ABSTRACT The type and magnitude of the physical and chemical processes controlling magma ascent, storage and solidification in zoned plutons remain controversial, even though such plutons are widespread in arcs. Linking both plutonic and host rock characteristics is key towards determining quantitative models for the growth and evolution of these plutons. We report a synthesis of an integrated study of the Loch Doon pluton, an archetypal zoned pluton in the Southern Uplands Terrane of Scotland, including linked plutonic and host rock field mapping, structural analysis, LA-ICP-MS U–Pb zircon geochronology, whole-rock and isotope geochemistry and magnetic susceptibility analyses. Our results reveal that trapping of rising magma and construction of the pluton in the upper crust at ∼397 Ma was dominated by vertical host rock displacement: likely a combination of downward host rock flow, stoping and floor subsidence, and to a lesser extent upward doming and faulting. Mixing of depleted mantle-derived and crustal melts, and fractionation and contamination of these at deeper crustal levels, was followed by continued in situ fractionation and mixing during crystallization that was extended by continual chamber rejuvenation and thermal buffering by recharge from magmas. Analysis of the pluton’s ubiquitous magmatic foliations and their relationship to internal magmatic and external host rock structures suggests that later, typically more evolved magma batches ‘nested’ within the rheologically weakest portions of earlier batches and made space by largely displacing these earlier intrusive batches rather than the external host rock. The active parts of the nesting system decreased with time, consistent with a freezing magma feeder system. This model is applicable to many other zoned late Caledonian Southern Uplands Terrane plutons of the United Kingdom. These Southern Uplands Terrane plutons appear to be magmatic ‘fingers’ rising from larger, deeper intrusive complexes that resided at depth. Our new data also allow us to present further constraints on the geodynamic model for this part of the Caledonian orogenic belt. Our interpretation of antecrystic and autocrystic zircon ages in Southern Uplands Terrane plutons indicates prolonged magmatism between ∼430–382 Ma and analysis of whole-rock Sr/Y ratios suggests crustal thicknesses during magma generation may have varied in space and, or, time between ∼30–45 km. Our data best fit a tectonic model in which magmatism was initiated at ∼430 Ma in response to break-off of subducted Iapetan lithosphere and was extended by additional melt/heat input during regional transtension at ∼415–400 Ma. Magmas evolved by complex open system fractionation, mixing, and contamination in a mid-lower crustal batholith before being emplaced as compositionally zoned nested plutons in the upper crust between ∼414–382 Ma. INTRODUCTION Zoned plutons and their host rocks commonly display complex geologic features, making it difficult to assess the type and magnitude of magmatic processes and host rock displacement that operated during their growth and subsequent evolution. Whilst U–Pb series zircon dating is increasingly being used to show that zoned plutons are long-lived, open systems with complex and dynamic mechanical behaviour (Matzel et al., 2006; Lipman, 2007; Miller et al., 2007; Miller, 2008; Memeti et al., 2010; Paterson et al., 2011; Schoene et al., 2012; Barboni et al., 2015), as yet, there is limited consensus on the exact type and spatio-temporal extent of the physical and chemical processes controlling magma ascent, storage and solidification. Lack of consensus in part results from most studies focusing on one aspect of these plutons. Understanding the response of magma-host rock systems to changes in bulk composition, fluids, temperature, viscosities, and deviatoric stress is key towards quantifying: (1) the material transfer processes involved during magma ascent and chamber growth, specifically, the mechanisms of host-rock displacement around a growing pluton; (2) how older crystal mushes are displaced during the arrival of younger pulses of magmas; and (3) the extent of syn- to post-emplacement fractional crystallization, mingling and recycling. Here, we explore several archetypal zoned plutons and provide a semi-quantitative analysis of emplacement and post-emplacement processes through integration of multiple host rock and plutonic datasets. Furthermore, given that melt generation and pluton growth evolve in response to local and, or, regional scale tectonism, we demonstrate the value of this integration in evaluating the timing and style of tectonic processes, even in well studied orogens such as the British Caledonides. The Loch Doon pluton (LDP) is an archetypal, normally zoned calc-alkaline pluton in the Caledonian orogenic belt of northern Britain (Figs 1 and 2). The composition of the pluton ranges from diorite (opx–cpx) at the margin, through quartz diorite (bio–opx–cpx; Fig. 3a), tonalite (bio–act–hbl; Fig. 3b), granodiorite (bio–act–hbl; Fig. 3c), granite (bio–act–hbl; Fig. 3d) and microgranite (bio–musc–cord; Fig. 3e) at its core (Tindle, 1982). Previous studies of this pluton have primarily