Magmatism during Gondwana break-up: new geochronological data from Westland, New Zealand

Abstract

Newly determined Late Cretaceous ⁴⁰Ar/³⁹Ar ages on megacrystic kaersutite from four lamprophyre dikes, and a U–Pb zircon age on a trachyte, from central and north Westland (New Zealand) are presented. These ages suggest that the intrusion of mafic dikes (88–86 and 69 Ma) was not necessarily restricted to the previously established narrow age range of 80–92 Ma. The younger lamprophyre and trachyte dikes (c. 68–70 Ma) imply that tensional stresses in the Western Province were either renewed at this time, or that extension and related magmatism continued during opening of the Tasman Sea. Extension-related magmatism in the region not only preceded Tasman seafloor spreading initiation (starting at c. 83 Ma, lasting to c. 53 Ma), but may have sporadically continued for up to 15 Ma after continental break-up.

Keywords:

• Gondwana break-up
• geochronology
• lamprophyre
• central Westland
• Fraser Complex

Introduction

Cretaceous–Paleogene rifting of the Gondwana margin thinned the continental crust of Zealandia and eventually culminated in the opening of the Tasman Sea between Australia and New Zealand (e.g. Bradshaw 1989; Tulloch & Kimbrough 1989; Laird & Bradshaw 2004; Schellart et al. 2006). In New Zealand, the earliest stages of thinning (c. 110–100 Ma) were associated with the formation of metamorphic core complexes (e.g. Tulloch & Kimbrough, 1989; Spell et al. 2000), accompanied by emplacement of granitoid plutons (Muir et al. 1997; Waight et al. 1997, 1998a; Tulloch et al. 2009). Progressive extension is expressed in the deposition of terrestrial Pororari Group sediments in extensional grabens across central Westland (e.g. Nathan 1978; White 1987; Beggs et al. 2008) and equivalents in the region offshore of Westland (Bishop 1992). Mafic dikes concentrated in the swarms of the Paparoa and Hohonu Ranges (Wellman & Cooper 1971; Hunt & Nathan 1976; Adams & Nathan 1978; Grindley & Oliver 1979; Waight et al. 1998b) intruded during the later stages of thinning. Felsic and mafic magmatic activity at c. 82 Ma (Waight et al. 1998b) is significant because it coincides with the age of the oldest seafloor in the Tasman Sea (Weissel & Hayes 1977; Veevers et al. 1991). It has therefore been regarded as magmatism coincident with the initiation of seafloor spreading that lasted until c. 53.3 Ma (Gaina et al. 1998). At the onset of seafloor spreading, the New Zealand microcontinent (Zealandia) formed through separation from Australia and Antarctica.

The granitoids and the core complex fabrics in the Western Province have been dated using U–Pb zircon and ⁴⁰Ar/³⁹Ar methods (Kimbrough & Tulloch 1989; Tulloch & Kimbrough 1989; Muir et al. 1997; Spell et al. 2000; Sagar & Palin 2011), and the occurrences of Cenozoic mafic magmatism in the Western Province and elsewhere in New Zealand have recently been reviewed by Timm et al. (2010). The timing of Late Cretaceous–Early Cenozoic lamprophyric and other mafic magmatism in the Western Province, coeval with extension and rifting, is much less precisely known. Initial K/Ar dating of dikes in the Hohonu Range and surrounding areas by Wellman & Cooper (1971) yielded a scatter of ages that they suggested represented an emplacement age of between c. 137 and 107 Ma. The majority of these ages have since been shown to be geologically implausible because the dikes intrude granitoids dated to be around 110 Ma (Waight et al. 1997). Waight (1995) and Waight et al. (1998b) argued on the base of cross-cutting relationships that lamprophyres in the Hohonu Range were instead intruded primarily before, but also synchronous with, the intrusion of the c. 82 Ma French Creek Granite (Tulloch et al. 1994; Waight et al. 1998b).

Thus far, the only reliable absolute ages of lamprophyric magmatism in Westland are those of Adams & Nathan (1978) who obtained K/Ar ages of 80–92 Ma (recalculated to decay constants of Steiger & Jäger 1977, see Table 1) from the Buller Gorge, in agreement with their cross-cutting relationships with coarse-grained non-marine sedimentary rocks of the Pororari Group. Although these and other older K/Ar ages effectively represent total gas ages with no control over the effect of possible open system behaviour in the whole rock (argon loss or uptake post-dating emplacement), they are at this time the best approximation of the actual intrusion age of most of the dikes in Westland.

Table 1 Recalculation of K/Ar ages of lamprophyre dikes and a basalt sill reported by Adams & Nathan (1978) to decay constants of Steiger & Jäger (1977).

Sample No.	Dated material	Intrusion type	Published age (Ma)	Recalculated age	± 2σ
P40127	Total Rock	Dike	82.5	84.4	1.2
P40726	Hornblende	Dike	83.2	85.1	1.2
P40727	Hornblende	Dike	84.2	86.2	0.4
P40728	Hornblende	Dike	78.1	79.9	1.2
P38713	Total Rock	Dike	79.3	81.2	1.2
P38704	Total Rock	Sill	89.9	92.0	1.6
			89.1	91.2	2

Table 2 Results of ⁴⁰Ar/³⁹Ar analyses by laser heating at Universität Potsdam and Lund University (OU79915, OU79943, OU79956 and KST-2).

