-monophyly of the Murihiku Terrane.

The older Torlesse terranes do not group with other greywacke terranes but occupy a basal position within the Eastern Province. They do not form a monophyletic group, although as in the case of the Buller Terrane of the Western Province this may be a result of incomplete resolution due to an insufficient number of characters used. The results show that the North Island Rakaia Terrane (Figure 1:4n) is probably best treated as a more highly metamorphosed part of the Kaweka Terrane (Figure 1:2) and that the name Kaweka be used exclusively for the older Torlesse in the North Island. If the older Torlesse terranes of the North and South Islands are shown to form a monophyletic group in the future, then both should be referred to as the Rakaia Terrane

Leaving aside the North Island Murihiku Terrane, which is discussed below, the remaining greywacke terranes form a monophyletic group that is a sister group to the CATs of the South Island. Within the greywacke terranes there is little resolution in sister group relationships apart from the North and South Island Pahau terranes. This sister group relationship supports the hypothesis that the younger Torlesse greywackes form a once contiguous terrane, and that these sediments were laid down in a single basin. The Murihiku Terrane of the North Island (Figure 1:7n) groups with the other greywacke terranes and is sister taxon to the younger Torlesse Pahau Terrane. It is unrelated to its supposed correlative in Southland/Otago or indeed any other Central Province terrane. It has long been known that there are significant differences between the Murihiku of the North and South islands in paleontology (Grant-Mackie 1985; Damborenea and Manceñido 1992), and lithology and petrology (Briggs et al. 2004). These differences have been explained as artefacts of facies differences or along-arc changes in petrology such as seen in the modern Taupo–Kermadec arc (Briggs et al. 2004). Our results suggest these differences reflect deposition in unrelated basins and do not support the hypothesis of a Murihiku Supergroup as currently defined (Campbell et al. 2003). The results of the analysis imply that the North Island Murihiku Terrane is a shelf equivalent of the deeper water, shelf-slope, younger Torlesse turbidites. A fault-bounded sliver of Murihiku rocks present in the Nelson district was not included in this analysis because insufficient data were available to do so. The CATs of the Nelson and Southland form a natural grouping as suggested by Adams and his co-workers (Adams et al. 2002). Although precise relationships between these terranes are not well resolved in the analysis, there is no clear support for a correlation between individual terranes of Nelson and Southland.

Conclusion

The appropriation of a method or technique by one branch of science from another has been a recurring theme in scientific research, providing valuable cross-fertilisation of ideas and stimulating novel research directions (Rinia et al. 2002). However, caution is needed to ensure that methods of analysis are truly applicable to the new field and to answering its aims. Cladistic methods were developed to group species together on the basis of shared-derived characters (homologues). An important assumption, on which cladistic methods are based, is that homologues are transmitted vertically within lineages and not horizontally between lineages. We believe this assumption holds good for geological examples such as terranes because there is no ‘horizontal flow’ of characters. A terrane acquires a characteristic, such as a particular zeolite age profile, which may be shared with other terranes, but which is uniquely retained in some measure by that terrane in perpetuity.

We are confident that this study demonstrates proof of principle because the analysis recovered natural groupings (monophyletic clades) recognised at the descriptive level. While the recovery of both the Western and Eastern provinces as natural groupings will come as no surprise to the New Zealand geological community, this does vindicate the hypothesis testing approach adopted in this study. Furthermore, the results offer supplementary information about unresolved issues within the geological discourse such as the possible artificiality of the Murihiku Supergroup, providing support for a separate Central Arc Terrane grouping, and outlining the close relationship of the North Island Murihiku basin to the younger Torlesse greywackes of the Pahau Terrane.

As a proof-of-concept study, we have shown that testing the historical relationships between terranes using geological characters can provide further analytical insight alongside current approaches. Terrane analysis of single areas is clearly a necessary starting point, but the artificiality of splitting the Australides into arbitrary geographic regions like New Zealand or South America does not take into account the movement of terranes within the accretionary zone. It is more than likely that some New Zealand terranes, for example, are more closely related to terranes (or source areas) in Australia (Adams et al. 2005) or South America (Wandres and Bradshaw 2005) than they are to other New Zealand terranes. In using terrane analysis approach, geologists have a robust tool that can test existing hypotheses of relationship, particularly terranes now dispersed across the southern continents during the long and complex Australides orogeny.

Acknowledgements

We thank an anonymous reviewer, Roger Cooper, and Nick Mortimer for helping improve the manuscript. EMD wishes to thank the UNSW for providing an Australian Postgraduate Award and PANGEA.

Disclosure statement

No potential conflict of interest was reported by the authors.