focused on the varied geochemical and isotopic signatures with little emphasis on the implications of the ubiquitous host rock and plutonic structures. A range of pluton emplacement mechanisms have been suggested, including laccolithic doming (Gardiner & Reynolds, 1932; Ruddock, 1969), in situ assimilation (Rutledge, 1950; Oertel, 1955) and multiple material transfer processes (Paterson & Vernon, 1995), including various amounts of assimilation, host rock shortening, host rock translation, stoping, and downward flow of early magma pulses. To explain the origin of the geochemical and isotopic zonation, a two- to three-stage in situ fractionation model has been widely accepted, where two or three pulses of magma intruded consecutively and subsequently fractionated in situ in a relatively closed magma system. Here, we show that the previous models do not suitably explain all the geologic datasets available for the pluton. We re-evaluate the previous literature and integrate it with our new field mapping, U–Pb zircon dating, geochemistry and magnetic analyses to provide new insight into pluton emplacement and magma chamber processes that attempts to satisfy all the various datasets. We extrapolate and test our model on other nearby zoned plutons in the Southern Uplands Terrane (SUT) of the UK. By addressing questions about zoned plutons, we hope to provide an important reference that may be applied to other zoned plutons worldwide. Fig. 1. View largeDownload slide Regional distribution of Caledonian plutonic and volcanic rocks, and major Caledonian faults in northern Britain (ages after: Oliver et al., 2008 and references therein; Neilson et al., 2009; Miles et al., 2014; Cooper et al., 2016; this study). Pluton outlines based upon BGS DiGMapGB-625, with permission of the British Geological Survey). Fig. 1. View largeDownload slide Regional distribution of Caledonian plutonic and volcanic rocks, and major Caledonian faults in northern Britain (ages after: Oliver et al., 2008 and references therein; Neilson et al., 2009; Miles et al., 2014; Cooper et al., 2016; this study). Pluton outlines based upon BGS DiGMapGB-625, with permission of the British Geological Survey). Fig. 2. View largeDownload slide Summary geological map of the Loch Doon pluton and its host rocks, including our new mapping of the ∼65 km2 southern half of the pluton (map after: Gardiner & Reynolds, 1932; Oertel, 1955; Ruddock, 1969; Tindle, 1982; and BGS DiGMapGB-50, with permission of the British Geological Survey). The extended key includes symbols and colours also used in subsequent maps. See Figure 4 and Supplementary Data A2 for age data. Fig. 2. View largeDownload slide Summary geological map of the Loch Doon pluton and its host rocks, including our new mapping of the ∼65 km2 southern half of the pluton (map after: Gardiner & Reynolds, 1932; Oertel, 1955; Ruddock, 1969; Tindle, 1982; and BGS DiGMapGB-50, with permission of the British Geological Survey). The extended key includes symbols and colours also used in subsequent maps. See Figure 4 and Supplementary Data A2 for age data. Fig. 3. View largeDownload slide Polished hand samples (a)–(e) and field photographs (f)–(i) of the Loch Doon pluton (see Fig. 2 for sample/photograph locations). (a) quartz diorite. (b) tonalite. (c) granodiorite. (d) granite. (e) microgranite containing dark green magmatic cordierite clots with leucocratic reaction halos. All photographs in (a)–(e) are oriented roughly vertically as in outcrop and are all equal scale, 9 cm high and 3 cm across. (f) View of the southeastern quadrant of the pluton showing how the pluton–host rock contact and internal plutonic contacts crosscut topography indicating the steep nature of the contacts. QzD, Quartz Diorite; Ton, Tonalite; GrD, Granodiorite; Gr, Granite. Scale: 8 m wide forest road right of centre. (g) A rare example of voluminous magma mingling in the March Burn. The sub-rounded mafic enclave is hosted by coarse grained granodiorite which is hosted by and inter-fingered with coarse grained granite. Scale: 25 cm long hammer. (h) Relationship between xenoliths, enclaves and magmatic foliations in granodiorite south of Craigminn. Scale: 1 cm wide pencil. (i) Relationship between enclaves and magmatic foliations in granodiorite west of Clashdan Brae. Scale: 1 cm wide pencil. Fig. 3. View largeDownload slide Polished hand samples (a)–(e) and field photographs (f)–(i) of the Loch Doon pluton (see Fig. 2 for sample/photograph locations). (a) quartz diorite. (b) tonalite. (c) granodiorite. (d) granite. (e) microgranite containing dark green magmatic cordierite clots with leucocratic reaction halos. All photographs in (a)–(e) are oriented roughly vertically as in outcrop and are all equal scale, 9 cm high and 3 cm across. (f) View of the southeastern quadrant of the pluton showing how the pluton–host rock contact and internal plutonic contacts crosscut topography indicating the steep nature of the contacts. QzD, Quartz Diorite; Ton, Tonalite; GrD, Granodiorite; Gr, Granite. Scale: 8 m wide forest road right of centre. (g) A rare example of voluminous magma mingling in the March Burn. The sub-rounded mafic enclave is hosted by coarse grained