Laser output (W)	^40Ar/^39Ar	^37Ar/^39Ar	^36Ar/^39Ar	K/Ca	% ^40Ar*	fr.% ^39ArK	^40Ar*/^39ArK	Age (Ma)	2σ
Sample ID: OU79943		J = 0.002650 (against 28.02 Ma FCT age)				Laboratory ID: C12023
1	750.32	130.97	2.7719	0.00	−7.75	0.03	−64.48	−338.0	± −380.5
1.2	523.66	102.18	1.6400	0.01	9.04	0.04	51.28	229.9	± 293.2
1.5	321.09	53.48	0.9452	0.01	14.36	0.08	48.05	216.2	± 126.9
1.8	31.29	62.60	0.2098	0.01	−81.88	0.07	−26.89	−133.3	± −158.0
2.2	32.86	23.04	0.0701	0.03	42.71	0.19	14.28	67.0	± 42.7
2.5	14.86	16.68	0.0028	0.03	99.96	32.63	15.04	70.5	± 4.9
2.7	14.53	10.34	0.0022	0.06	99.96	31.94	14.64	68.7	± 3.5
2.9	14.68	15.72	0.0016	0.04	99.95	28.07	14.85	69.6	± 4.6
3.1	15.08	13.80	0.0021	0.04	99.81	6.95	15.21	71.3	± 23.2
				Plateau age (four steps from 5.0 to 6.2%):				69.4	± 2.4
Sample ID: OU79956		J = 0.002654 (against 28.02 Ma FCT age)				Laboratory ID: C12031
0.7	116.96	25.88	0.3652	0.02	9.52	0.16	11.35	53.6	± 45.3
0.9	23.22	0.60	0.0095	0.97	88.15	7.12	20.48	95.5	± 1.3
1.1	20.74	0.20	0.0016	2.92	97.81	21.33	20.29	94.6	± 1.2
1.3	20.15	0.18	0.0006	3.27	99.13	23.94	19.98	93.2	± 1.5
1.5	20.44	0.21	0.0005	2.81	99.30	20.53	20.30	94.7	± 1.2
1.7	20.69	0.35	0.0006	1.68	99.30	12.30	20.55	95.8	± 1.3
1.9	21.18	0.41	0.0007	1.44	99.24	10.54	21.03	98.0	± 1.0
2.1	20.83	1.06	0.0007	0.56	99.47	4.07	20.74	96.7	± 1.6
				Plateau age (seven steps from 1.8 to 4.2%):				95.7	± 0.4
Sample ID: OU79915		J = 0.002658 (against 28.02 Ma FCT age)				Laboratory ID: C12032
0.8	35.07	1.41	0.0683	0.42	42.81	2.46	15.03	70.7	± 3.7
1	20.06	0.19	0.0038	3.15	94.50	18.56	18.96	88.7	± 1.3
1.2	19.72	0.18	0.0014	3.19	97.94	18.81	19.31	90.3	± 1.4
1.4	20.36	0.26	0.0009	2.24	98.76	13.24	20.11	94.0	± 1.3
1.6	20.07	0.22	0.0007	2.65	99.00	15.64	19.87	92.9	± 1.2
1.8	20.04	0.27	0.0007	2.18	99.12	12.86	19.87	92.9	± 1.0
2	19.91	0.28	0.0006	2.11	99.27	12.47	19.77	92.4	± 0.9
2.2	20.42	0.58	0.0005	1.01	99.48	5.96	20.32	94.9	± 1.2
				Plateau age (four steps from 2.8 to 4.0%):				92.9	± 0.6
Sample ID: KST2		J = 0.005848 (against 28.34 Ma TCR age)
2.1	3206.159	2.811398	8.8423	0.1533	18.5	0.014051	594.65389	2704.4	± 339.9
2.3	837.8695	0.5765036	2.2863	0.7477	19.4	0.028102	162.37311	1204.4	± 303.5
2.5	501.8065	1.272675	1.4349	0.3387	15.5	0.049179	77.95771	677.7	± 93.3
2.7	268.2241	1.794722	0.7938	0.2402	12.6	0.06323	33.82935	325.7	± 56.2
2.8	164.2169	0.8259922	0.5037	0.5218	9.4	0.117093	15.44906	156.0	± 34.5
2.9	144.0491	0.8190386	0.4583	0.5263	6	0.128803	8.70149	89.5	± 31.2
3.1	146.9716	0.8983874	0.4683	0.4798	5.9	0.161589	8.65932	89.1	± 27.8
3.3	99.43718	0.9563877	0.3031	0.4507	10	0.199059	9.9623	102.2	± 18.4
3.5	62.63894	0.9151869	0.1829	0.471	13.8	0.273998	8.67339	89.3	± 14.1
4	37.43918	1.446422	0.0871	0.298	31.6	0.634646	11.83436	120.7	± 6.6
4.5	15.14282	2.572341	0.0174	0.1676	67.4	3.592422	10.22118	104.7	± 2.5
5	9.190742	2.966431	0.0031	0.1453	92.5	94.73783	8.52067	87.7	± 0.3
				Total gas age:				88.5	± 0.6

Decay constants after Steiger & Jäger (1977), 4 amu discrimination = 1.0574 ± 0.11%, ⁴⁰K/³⁹K = 0.022 ± 69.30%, ³⁶Ca/³⁷Ca = 0.000264 ± 3.44%, ³⁹Ca/³⁷Ca = 0.000672 ± 0.94%.

Camptonitic lamprophyre dikes also occur in the Victoria Range (Tulloch 1979) and have now been located in the Bonar Range, c. 130 km south of the Buller area, where they cut sillimanite-grade gneisses (Jongens 2006; Scott et al. 2011a). Lamprophyric dikes are also reported between Bald Hill and the southernmost parts of Westland (south Westland: Nathan 1977, 1978; Bald Hill: Ryland 2011; Haast area: Randall 2004). In south Westland, the 64.8 ± 0.8 Ma Arnott Basalt (Timm et al. 2010) and a 61.4 ± 0.8 Ma rhyolite clast (Phillips et al. 2005) remain the only accurately dated magmatic events during Tasman Sea spreading. Mafic dikes have also been found in the deformed Fraser Complex and Granite Hill Complex (Rattenbury 1991; Waight et al. 1997). The vast majority of the dikes found in the Fraser Complex were deformed during mylonitization of the surrounding country rock (Rattenbury 1991) but this does not seem to be the case for dikes found on Granite Hill (Waight 1995; Waight et al. 1997).

Apart from mafic lamprophyres, more felsic quartz trachyte and phonolitic dikes occur sporadically. Two trachyte dikes have previously been dated by K/Ar in biotite, yielding ages of 64 and 73 Ma (Tulloch & Kimbrough 1989) These trachytes occur in the upper plate of the Paparoa metamorphic core complex and in the Victoria Range, and record extension during Tasman Sea spreading (Tulloch & Kimbrough 1989; Beggs et al. 2008).

In this paper, we document new radiometric ages for a selection of rocks from central Westland. ⁴⁰Ar/³⁹Ar ages are reported for lamprophyre dikes from Lake Brunner (or Lake Moana), the Hohonu Ranges and the Bonar Range (Fig. 1); these ages are supported by biotite ⁴⁰Ar/³⁹Ar cooling ages from Bonar Range orthogneiss. This new information reveals that the lamprophyre dike swarms represent either one prolonged or two (or more) distinct periods of extension and magmatism. We also report a U–Pb zircon age for a deformed trachyte dike in the Fraser Complex. Although imprecise, the interpreted age overlaps with ages of undeformed dikes outside the complex. The Fraser Complex protolith therefore appears to have experienced magmatism and tectonism similar to that of basement rocks of the Western Province.