granodiorite which is hosted by and inter-fingered with coarse grained granite. Scale: 25 cm long hammer. (h) Relationship between xenoliths, enclaves and magmatic foliations in granodiorite south of Craigminn. Scale: 1 cm wide pencil. (i) Relationship between enclaves and magmatic foliations in granodiorite west of Clashdan Brae. Scale: 1 cm wide pencil. The general calc-alkaline geochemical affinities, post-collision ages and variable isotopic characteristics of plutonic and lamprophyre magmas in the SUT has led to considerable debate about the petrogenetic processes and tectonic controls of late Caledonian magmatism. Models involving slab break-off (e.g. Atherton & Ghani, 2002; Oliver et al., 2008; Neilson et al., 2009; Miles et al., 2016), regional transtension (Brown et al., 2008), lithospheric delamination (Miles et al., 2016), and Rheic Ocean subduction (McCarthy, 2013) have proved difficult to reconcile with all of the available data, in particular the geochronological evidence which suggests that magmatic activity in the SUT likely initiated sometime during the final convergence of Laurentia and Avalonia and continued for ∼40–50 Myr. In addition, the LDP is often excluded from the models used to describe the genesis of its neighbouring plutons in the SUT. Our new data and re-interpretation of previous studies allows us to provide valuable constraints on the geodynamical model and tectonic controls on magmatism in this part of the Caledonian orogenic belt. TECTONIC SETTING The Caledonian orogeny is widely acknowledged as a protracted and complex mountain building episode spanning ∼100 Myr from the early to mid-Paleozoic. In northern Britain and Ireland, the orogeny encompasses a series of discrete tectonometamorphic and tectonomagmatic phases associated with the continued subduction and closure of the Iapetus Ocean, accretion of island arcs, and collision of the Laurentia, Avalonia and Baltica continents (Mckerrow et al., 2000; Oliver et al., 2008; Bird et al., 2013). The final convergence of so called Eastern Avalonia and Laurentia during the Scandian phase (∼430–420Ma) occurred along the Iapetus Suture which now sits between the SUT of southern Scotland and Ireland, and the Lakesman-Leinster Terrane of northern England and Ireland (Fig. 1). Prehnite–Pumpellyite facies metamorphism (Merriman & Roberts, 2001; Stone, 2014) and a lack of regionally penetrative deformation fabrics (Kemp, 1987; Soper et al., 1992) around the Iapetus Suture zone suggest that Scandian convergence of Laurentia and Eastern Avalonia was relatively passive and strongly oblique, causing only minor amounts of regional uplift. Oliver et al. (2008) provided a detailed chronology of Caledonian granite emplacement across northern Britain (Fig. 1), with reference to five distinct tectonomagmatic events: (1) ∼470 Ma S-type granites emplaced after collision of the Laurentian margin and the Midland Valley Arc (Grampian Event); (2) ∼455 Ma S-type granites emplaced during decompression of the over-thickened Grampian orogenic belt; (3) ∼430 Ma I-type granites emplaced during the subduction of Iapetus oceanic lithosphere under the Laurentian margin; (4) ∼420 Ma I-type volcanic rocks erupted during slab roll-back after initial collision of Laurentia and Eastern Avalonia; (5) ∼410 Ma I-type and S-type granites emplaced during bilateral slab break-off and lithospheric delamination around the Iapetus Suture zone. The LDP is one of several major late Caledonian plutons of similar composition and age exposed just north of the Iapetus Suture zone within the SUT (Fig. 1). The tectonostratigraphic arrangement of the SUT, which consists predominantly of steeply dipping greywackes and shales, is interpreted as a NE–SW striking imbricate thrust belt that formed adjacent to the Laurentian continental margin after the accretion of the Midland Valley Arc to Laurentia during the Grampian orogeny (Mitchell & Mckerrow, 1975; Leggett et al., 1979; Needham, 2004). Although there are conflicting models for its tectonosedimentary evolution, recent consensus is that the SUT formed as a supra-subduction accretionary prism during the closure of the Iapetus Ocean prior to the ∼430–425 Ma docking of Laurentia and Eastern Avalonia (Phillips et al., 2003; Stone & Merriman, 2004; Oliver et al., 2008; Neilson et al., 2009; Bluck, 2013; Waldron et al., 2014). Previous work on the Loch Doon pluton The 130 km² LDP was initially mapped by the celebrated team of Geikie, Horne, Irvine and Peach in 1860 s–70 s with Teall (1899) later documenting and describing the wide petrological variation within the pluton. Gardiner & Reynolds (1932) utilized field mapping, petrographic and density analytical techniques to produce a remarkably detailed geological map that has changed in only minor details with subsequent studies (Fig. 2). A concentric zonal arrangement of plutonic rocks was recognized and attributed to be the product of three major pulses of increasingly felsic magma that were produced from a single parent magma that differentiated in a deep-seated magma chamber. The complex as a whole was interpreted as a laccolith intrusion based on the greater lateral extent of the thermal aureole to the east of