Figure 1 Simplified geological map of central Westland, modified after Nathan et al. (2002). Sample locations are indicated by the green dots, accompanied by sample number and obtained dike intrusion ages. B, Bonar Range; Bu, Buller Gorge; C, Mt Camelback; G, Granite Hill; H, Hohonu Range; K, Mount Te Kinga; L, Lake Kaniere; M, Lake Brunner; PC, Paparoa Core Complex with arrows indicating the direction of stretching lineations (Tulloch & Kimbrough 1989); T, Mount Turiwhate.

Methods

⁴⁰Ar/³⁹Ar geochronology

Hand-picked mineral separates were analysed using the ⁴⁰Ar/³⁹Ar method at the University of Nevada Las Vegas (UNLV) (samples Hoh-1, Moa-16), Universität Potsdam in Germany (OU 79943, 79956 and 79915) and Lund University in Sweden (KST-2). The results are reported in Table 2. Details of the analytical methods used at UNLV have been reported in the Appendix to Scott et al. (2011b), and analytical procedures of ⁴⁰Ar/³⁹Ar dating at Universität Potsdam are summarized in Wiederkehr et al. (2009). Sample KST-2 was analysed at Lund University, Sweden following the methods described by Hermansson et al. (2008). Data from UNLV and from Universität Potsdam were standardized against FC-2, Fish Canyon tuff sanidine (with an assumed age of 28.02 ± 0.28 Ma after Renne et al. (1998)) at UNLV, whereas an age of 27.5 Ma (after Uto et al. 1997) was assumed for analyses carried out at Universität Potsdam, the reported data are corrected for this minor difference and are all standardized against the age after Renne et al. (1998). Data from Lund University were standardized against Taylor Creek rhyolite (TCR) sanidine (28.34 ± 0.28 Ma, Renne et al. 1998).

Total gas ages are calculated by weighting by the amount of ³⁹Ar released. A plateau age should consist of three or more contiguous gas fractions with analytically indistinguishable ages (i.e. all plateau steps overlap in age at ± 2σ analytical error) and comprise a significant portion of the total gas released (typically > 50%). For each sample, if possible, inverse isochron diagrams were constructed to check for the effects of excess argon. Reliable isochrons are based on the mean squared weighted deviates (MSWD) criteria of Wendt & Carl (1991) and, as for plateaux, must comprise contiguous steps and a significant fraction of the total gas released. Ages reported are at the 2σ (standard deviation) confidence level. For samples that did not yield a plateau age by these criteria, the age is approximated with a weighted mean of representative heating steps. In this case, the heating steps represent a significant fraction of the gas released but do not produce a real plateau age because one or more steps deviate more than 2σ. The determination of the preferred age for each sample is discussed below.

U–Pb geochronology

Zircon grains were extracted using conventional crushing and heavy liquid separation techniques, mounted in epoxy and then imaged under cathodoluminescence using the JEOL JXA-8600 microprobe at the University of Otago geology department. U–Th–Pb analyses were conducted by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) at the Australian National University, Canberra. All data were standardized against TEMORA-2 zircon and NIST-610 glass. The data acquisition method is the same as reported in detail by Scott & Palin (2008). Data points were excluded from age calculations if they had an internal MSWD of > 10 or if they were more than 10% discordant. Reported ages are at 2σ confidence level.

Sample setting and results

Lake Brunner and the Hohonu Range

The geology of the region surrounding Lake Brunner and the Hohonu Range is dominated by Cretaceous granitoids intruding Early Paleozoic Greenland Group metasedimentary rocks of the Buller Terrane (Waight et al. 1997, 1998a, 1998b; Suggate & Waight 1999). The Cretaceous granitoids (114–109 Ma), which are the most voluminous group of basement rocks in this area, are part of the I–S-type Rahu Suite (Waight et al. 1997, 1998a). A prominent exception is the volumetrically minor A-type French Creek Granite (81.7 ± 1.2 Ma) on the western flank of the Hohonu Range (Waight et al. 1997, 1998b). Most plutons are intruded by lamprophyric to doleritic and rare phonolitic/trachytic dikes of the Hohonu Dike Swarm. On the basis of field relationships, Waight et al. (1998b) inferred that the dikes are primarily older than, but also extend to younger than, the French Creek Granite. Rare composite dikes indicate contemporaneous mafic and silicic magmatism. The dikes may represent up to 40% of outcrop or stream boulders and several stream traverses indicate that c. 7% crustal dilation occurred during dike emplacement (Waight et al. 1998b). This indicates that, at least locally, dike magmatism accompanied significant crustal extension. Dike abundances decrease dramatically to the south and north of the Hohonu Range and Mount Te Kinga, with c. 30 dikes observed on Mount Turiwhate, and only two dikes found on Mount Tuhua directly west of Lake Kaniere (Fig. 1). The dike swarm shows a strong WNW–ESE preferred orientation, which parallels the line of opening of the Tasman Sea, and the age of the French Creek Granite (Waight et al. 1998b) overlaps with the oldest known oceanic crust in the Tasman Sea (Weissel & Hayes 1977), all strongly indicative that the alkaline magmatism of the Hohonu Dike Swarm and French Creek Granite was intimately linked to crustal extension and the early stages of break-up of New Zealand and Australia (Waight et al. 1998b).

Sample description

The dated samples (Hoh-1, Moa-16 and KST-2) from Lake Brunner and the Hohonu Range are lamprophyres that were targeted due to the presence of megacrystic kaersutite (K–Ti-rich amphibole, see Fig. 2). KST-2 was collected from an in situ composite dike in an unnamed creek draining into Lake Brunner. The dike is c. 2 m wide and comprises an outer margin of phonolite grading over a short c. 10 mm distance into lamprophyre containing abundant kaersutite megacrysts and spinel harzburgite xenoliths. It is likely to be a similar dike to that collected as a stream boulder for this study (Moa-16) and to the dike containing the mantle nodules described in detail by Tulloch & Nathan (1990). Figure 2 shows the boulder from which Moa-16 originates, with the visible large black kaersutite grains and similar sized peridotite xenoliths. In thin section, megacrystic kaersutite is orange brown–pale brown pleochroic and clear, rarely containing sparse up to 2 mm opaque inclusions. Sub-mm-wide veinlets of groundmass material intrude megacrystic kaersutite from the rims (Fig. 2B). The megacrystic kaersutite differs slightly from lamprophyre groundmass kaersutite because of the smooth shape (as opposed to euhedral crystals) and a commonly occurring sub-mm-thick (commonly partially resorbed) rim around megacrystic kaersutite on the outermost edges. The size of kaersutite ranges from several cm, visible in outcrop, down to c. 0.05 cm, slightly larger than euhedral kaersutite found in the groundmass. Different sections of the dike show a spread of matrix kaersutite ranging between 0.05 and 0.2 mm in the smallest dimension, commonly showing an aspect ratio of c. 1:4, kaersutite makes up between 10 and 50% of the groundmass. The remainder of the groundmass is made up of subhedral c. 0.1 × 0.7 mm grainy feldspar; extremely fine but up to 0.2 mm-long apatite needles are pervasive through the feldspar fabric. Opaque minerals range in size from < 0.01 mm, commonly up to 0.2 mm, and sometimes ranging up to a cm in size. Interstitial spaces between feldspar are filled with green, fine-grained aggregates of epidote and up to 1 mm sized carbonate. The typical carbonate + epidote and rarely fine-grained albite ocelli occur as up to 1 cm large patches, interpreted to represent the latest crystallizing volatile-rich phases. Anhedral to euhedral clinopyroxene that is rarely zoned and commonly partially altered to a fine-grained epidote aggregate occurs in sizes similar to megacrystic kaersutite (in thin section) and is interpreted to represent the xenocrystic equivalent originating from peridotite xenoliths with the rims representing possible magmatic overgrowth. The host dike can be classified as a camptonite (Rock 1987).

Figure 2 A, Photograph of the stream boulder sampled for Moa-16. Greenish grey lamprophyre contains abundant megacrystic black kaersutite and peridotite/pyroxenite xenoliths of several cm in size. Hammer shown for scale. B, Photograph of a thin section (plane polarized light) of sample KST-2, scale in figure. The thin section shows part of a megacrystic kaersutite grain (7 mm across) with a c. 0.2 mm thick rim and matrix made up primarily of smaller kaersutite and biotite with feldspar and opaque minerals and a fine-grained epidote aggregate (Ep). C, Photograph of thin section (plane polarized light) of sample Hoh-1, scale in figure, showing similar mineralogy to KST-2.

The dated kaersutite sample Hoh-1 comes from a lamprophyre dike intruding granite on a ridge crest in the Hohonu Range. This dike also contains xenoliths of granite and small xenoliths of spinel peridotite, along with xenocrysts of clinopyroxene, feldspar and spinel. The lamprophyre is inequigranular and composed of phenocrysts of augite (1 mm) and chlorite–tremolite pseudomorphs of olivine (1.5 mm) set in a groundmass composed of kaersutite, biotite, plagioclase and magnetite. Kaersutite megacrysts display a slight zoning at the rims (Fig. 2C). There are abundant xenocrysts of quartz or feldspar, and xenoliths of granite, always with a reaction corona of fine prismatic crystals of augite around their margins. Many mafic mineral species show marginal chloritization. Ocelli of coarse-grained calcite occur sporadically. Mineralogically, the rock is an alkali lamprophyre and the dominance of groundmass plagioclase indicates that it is a typical camptonite. The presence of granitic material in disequilibrium with the mafic assemblage in the Hoh-1 sample indicates minor contamination by the granitic wallrock.

Geochronology

⁴⁰Ar/³⁹Ar data for Hoh-1 do not fit criteria for calculating a plateau or isochron age. The total gas age is c. 80 Ma. However, the first three steps (15% of the ³⁹Ar released) have distinctly young ages (33–77 Ma and disrupted Ca/K values, see Table 3), possibly representing young disturbance. If these three steps are excluded from the age calculation on the basis that they represent out-gassing of alteration patches or inclusions in the kaersutite, the weighted average age is 86.8 ± 1.0 Ma, which incorporates 80.8% of the ³⁹Ar released (Fig. 3). This age is taken to be the age of dike intrusion. Moa-16 also fails to meet criteria for a plateau or an isochron (Fig. 3). The total gas age is c. 86 Ma; however, this pooled age includes the disturbed first four steps in the age spectrum (3.8% ³⁹Ar released, with variable Ca/K values, see Table 3). If these data points are excluded, the remaining data give a weighted average age of 86.4 ± 2.7 Ma (Fig. 3).

Figure 3 ⁴⁰Ar/³⁹Ar step heating age spectra of kaersutite from the four studied dikes. Boxes indicate 2σ errors on heating steps and on calculated averages. Ages displayed in figures are considered the best approximation of the age when no plateau or isochron age could be calculated.

Table 3 Results of ⁴⁰Ar/³⁹Ar analyses at UNLV (Hoh-1 and Moa-16).

T (°C)	^36Ar	^37Ar	^38Ar	^39Ar	^40Ar	%^40Ar*	% ^39Ar rlsd	Ca/K	^40Ar*/^39ArK	Age (Ma)	± 2σ
Sample ID: Moa-16			J = 0.001732 (against 28.02 Ma FCT age)
850	1.07	3.82	0.62	8.13	619.17	52.40	1.19	2.04	39.92	120.62	± 0.93
950	0.15	1.53	0.15	6.97	192.48	84.40	1.02	0.95	22.21	68.11	± 0.59
1050	0.10	2.51	0.15	5.79	172.24	90.20	0.85	1.89	25.38	77.61	± 0.59
1090	0.11	6.73	0.24	4.95	179.58	89.90	0.73	5.91	30.95	94.22	± 0.79
1100	0.07	8.11	0.24	5.34	169.84	96.20	0.78	6.60	28.97	88.33	± 0.66
1110	0.11	26.43	0.74	17.12	517.14	97.10	2.51	6.71	29.10	88.73	± 0.65
1120	0.24	105.20	2.91	68.27	1988.85	98.00	10.00	6.70	28.79	87.78	± 0.56
1130	0.31	227.88	6.09	149.80	4251.61	99.00	21.95	6.61	28.43	86.72	± 0.55
1140	0.26	269.85	7.10	175.43	4878.06	99.50	25.71	6.69	28.01	85.45	± 0.59
1160	0.23	295.16	7.67	194.23	5383.98	99.80	28.46	6.61	28.00	85.45	± 0.57
1190	0.09	34.20	0.91	22.71	655.37	98.80	3.33	6.55	28.34	86.44	± 0.71
1400	0.21	36.28	1.02	23.74	725.32	95.90	3.48	6.64	28.71	87.56	± 0.56
									Total gas age	86.41	± 0.53
Sample ID: Hoh-1			J = 0.001769 (against 28.02 Ma FCT age)
950	0.73	35.46	1.06	48.57	704.84	72.90	11.23	3.19	10.56	33.38	± 0.29
1050	0.14	19.76	0.35	8.37	206.97	87.10	1.93	10.33	20.65	64.72	± 0.56
1100	0.10	19.13	0.44	10.04	270.05	94.60	2.32	8.33	24.73	77.25	± 0.65
1110	0.12	32.46	0.76	15.87	438.42	96.20	3.67	8.94	26.29	82.01	± 0.64
1120	0.22	106.14	2.41	50.96	1451.12	97.60	11.78	9.11	27.95	87.08	± 0.68
1130	0.27	131.38	3.01	63.82	1799.76	97.50	14.75	9.00	27.71	86.32	± 0.69
1140	0.14	72.96	1.68	35.54	997.19	98.20	8.22	8.98	27.61	86.04	± 0.68
1150	0.12	37.03	0.84	17.65	511.78	97.10	4.08	9.17	27.83	86.69	± 0.74
1160	0.16	67.99	1.53	32.39	935.18	97.50	7.49	9.18	28.19	87.80	± 0.69
1190	0.31	162.06	3.58	76.84	2174.28	97.60	17.76	9.22	27.88	86.84	± 0.75
1250	0.30	126.16	2.78	60.07	1721.04	69.90	13.89	9.19	27.99	87.18	± 0.70
1400	0.19	25.97	0.63	12.47	392.45	93.00	2.88	9.11	27.79	86.57	± 0.77
									Total gas age	80.00	± 0.53