the pluton towards and encompassing the Burnhead intrusion (Fig. 2). However, no explanation was provided as to why the rest of the pluton exhibits a consistently narrower thermal aureole in comparison. Rutledge (1950) and Oertel (1955) produced structural maps of the southwestern and northern regions of the pluton, respectively, which indicated a concentric symmetry in the pattern of ubiquitous planar plutonic foliations about the centre of the pluton. Oertel (1955) argued for granitization, in which solid-state metasomatic modification of metasedimentary host rocks formed the outer mafic units and continued metasomatic modification of the outer units formed inner units, to produce the concentric compositional zoning. The concentric foliations were interpreted to have formed in a solid mass during a ∼35% volume increase caused by the metasomatic changes. The dominant concentric mineral alignment occurred along conjugate shear planes during the volume increase and a weak, radial second alignment possibly formed by mineral growth along stress-controlled joints. Using whole-rock geochemical analyses, Ruddock (1969) returned to a magmatic petrogenesis for the pluton, supporting the evolution of the complex as a three-stage intrusion sourced from a single differentiating parent magma of intermediate composition. Brown et al. (1979) used whole-rock geochemistry, microprobe analysis, and fission track radiography to develop a two-stage in situ fractionation model that was consistent with the removal of plagioclase and pyroxene from a single monzodioritic parental magma. The Rb, Sr and O isotopic study by Halliday et al. (1980) concluded that the plutonic margins are more primitive ‘I-type’ granites (cf Chappell & White, 1974) compared to the more evolved ‘S-type’ granites in the centre of the pluton. Three isotopically distinct pulses of magma (dioritic, granodioritic and microgranitic) were identified and attributed to increasing degrees of contamination of a primitive magma by Lower Palaeozoic meta-sedimentary rocks. In a regional study, Blaxland et al. (1979) interpreted distinctly radiogenic feldspars and the high 207Pb/204Pb, 206Pb/204Pb, 208Pb/204Pb ratios seen within the SUT granitoids to be caused by melting of relatively primitive mantle-like protoliths, with an absence of older crustal source material. The trends in many of the whole-rock major and trace-element relationships for the LDP have led the most recent studies to conclude that the concentric compositional zoning within the pluton is the result of either a two- or three-stage in situ fractional crystallization process (Brown et al., 1979; Halliday et al., 1980; Tindle & Pearce, 1981; Stephens & Halliday, 1984). Tindle & Pearce (1981), using petrogenetic modelling and least square analyses of trace element concentrations, concluded that fractionation was controlled by plag–opx–cpx–bio during crystal settling and filter pressing of two distinct magmas. In their model, the first magma pulse, interpreted to be a quartz diorite intruded what is now the northwestern part of the pluton. It underwent in situ fractionation by crystal settling and filter pressing to form zones of diorite to quartz diorite. The second arriving pulse, interpreted to be granodiorite, also underwent in situ fractionation by crystal settling and filter pressing to form zones of quartz diorite to granite. The northwestern diorite to quartz diorite was envisaged to be at least partially crystallized before arrival of the second magma batch, supported by the presence of cognate inclusions of the former in the latter. Crystal settling was said to dominate after the initial emplacement of each magma pulse, with filter pressing and inward and upward migration of melts becoming more important with increasing degrees of fractionation. Rb–Sr mineral whole-rock geochronology techniques for the LDP have yielded ages between 407·2 ± 2·8 Ma and 409·1 ± 2·8 Ma, consistent with K–Ar techniques which give ages of 408 ± 4 Ma (Halliday et al., 1980). A zircon U–Pb age for the diorite was reported as 414 ± 1 Ma (Aftalion et al., 1984), and older ages for the granite (Halliday et al., 1979) of 422 Ma and 439 Ma were considered to indicate minor inherited components. A comparison of published ages for the LDP and other nearby SUT plutons is provided in Table 1. Table 1: A summary of published ages for SUT plutons Pluton Isotopic System Published Age (Ma) Loch Doon Rb–Sr 407-409 A K–Ar 408 A U–Pb 414 B Fleet Rb–Sr 392 A U–Pb 410–387 B,C,D,E Criffel Rb–Sr 397 A K–Ar 399–389 A U–Pb 410–391 B,E Carsphairn Rb–Sr 410 F Cheviot Pluton Rb–Sr 396 F Cheviot Lavas Rb–Sr 396 F Newry Igneous Complex Rb–Sr 399 G U–Pb 414–407 H Portencorkrie Rb–Sr 392 I U–Pb 395 J Pluton Isotopic System Published Age (Ma) Loch Doon Rb–Sr 407-409 A K–Ar 408 A U–Pb 414 B Fleet Rb–Sr 392 A U–Pb 410–387 B,C,D,E Criffel Rb–Sr 397 A K–Ar 399–389 A U–Pb 410–391 B,E Carsphairn Rb–Sr 410 F Cheviot Pluton Rb–Sr 396 F Cheviot Lavas Rb–Sr 396 F Newry Igneous Complex Rb–Sr 399 G U–Pb 414–407 H Portencorkrie Rb–Sr 392 I U–Pb 395 J References: AHalliday et al. (1980) BAftalion et al. (1984) CPidgeon & Aftalion (1978) DFettes & Henney (1999) EMiles et al. (2014) FThirlwall (1988). GO’Connor (1975). HCooper et al. (2016). IEvans (1995). JOliver et al. (2008). Table 1: A summary of published ages for SUT plutons Pluton Isotopic System Published Age (Ma) Loch Doon Rb–Sr 407-409 A K–Ar 408 A U–Pb 414 B Fleet Rb–Sr 392 A U–Pb 410–387 B,C,D,E Criffel Rb–Sr 397 A K–Ar 399–389 A U–Pb 410–391 B,E Carsphairn Rb–Sr 410 F Cheviot Pluton Rb–Sr 396 F Cheviot Lavas Rb–Sr 396 F Newry Igneous Complex Rb–Sr 399 G U–Pb 414–407 H Portencorkrie Rb–Sr 392 I U–Pb 395 J Pluton Isotopic System Published Age (Ma) Loch Doon Rb–Sr 407-409 A K–Ar 408 A U–Pb 414 B Fleet Rb–Sr 392 A U–Pb 410–387 B,C,D,E Criffel Rb–Sr 397 A K–Ar 399–389 A U–Pb 410–391 B,E Carsphairn Rb–Sr 410 F Cheviot Pluton Rb–Sr 396 F Cheviot Lavas Rb–Sr 396 F Newry Igneous Complex Rb–Sr 399 G U–Pb 414–407 H Portencorkrie Rb–Sr 392 I U–Pb 395 J References: AHalliday et al. (1980) BAftalion et al. (1984) CPidgeon & Aftalion (1978) DFettes & Henney (1999) EMiles et al. (2014) FThirlwall (1988). GO’Connor (1975). HCooper et al. (2016). IEvans (1995). JOliver et al. (2008). Subsequent to the flurry of 1970 s and 1980 s geochemical papers, culminating in Tindle & Pearce’s (1981) work, the LDP has been largely neglected. Thus, it is an excellent time for a re-evaluation of the evolution of this pluton. In particular, the previous lack of integration between geochemical and structural datasets, both within and external to the pluton imparted an inherent prejudice towards certain models. We employ recent advances in our understanding of magma genesis, granite processes and Caledonian geodynamics to shed new light on the evolution of this normally zoned pluton and other zoned plutons within the SUT. We begin by utilizing a series of summary maps that we have constructed from published work that show both host rock structures and internal characteristics of adjacent plutons in the SUT (e.g. Fig. 2). We then compare the characteristics of all of these to the LDP in order to search for common structural and geochemical patterns that will help to establish a generic emplacement and geochemical model for these plutons. RESULTS The publications noted above provide a wealth of descriptions on the petrology and geochemistry of the LDP. The results presented here focus on both our new studies and key elements of previous studies that we believe have alternative interpretations. The following detailed host rock and pluton field relationships were recorded during our mapping of ∼65 km2 of the southern half of the LDP and its aureole. Host rock stratigraphy and structure The LDP intrudes strongly deformed sequences of Mid–Upper Ordovician age greywackes and argillites of the Leadhills Supergroup and Moffat Shale Group (Fig. 2). These sequences are defined in a series of strike-parallel, tectono-lithostratigraphic packages extending the full length of the SUT. The contact between the Portpatrick Formation and Glenwarghen Formation is depositional, whereas the contacts between them and the neighbouring Kirkcolm and Shinnel Formations are along the major strike-parallel Leadhills and Fardingmullach faults, respectively (Fig. 2; Floyd, 1999). These faults are poorly exposed in the field area, but given their linear nature along the entire length of the SUT, the faulted contacts are interpreted to dip steeply. The laterally discontinuous Moffat Shale Group stratigraphy (Caradoc–Ashgill) usually crops out as inliers at the base of each succession immediately NW of the Leadhills and Fardingmullach faults, but is also exposed further up the successions along more minor strike-parallel faults. The Moffat Shale Group units are usually conformably overlain by the aforementioned lithostratigraphic successions. The Kirkcolm, Glenwarghen, Portpatrick, and Shinnel Formations are generally composed of greywackes, arenites and siltstones, and occasional conglomerates. Lithologies are predominantly quartzose, but are often rich in detrital pyroxene, amphibole, epidote and andesitic lithoclasts. Beds are laterally continuous and vary in thickness from centimetre to metre scale and frequently show primary sedimentary structures including bedding (S0), graded bedding, laminations, ripples, flute casts and scour marks. The Moffat Shale Group units are generally composed of graptolitic mudstones with local chert and rare basaltic pillow lavas. Beds are laterally discontinuous, centimetre to decimetre scale in thickness and frequently show primary sedimentary structures, including bedding (S0), laminations, and ripples. All lithostratigraphic bedding in the Loch Doon area dips at or near vertical and is commonly slightly overturned. Tilting and folding is said to have originally occurred along low angled thrust faults during frontal accretion at the southern toe of the Southern Uplands accretionary wedge, e.g. the Leadhills and Fardingmullach faults in the field area (Leggett et al., 1979; Needham, 2004). The faults were rotated more steeply as further material was accreted and underplated, rotating them to the vertical dips seen in present day outcrop. Regional deformation in the SUT was diachronous throughout Ashgillian–Wenlockian (∼449–427 Ma) subduction of the Iapetus Ocean and becomes