Isotope beams in mV, rlsd, released; error in age includes J error. ³⁶Ar to ⁴⁰Ar are measured beam intensities, corrected for decay for the age calculations. Decay constants after Steiger & Jäger (1977), 4 amu discrimination = 1.0574 ± 0.11%, ⁴⁰K/³⁹K = 0.022 ± 69.30%, ³⁶Ca/³⁷Ca = 0.000264 ± 3.44%, ³⁹Ca/³⁷Ca = 0.000672 ± 0.94%.

⁴⁰Ar/³⁹Ar data in KST-2 do not fit the criteria for calculating a plateau or isochron age because during step heating c. 95% of total argon was released in the last step (Fig. 3). The first heating steps yielded highly variable ages between 60 and 120 Ma; these steps are excluded from the age calculation on the basis that they may represent out-gassing of alteration patches or inclusions in the kaersutite. The age of the final heating step (87.7 Ma) may provide a closer approximation of the actual age of intrusion. This age overlaps with those obtained for Moa-16 and Hoh-1 (Fig. 3).

Bonar Range

The Bonar Range is composed of Devonian–Carboniferous orthogneisses and Early Paleozoic metasedimentary gneisses representing reworked Palaeozoic S-type granites and Greenland Group metasedimentary rocks, respectively (Jongens 2006; Scott et al. 2011a). U–Th–Pb ages of 370–350 Ma for monazite that has grown along biotite cleavage in one paragneiss was interpreted as indicating that metamorphism and deformation occurred during the Devonian–Carboniferous, coeval with the extensive S-type Karamea Suite granitic magmatism in the Western Province (Scott et al. 2011a). Although several mafic dikes cutting the gneissic country rock have been identified, they are less common than in the Hohonu–Lake Brunner area.

Biotite fractions from two orthogneisses (OU 79956 and OU 79915, see Fig. 4), and a kaersutite fraction from a phenocryst in a lamprophyre dike (OU 79943, see Fig. 3) were selected for ⁴⁰Ar/³⁹Ar chronology. Biotite in OU 79956 has a total gas age of 94.9 ± 0.5 Ma. The data meet the criteria for a plateau, which has an age of 95.7 ± 0.4 Ma for 99.8% of the ³⁹Ar released, as well as an isochron, which has an age of 95.5 ± 1.2 Ma and a ⁴⁰Ar/³⁶Ar value of 289 ± 32. Although the ages are effectively the same, the isochron ⁴⁰Ar/³⁶Ar value is statistically indistinguishable from atmospheric Ar (295.5) and is taken as the best result. ⁴⁰Ar/³⁹Ar data for the biotite fraction in OU 79915 have a total gas age of 91.3 ± 0.4 Ma, but this age incorporates a distinctly young initial step. The four central steps (54% of the ³⁹Ar released) meet the criteria of a plateau, and resolve an age of 92.9 ± 0.6 Ma. This age is considered the best result because the isochron age, 93.0 ± 1.8 Ma, has an unsatisfactory ⁴⁰Ar/³⁶Ar value for atmospheric Ar (129 ± 652).

Figure 4 ⁴⁰Ar/³⁹Ar step heating age spectra and isochrons of biotite from the Bonar Orthogneiss. Includes plateau ages and isochrons, errors are 2σ.

Although the biotite-bearing orthogneisses are cross-cut by lamprophyre and basaltic dikes, in situ samples were extensively altered and so a fresh lamprophyre specimen found as a stream boulder in the centre of the range was analysed. A kaersutite phenocryst from this lamprophyre (OU 79943) yielded a total gas age of 69.8 ± 2.9 Ma. Of the ³⁹Ar, 99.6% was released in the final four heating steps, which also meet the criteria for a plateau at 69.4 ± 2.4 Ma (Fig. 3). As the errors for isochron ages are high, no meaningful isochron age could be constructed. The most likely intrusion age is 69 Ma.

Fraser Complex

The Fraser Complex is an elongated group of deformed rocks sandwiched between the Alpine Fault and the Fraser Fault. The complex is composed of amphibolite–facies orthogneiss and paragneiss (Rattenbury 1991). The protolith may represent a metamorphosed equivalent of regional granitic and gneissic basement to Greenland Group sediments (Rattenbury 1991). More recently it has been suggested to represent metamorphosed Buller Terrane sedimentary rocks and intrusives, including 110 Ma Rahu Suite granitoids as the youngest unit identified within the protolith (Hiess et al. 2010). Equivalent but typically less-deformed lithologies to the Fraser Complex are found on Granite Hill east of Lake Brunner (Waight et al. 1997; Fig. 1). The Fraser Complex was metamorphosed at kyanite grade (Rattenbury 1991) in the Cretaceous (Hiess et al. 2010), and was subsequently mylonitized. K/Ar dating of biotite (Rattenbury 1987) suggests cooling through the biotite closure temperature (c. 450 °C, Villa 1998) at c. 44–61 Ma. Basaltic, lamprophyric and trachyte dikes in the Fraser Complex were deformed during mylonitization.

Because the lamprophyre dikes in the Fraser Complex are mylonitized and close to the active Alpine Fault, any determined ⁴⁰Ar/³⁹Ar ages could yield the age of cooling after deformation or be significantly disturbed rather than providing the timing of dike emplacement. Therefore, one trachyte dike (OU 55684) was selected for U–Th–Pb zircon geochronology. The selected sample was collected from the Tuke River and has been described by Rattenbury (1991). A total of 31 analyses were carried out on prismatic zircon grains from the trachyte by LA-ICP-MS (see Table 4). Some of the apparently youngest ages come from inclusion-rich zones in zircon grains. Qualitative electron microprobe investigations of these inclusions using energy dispersive spectrometry indicates that they comprise Th-rich phases [likely thorite (Th,U)SiO₄, see Fig. 5] and another silicate phase, probably biotite (too small to analyse but qualitative major element abundances suggest a biotite-like composition). Many analyses (including those in the thorite-bearing domains) were either discordant or had internal MSWD values of > 10; these data are excluded from any further age calculation. The remaining analyses yield ²⁰⁸Pb-corrected ages between 1241 and 63 Ma, with the older ages clearly inherited from the source region or contamination during ascent. The youngest six grains, which are from subhedral crystals that are larger than the older (inherited grains), yield a pooled age of 68.4 ± 3.5 Ma with an MSWD of 3.9. The high MSWD indicates that this data set could incorporate more than one population, but with such a small dataset, it is not possible to meaningfully refine this age. We interpret this to indicate that the emplacement of this trachyte dike occurred between 65 and 72 Ma (see Fig. 6). Although this age is imperfect, we note that it matches the age range of the two existing trachyte ages from Westland. A summary of the new geochronological results from this study is presented in Table 5.

Figure 5 Backscatter electron image of a zircon grain previously targeted for LA-ICP-MS analysis. LA-ICP-MS spots and the age of the specific spot are shown in pink. Bright areas as a result of secondary electron charging indicated by ‘C’. The image shows a coarse grain with inclusion-rich regions, examples of thorite (bright phases) and presumed biotite (dark phases) inclusions are indicated. Scale in figure.

Figure 6 Concorida plot including analysed grains up to 400 Ma, error ellipsoids indicate 2σ error.

Table 4 Full LA-ICP-MS results for sample OU55684.

Spot #	Pb* (ppm)	U (ppm)	Atomic Th/U	Uncorrected ^206Pb/^238U	± % (1σ)	Uncorrected ^207Pb/^235U	± % (1σ)	Uncorrected ^207Pb/^206Pb	± % (1σ)	Uncorrected ^208Pb/^232Th	± % (1σ)	^208Pb corrected ^206Pb*/^238U age (Ma)	± (1σ)	^208Pb corrected ^207Pb*/^235U age (Ma)	± (1σ)	^208Pb corrected ^207Pb*/^206Pb* age (Ma)	± (1σ)	Spot MSWD	Selected age (Ma)	± (1σ)
30	92.59	6219	2.13	0.00996	1	0.1212	3.2	0.0883	3.1	0.00347	4.9	61.9	1	66.3	30.8	261.9	111	14.52
21	3.54	347	0.07	0.0105	1.2	0.1531	3.7	0.1058	3.5	0.02183	4.6	63.1	0.8	72	6.6	381.1	29.7	1.5	63.1	1.9
23	114.2	5139	4.63	0.01002	0.9	0.085	3.5	0.0615	3.3	0.00304	1.9	65.3	2.5	193	54.6	2192.4	274	54.8
31	62.26	3833	2.22	0.01047	1.2	0.0899	2.7	0.0623	2.4	0.00365	3.5	65.5	1.6	32.4	23.7	0	0	3.2	65.5	2.4
27	123.4	6930	3.04	0.01015	0.8	0.0789	2.2	0.0564	2	0.00309	1.7	65.7	1	115.4	22.7	1272.2	150	9.16	65.7	2
28	115.1	5882	3.66	0.01009	0.6	0.0702	1.5	0.0505	1.4	0.00311	1	67.4	0.7	108.6	17.1	1155.5	114	30.07
29	73.3	3845	3.31	0.01041	0.7	0.0921	18.7	0.0642	18.7	0.0034	3.7	68.4	0.9	56.2	48.6	0.3	0.3	8.81	68.4	2.1
0	112.4	4458	4.78	0.0112	1.3	0.0785	6.6	0.0508	6.5	0.00362	2.1	70.4	4.2	13	74.2	0	0	3.27	70.4	4.6
15	3.03	239	0.4	0.01196	1.3	0.1614	5.6	0.0979	5.5	0.00729	3.8	72.5	1	82.1	10	370.7	39.3	1.5	72.5	2.2
18	4.8	426	0.03	0.01196	1.8	0.1103	10.4	0.0669	10.2	0.0311	13.5	74.7	1.4	71.5	12.1	3.2	0.6	0.57	74.7	2.4
19	7.62	471	0.26	0.01643	0.8	0.1212	4	0.0535	3.9	0.00593	3.2	104.6	0.9	106.1	5.3	143.3	7	1.5	104.6	3
20	6.73	323	0.87	0.018	1.2	0.1982	6.3	0.0798	6.1	0.00648	3.2	113	1.4	144.9	14.8	711.9	55.6	1.59	113	3.4
12	5.41	253	0.85	0.01871	1.7	0.2139	7	0.0829	6.7	0.00658	4	117.8	2.1	166.2	17.1	929.5	67	2.05	117.8	3.8
8	3.7	170	0.98	0.01843	1.3	0.1527	7.1	0.0601	7	0.00587	3.3	118	1.7	140.1	16.2	542.3	51.5	1.25	118	3.6
17	17.25	951	0.04	0.01942	0.7	0.1482	5.4	0.0553	5.4	0.02177	5.1	122.2	0.9	108.5	7.6	0.6	0	1.08	122.2	3.4
2	13.41	739	0.07	0.01955	1.4	0.1387	4.9	0.0515	4.7	0.00707	10.5	124.6	1.8	128.8	6.6	206.6	9.9	1.61	124.6	3.8
24	16.26	841	0.11	0.0205	1.6	0.1735	6.7	0.0614	6.5	0.00809	12.2	130.5	2.1	156.9	12.5	576.7	37.2	1.85	130.5	4.2
11	5.92	220	0.96	0.02333	1.6	0.2242	8.1	0.0697	7.9	0.00642	2.6	152.4	2.5	255.4	16.4	1359.9	53.2	1.91	152.4	4.8
10	33.27	1115	0.02	0.03212	2.3	0.3974	6	0.0897	5.5	0.10149	11.3	199.4	4.5	277.5	19	1000.7	47.6	3.1	199.4	7
13	3.95	81	0.64	0.04335	3	0.433	9.2	0.0724	8.7	0.01871	11	261.6	8.7	210.8	65.2	0.3	0.1	0.81	261.6	11.2
1	78.19	1904	0.04	0.04449	0.8	0.3531	1.9	0.0576	1.7	0.02234	15.2	279.8	2.1	295.8	5.8	423.8	7.3	1.71	279.8	7.9
26	56.15	1338	0.01	0.0459	0.8	0.3481	2.1	0.055	1.9	0.04586	12.4	288.9	2.4	298	5.6	369.7	6.2	2.93	288.9	8.2
3	28.35	571	0.25	0.05071	1.1	0.3708	3.2	0.053	3	0.01484	3.3	319.7	3.5	331.6	10.6	416.3	12.3	1.99	319.7	9.4
14	17.85	342	0.34	0.05184	1	0.3983	4.1	0.0557	4	0.0166	4.1	324.9	3.4	333.1	22.3	387.8	25.1	1.55	324.9	9.5
25	235.1	4964	0.01	0.05172	0.8	0.3673	1.7	0.0515	1.5	0.0158	11.9	325.1	2.4	318	4.6	267.2	3.6	3.09	325.1	9.2
4	13.94	252	0.38	0.05412	1.1	0.4305	4.9	0.0577	4.8	0.01771	3.3	338.9	3.6	350.3	19	427.1	22	1.39	338.9	9.9
7	33.94	590	0.42	0.05568	0.6	0.4413	2.2	0.0575	2.1	0.01824	2.8	348.3	2.1	358.5	13.9	424.2	15.9	1.13	348.3	9.7
6	28.96	241	0.33	0.11857	1.2	1.4068	4	0.086	3.8	0.04295	11	717.9	9.3	833.5	46.9	1160.5	56.3	4.16	717.9	21.7
5	30.34	247	0.2	0.12557	0.8	1.2965	4	0.0749	3.9	0.04091	5.5	761.9	5.8	831.6	26.4	1024.4	29.6	1.04	761.9	21.5
16	65.97	346	0.35	0.18667	0.8	2.0012	1.8	0.0777	1.6	0.05083	2.4	1105.6	7.8	1154.1	15.6	1241.3	15.2	0.72	1241.3	51.9
9	200	603	0.3	0.31927	1.3	6.5798	2.1	0.1495	1.6	0.12216	4	1773.7	20	1978.4	22	2208.3	19.5	1.19	2208.3	90.3