progressively younger to the south. The regionally developed, weak, slaty foliation (S1) in argillites is absent in juxtaposed arenites and greywackes indicating that the regional deformation was relatively weak. This is consistent with the weak metamorphic signatures of the SUT, showing prehnite-pumpellyite facies/low anchizone conditions during burial in the accretionary wedge (Merriman & Roberts, 2001; Stone et al., 2012; Stone, 2014). Thermal and structural aureole Where pluton-host rock contacts are observed in the field, they are intrusive and their orientations can be constrained better than previously inferred. The pluton has an approximately vertical host rock contact constrained by relatively linear contacts across >400 m of local relief (Fig. 3f). In contrast to the shallow-dipping roof zone interpreted by Stephens & Halliday (1979) around the southern quartz diorite, traces of pluton contacts across local streams in this region are not sub-parallel to contours but instead must dip at least 60° outwards in this area. The steeply dipping pluton-host rock contact is consistent with an ∼1·5 km wide thermal aureole, which decreases in grade rapidly from a coarse grained andalusite–cordierite hornfels at the pluton-host rock contact to distal weakly metamorphosed shales and greywackes of regional prehnite-pumpellyite grade (Gardiner & Reynolds, 1932; Leake et al., 1981). Commonly, the chilled margin of the pluton and partially assimilated host rock (metre scale) have similar hornfelsic appearances in the field. To the east of the pluton, the thermal aureole extends laterally to encompass the compositionally zoned Burnhead intrusion (Fig. 2), leading some authors to suggest a shallow subsurface link between the two granitoids (Gardiner & Reynolds, 1932; Fettes & Henney, 1999). However, residual gravity anomaly surveys (Powell, 1970; Dawson et al., 1977; Stone et al., 2012) indicate steep pluton-host rock contacts to at least 10 km depth and show only a minor outward perturbation at the eastern contact of the LDP. Instead, we believe the relatively wide aureole to the east of the LDP is likely due to additional minor intrusives in this area which contributed to the thermal metamorphism of the host rock. For example, the granite intrusion at Mid Hill ∼500 m south of the Burnhead intrusion (Fig. 2), and the area between the LDP and the Burnhead intrusion also have a relatively high concentration of lamprophyre dykes (British Geological Survey, 1994). It is, therefore, likely that the LDP does not connect with the Burnhead intrusion at shallow levels, and instead their thermal aureoles are simply ‘overlapping’. A similar outward perturbation of the thermal aureole can be observed to the SW of the LDP where subsidiary granitoids and dykes are also observed (Fig. 2; British Geological Survey, 1994). In the thermal aureole, the contact metamorphic mineral assemblages and hornfels texture overprint the prehnite–pumpellyite regional metamorphic mineral assemblage and also overprint S1 cleavages. Trajectories of host rock regional structures around the LDP show varying degrees of deflection from regional strikes. Deflections of bedding, F1, host rock magnetic fabrics (Kafafy & Tarling, 1985), and lithostratigraphic bounding faults range from none at the northern pluton-host rock contact to ∼70° at 1·5 km from the contact in the southwest (Fig. 2). The larger deflections only occur locally, and the majority of the aureole has minor or no deflection of regional host rock structures. Deflections are consistent with petrographic textural observations that the pluton post-dates the formation of regional host rock structures. There is no evidence of appreciable thickening or thinning of regional tectonostratigraphic packages with decreasing distance to the pluton contact. There is also no evidence of faulted pluton-host rock contacts, nor evidence of faults within the host rock that are parallel to the margin of the pluton. Pluton structure and lithology The LDP has a length: width ratio of ∼2:1 to ∼4:1 and a prominent necking in its centre, giving it a distinctive hour-glass shape in map view (Fig. 2). It has a maximum length of 19·2 km and ranges in width from 4·6 km to 10·5 km, covering an area of 130 km2. Unlike many other plutons intruding the SUT, the long axis of the LDP strikes north–south, completely discordant to the regional NE–SW structural grain of the SUT. The pluton displays a spectrum of petrological compositions in a concentric arrangement, giving it a strong normal zonation pattern (Figs 2 and 3a–e). Averaged modal analyses for each rock type are presented in Table 2 using data from Tindle (1982). The central microgranite contains cordierite which occurs in cm-scale ophitic clots indicating its magmatic origin and has been pervasively pseudomorphed to biotite and muscovite (Tindle, 1982). The clots occupy up to 7% of the rock and have cm-scale haloes depleted in mafic constituents (Fig. 3e). Table 2: Averaged modal analyses of the main rock types in the LDP Rock Type n Plag Kspar Qz Hyp Aug Amph Bio Accessories Dio & Qz Dio 15 62 – 9 9 5 5 9 Ap Py Ilm Cp Zrn Tonalite 10 48 4 18 – 1 9 19 Ap Ilm Py Sph Zrn Granodiorite 30 43 11 24 – – 7 14 Ap Sph Ilm Aln Zrn Granite 11 37 22 26 – – 4 10 Ap Ilm Sph Aln Zrn Microgranite 4 38 25 30 – – – 4 Ap Ilm Zrn Cord Musc Mnz Rock Type n Plag Kspar Qz Hyp Aug Amph Bio Accessories Dio & Qz Dio 15 62 – 9 9 5 5 9 Ap Py Ilm Cp Zrn Tonalite 10 48 4 18 – 1 9 19 Ap Ilm Py Sph Zrn Granodiorite 30 43 11 24 – – 7 14 Ap Sph Ilm Aln Zrn Granite 11 37 22 26 – – 4 10 Ap Ilm Sph Aln Zrn Microgranite 4 38 25 30 – – – 4 Ap Ilm Zrn Cord Musc Mnz Accessories: Aln. Allanite; Ap, Apatite; Cp, Calcopyrite; Cord, Cordierite; Ilm, Ilmenite; Mnz, Monazite; Musc, Muscovite; Py, Pyrite; Sph, Sphene; Zrn, Zircon. Note amphibole occurs as a subsolidus replacement of primary pyroxene. Table 2: Averaged modal analyses of the main rock types in the LDP Rock Type n Plag Kspar Qz Hyp Aug Amph Bio Accessories Dio & Qz Dio 15 62 – 9 9 5 5 9 Ap Py Ilm Cp Zrn Tonalite 10 48 4 18 – 1 9 19 Ap Ilm Py Sph Zrn Granodiorite 30 43 11 24 – – 7 14 Ap Sph Ilm Aln Zrn Granite 11 37 22 26 – – 4 10 Ap Ilm Sph Aln Zrn Microgranite 4 38 25 30 – – – 4 Ap Ilm Zrn Cord Musc Mnz Rock Type n Plag Kspar Qz Hyp Aug Amph Bio Accessories Dio & Qz Dio 15 62 – 9 9 5 5 9 Ap Py Ilm Cp Zrn Tonalite 10 48 4 18 – 1 9 19 Ap Ilm Py Sph Zrn Granodiorite 30 43 11 24 – – 7 14 Ap Sph Ilm Aln Zrn Granite 11 37 22 26 – – 4 10 Ap Ilm Sph Aln Zrn Microgranite 4 38 25 30 – – – 4 Ap Ilm Zrn Cord Musc Mnz Accessories: Aln. Allanite; Ap, Apatite; Cp, Calcopyrite; Cord, Cordierite; Ilm, Ilmenite; Mnz, Monazite; Musc, Muscovite; Py, Pyrite; Sph, Sphene; Zrn, Zircon. Note amphibole occurs as a subsolidus replacement of primary pyroxene. There are two types of dyke that cross-cut the main LDP intrusive units. The most common type is equigranular, leucogranitic with an average composition of 36% plagioclase, 32% quartz, and 32% orthoclase and accessory muscovite, xenotime, monazite, and tourmaline. The narrower leucogranitic dykes (cm-scale) are typically fine grained compared to the wider dykes (m-scale), which are typically medium grained. They are particularly common throughout the granitic and granodiorite units. They predominantly strike radially from the centre of the pluton with steep to vertical dips, although they are sometimes observed with shallow dips and with random strike trends. The less frequent type of dyke is porphyritic with a fine-grained groundmass containing medium grained hornblende, biotite and plagioclase phenocrysts. These dykes are compositionally diverse and are collectively referred to in the literature as lamprophyres. They occur throughout all plutonic units, but are more commonly found in the host rock metasediments. They are too infrequent in the mapped area to reliably identify patterns in strike in the plutonic units. Pluton and dyke contacts All plutonic units become finer grained within close proximity (m) to the pluton-host rock contact, indicating a degree of rapid cooling. The diorite and quartz diorites in particular exhibit wider chilled margins (<100 m) compared to the granodiorite and tonalite (<10 m). Importantly, the internal compositional contacts between individual plutonic units are mostly gradational and it is rare to find examples in the field that show the juxtaposition along sharp contacts of voluminous igneous rock types. As such, the contacts drawn on Figure 2 are constrained by the 70 modal analyses of Tindle (1982) and by comprehensive field mapping. In the NW part of the LDP, the contact between the diorite and quartz diorite cannot be observed at the metre scale, but is gradational over 100 s of metres. Given the local topographic relief across the contact is <50 m, we cannot make any interpretations on the likely dip of this contact. The contact between the northwestern quartz diorite and tonalite is also gradational; however, it is strongly discordant to the ∼150 m topographic relief, indicating that it is a relatively steep contact. The contact between the tonalite to granitic units in the north of the pluton is gradational over 100 s of m, and again shows discordance to the ∼200 m topographic relief, indicating a steep contact. In the southern part of the pluton, the contact between the quartz diorite and tonalite occurs laterally over a much smaller distance (10 s of m), which then also grades into granodiorite over a similar distance. Here, these contacts clearly cut across the ∼140 m topographic relief indicating a steep contact. This further negates the possibility of a roof zone in this area of the pluton as suggested by previous studies (Stephens & Halliday, 1979; Tindle, 1982). In this area veins of granodiorite (cm-scale) are commonly seen interfingering with the quartz diorite. The contacts between the granodiorite and granite units are the most gradational of all, occurring over several 100 s of m. Significantly, we map the granite-microgranite contact differently than Tindle & Pearce (1981) and believe that the microgranite has an area of ∼1 km2, half the area previously mapped. The contact is identified by a marked inward decrease in grain size over ∼10–20 m, which is coincident with the occurrence and inward increase of altered cordierite clots. The identification of this relatively sharp contact and reduced area of the inner