Notes: MSWD, internal spot MSWD < 10 for used ages. Data sorted by corrected ²⁰⁸Pb age. Ages were calculated from ²⁰⁶Pb*/²³⁸U ratios, where * indicates radiogenic Pb, total Pb minus common Pb. Common Pb corrections were applied based on the difference between measured and expected ²⁰⁸Pb/²⁰⁶Pb ratios for the measured ²³²Th/²³⁸U value, according to methods described by Compston et al. (1984). Concordance was calculated on the basis of agreement between ²⁰⁷Pb*/²³⁵U and ²⁰⁶Pb*/²³⁸U ages within their reported 1σ internal errors. Analyses > 10% discordant were not included in the age calculation. Concordance was calculated on the basis of agreement between ²⁰⁷Pb*/²³⁵U and ²⁰⁶Pb*/²³⁸U ages, and has a typical uncertainty of 5–10%.

Table 5 Summary of new age results and their locations.

Sample	Grid reference (NZMS260)	Lithology	Age (Ma)	Method
Lake Brunner
Hoh-1	K32 755 323	Camptonitic lamprophyre	87 ± 1.0	Weighted mean
Moa-16	K32 847 386	Camptonitic lamprophyre	87 ± 2.7	Weighted mean
KST-2	K32 844 388	Camptonitic lamprophyre	88 ± 0.3	Final heating step
Bonar Range
OU79943	I34 201 907	Camptonitic lamprophyre	69 ± 2.4	Plateau age
OU79915	I34 199 894	Bonar orthogneiss	93 ± 0.6	Plateau age
OU79956	I34 236 916	Bonar orthogneiss	96 ± 1.2	Isochron age
Fraser complex
OU55684	J34 368 952	Quartz trachyte (deformed)	68 ± 4	Youngest Zr population

Ages are ⁴⁰Ar/³⁹Ar ages considered most accurate, as explained in the text, errors are 2σ. Except OU55684 where a zircon age is reported and the reported error indicates the spread in the youngest zircon age population (N = 6).

Discussion

The new geochronological data extend the known range of dike magmatism in Westland and shows that at c. 68 Ma magmatism was bimodal with mafic as well as trachytic dikes being emplaced. Although not very voluminous, post-110 Ma magmatism preceding Gondwana break-up is widely distributed through central (Adams & Nathan 1978; Tulloch 1979; White 1987; Rattenbury 1987, 1991; Tulloch & Kimbrough 1989; Waight et al. 1998b; Jongens 2006) and north Westland (Hunt & Nathan 1976). Additionally, post-110 Ma felsic (e.g. Phillips et al. 2005; Tulloch et al. 2009) and mafic (e.g. Nathan 1978; Randall 2004; Timm et al. 2010; Ryland 2011) magmatic rocks occur sporadically throughout south Westland. Geochronological studies of central-north Westland, including the newly presented data, are summarized from south to north in Fig. 7.

Figure 7 Central–north Westland magmatic events between 100 and 50 Ma, occurrences sorted from south to north. Ranges indicate the spread in data from the same locality (Mt Camelback and Arahura-1 drillcore, Hohonu and Paparoa) and uncertainty on the single measurements for the Bonar Range, Fraser Complex and Southern Paparoa – Victoria ages. Dotted lines indicate the approximate ages for the beginning and end of Tasman Sea spreading (Weissel & Hayes 1977; Veevers et al. 1991; Gaina et al. 1998).

Deformation and magmatism in the Fraser Complex

Although imprecise, the interpreted 68 ± 4 Ma intrusion age of the deformed trachyte dike in the Fraser Complex overlaps with ages of mafic, as well as trachytic, magmatism outside the complex, indicating coupled magmatic and regional stress relations. Additionally, the age demonstrates that mylonitization of this Westland unit cannot be exclusively related to Early Cretaceous core complex formation (cf. Kimbrough et al. 1994) and is therefore likely to be related to the adjacent Alpine Fault, as speculated by Rattenbury (1987, 1991). Combined with the biotite K/Ar cooling ages of 44–61 Ma (Rattenbury 1987) this would constrain the ductile mylonitization that affected the trachyte dike to have occurred between an interval with 72 and 44 Ma as the oldest and youngest possible ages, respectively. However, this age range precedes the onset of the Alpine Fault at 30–25 Ma (Cooper et al. 1987; Schellart et al. 2006; Furlong & Kamp 2009). Therefore, the ages in the Fraser Complex would imply that a previously unidentified stage of deformation between 72 and 44 Ma is recorded within the Fraser Complex that has no equivalent elsewhere in the Western Province. A better explanation is that the biotite cooling ages of 44–61 Ma are inaccurate and too old. This could be due to total gas ages representing a mixture of different age domains within the analysed mineral, as shown in detail for rocks adjacent to the Alpine Fault by Batt et al. (2004).

Intrusion of dikes in the Bonar Range

The biotite plateau and isochron ages of 93–96 Ma for the Bonar Range orthogneiss represent a cooling of the rock below c. 450 °C, the suggested closure temperature for K/Ar in biotite (Villa 1998). This age does not appear to be reset by the intrusion of c. 69 Ma dikes or later activity related to the Alpine Fault. Rather, it coincides with cooling ages of the lower plate of the Paparoa metamorphic core complex to the north (Spell et al. 2000). Based on zircon fission-track analyses Ring & Bernet (2010) have argued that in contrast to coeval core complex formation, cooling of the Bonar Range was slow and controlled by erosional denudation. Some of the zircon fission-track cooling ages from the Bonar Range, however, are older than the newly acquired biotite ages from the Bonar Orthogneiss by up to c. 30 Ma. These data fail to describe a plausible cooling history because the annealing temperature of fission tracks in zircon (c. 280 °C, Bernet 2009) is lower than the K/Ar closure age in biotite (c. 450 °C, Villa 1998). Therefore the actual uplift history of the Bonar Range must be more complex, but it is beyond the scope of this article to discuss this in detail.

Because the Ar closure temperature for amphibole (c. 600 °C, Kamber et al. 1995) is higher than that of biotite, the older biotite age of the country rock confirms that the lamprophyre kaersutite age of c. 69 Ma represents the lamprophyre dike intrusion age and that it has not been reset by metamorphism.

Regional age correlations

Both lamprophyre dike locations dated in the Hohonu Range have similar ages of c. 88 Ma. One of the previously reported ages, K/Ar in hornblende of 92 ± 5 Ma (GA3464, Wellman & Cooper 1971), obtained from a dike along the northern slope of Mt Turiwhate (see Fig. 1) overlaps with the new age data and possibly reflects an accurate (albeit imprecise) intrusion age. Because field relationships indicate that dike intrusion was also contemporaneous with and post-dates the c. 82 Ma French Creek Granite, the new ages of 88 Ma indicate that the intrusion of dikes in the Hohonu Range lasted at least 6 Ma (from 88 to 82 Ma) and continued for an unknown period. The intrusion ages of lamprophyre dikes in the Hohonu Range therefore predate and overlap with the beginning of Tasman Sea spreading. This temporal relationship is also observed in basalts and lamprophyre dikes dated at 80–92 Ma (Adams & Nathan 1978) intruding Pororari Group sedimentary rocks in the Buller Gorge, directly north of the Paparoa metamorphic core complex. These ages indicate that mafic magmatism closely followed the cessation of rapid uplift of the core complex (Spell et al. 2000) that was later cut by trachyte dikes dated at c. 64 Ma with possible equivalents dated at c. 74 Ma in the Victoria Ranges to the east of the Paparoa Core Complex (Tulloch & Kimbrough 1989).

Bonar Range lamprophyre kaersutite ⁴⁰Ar/³⁹Ar (69 Ma) and Fraser Complex zircon U–Pb ages (68 ± 4 Ma) overlap with the previously dated (whole rock K/Ar) basalts at Mt Camelback and the base of the Arahura-1 drillcore (both c. 68 Ma, Sewell et al. 1988), situated between the Bonar and Hohonu Ranges and to the west of the Fraser Complex (see Fig. 1). The restricted spread in age of the c. 69 Ma lamprophyre, basalts and trachyte found in central Westland differs from the wide spread in generally older ages (80–92 Ma) found in mafic dikes intruding Pororari Group sediments and Rahu Suite granitoids in the Paparoa Ranges (Adams & Nathan 1978; Tulloch & Kimbrough 1989). The occurrence of Pororari Group sediments on and close to Mt Camelback (Gage & Wellman 1944) indicates a similar age of the onset of extension. Therefore the age of c. 69 Ma may represent an interval of local mafic magmatism in central Westland during the Tasman Sea rifting stage, post-dating the lamprophyres associated with initial break-up. Alternatively, the young ages might represent the later stages of a continuous period of mafic magmatism related to extensional tectonics. To resolve these questions, additional absolute dating, geochemistry and isotopic analysis of a larger set of lamprophyre dikes from across Westland is in progress.

Significance of c. 68 Ma magmatism

Dike emplacement at (88–82 Ma) is clearly related to regional tension that ultimately led to continental break-up (e.g. Waight et al. 1998b). A cause for the c. 68 Ma dikes and basaltic magmatism in Westland seems less obvious. Paleogeographic reconstructions at 68 Ma place the Western Province adjacent to the offshore Campbell Plateau, south of the South Island (e.g. Sutherland 1999). Petroleum wells and stratigraphic sections on Campbell Island indicate a sharp increase in subsidence rate for the Campbell Plateau at 68 Ma (Cook et al. 1999). This has been explained by the migration of the plateau away from low density mantle now beneath Eastern Antarctica (Sutherland et al. 2010). The c. 69 Ma dikes in the Western Province may also be related to this tectonic change.

Conclusions

1. New ⁴⁰Ar/³⁹Ar ages indicate that Gondwana break-up-related magmatism in the Western Province was mostly concentrated at c. 88 Ma and at c. 68 Ma. Limited data so far suggest two discrete pulses of magmatism but a continuous period of magmatism cannot be ruled out.

2. The emplacement of dikes in Westland at c. 69 Ma coincides with tectonic changes on the offshore Campbell plateau.

3. U–Pb zircon dating of a deformed trachyte dike in the Fraser Complex supports the interpretation that deformation of the Fraser Complex occurred in the Cenozoic. Deformation is likely coupled to activity of the Alpine Fault.

4. Biotite ⁴⁰Ar/³⁹Ar dating of Bonar Range gneisses appears to conflict with previous zircon fission-track thermochronology.

Acknowledgements

Two anonymous reviewers are thanked for their review of the manuscript. Matthew Leybourne is thanked for comments and efficient and involved editorial handling. Andy Tulloch is thanked for constructive discussion and help in constructing Table 5. Joseph Kula is thanked for assistance in data interpretation. JMS acknowledges financial support from FRST contract UOX1004. The Danish Research Council for Nature and Universe is acknowledged for their support to TEW and QHAVDM (grant no. 10-085245).