microgranite is the most important deviation from previous mapping. Chapman & Cameron (1978) recorded similar sharp igneous contacts between the granite and microgranite, although these are said to be somewhat spatially limited in outcrop. One example of clear sharp contacts was found in the March Burn (Fig. 3g) near the pluton-host rock contact where three magma types are preserved in a single outcrop. Here, a fine grained mafic enclave is hosted by a coarse grained granodiorite, which in turn is intensely intruded by coarse grained granite. The granite is isolated, surrounded by granodiorite and cannot be linked to the granite at the centre of the pluton. The leucogranite dykes are completely discordant to the main concentric petrographic zonation, indicating that they post-date the main plutonic phase, although there are subtleties to this. The wider (dm- to m-scale), medium grained dykes commonly show irregular cuspate contacts with the host plutonic rock, with no evidence of chilled or baked margins. Conversely, the narrow (cm-scale), fine grained dykes commonly display sharp contacts with the host plutonic rock and also show cooling joints perpendicular to the contact. Occasional fine-grained conjugate leucogranitic dykes are observed in the granodiorite. Given the range of observed relationships, it is concluded that the intrusion of leucogranitic dykes was a protracted event, occurring progressively throughout the cooling of the pluton. The less frequently observed lamprophyre dykes typically show relatively sharp contacts and weakly chilled margins. Xenoliths and enclaves Xenoliths of metasedimentary host rock are present throughout all the main plutonic units and are easily identified from host plutonic rock by their distinct darker colour, finer grain size, and sometimes preservation of relic sedimentary laminations. They locally occupy up to 15% of any given outcrop close to the pluton-host rock contact and ≪2% at distances more than 200 m from the contact. They become increasingly scarce, but not completely absent, towards the centre of the pluton. Xenoliths typically have a hornfelsic texture (sometimes with fine grained centres and coarser grained margins), range between 1–50 cm in length, and have a range of axial ratios and angularity. Xenoliths with relatively high axial ratios commonly have steeply plunging long axes, although it is not uncommon to see xenoliths of different sizes, axial ratios and orientations within individual exposures (Fig. 3h). Xenoliths of host rock material up to 400 m in length have been observed in the tonalite close to the northern pluton-host rock contact (British Geological Survey, 1994) although xenoliths of this size are rare elsewhere in the pluton. Co-magmatic microgranitoid enclaves are present throughout all phases of the pluton and are usually identified by their more mafic composition and medium grain size compared to their host plutonic rock. They are identified from xenoliths by their coarser grain size and plutonic textures (Fig. 3h–i). They typically occupy <1% of any given outcrop, but this increases towards the major compositional contacts, particularly around the southern quartz diorite–tonalite–granodiorite contact zone where they locally occupy up to 10% of any given outcrop. Enclaves have a range of sizes, from mm- to dm-scale and typically, but not exclusively, have high axial ratios. Just like xenoliths, those with high axial ratios commonly have a steeply oriented plunge direction, although it is not uncommon to see enclaves of different sizes, axial ratios and orientations within individual exposures. Commonly, enclaves have very similar compositions and textures to their hosts and are observed with very gradational host plutonic rock contacts, suggesting a locally high degree of mixing (Fig. 3i). Magmatic fabrics The principal plutonic fabric is a variably developed, steeply dipping foliation, which is ubiquitous throughout all the plutonic units. The foliation is predominantly defined by the alignment of euhedral feldspar phenocrysts, euhedral–subhedral pyroxene (and secondary amphibole), biotite crystals and microgranitoid enclaves. Commonly, minerals defining the foliation also define a steeply plunging mineral lineation, although a lack of three-dimensional exposures limits their routine analysis. Mineral petrographic analysis shows only minor evidence of crystal–plastic strain, i.e. mica kink bands, deformation twinning, weak subgrain development and undulose extinction in the quartz groundmass, suggesting the pluton underwent only relatively mild near- to sub-solidus recrystallization. The combination of euhedral–subhedral crystal habits, concentric mineral zonation, growth twinning, and a randomly orientated fine-grained groundmass suggests that crystal growth and alignment in foliation planes progressed under melt-present conditions (e.g. Paterson et al., 1998) and that the textures are, therefore, magmatic in origin. The lack of concordant foliations within xenoliths of metasedimentary host rock and cognate inclusions (fully crystallized blocks of earlier intrusive units) further precludes fabric development under sub-solidus conditions. Our detailed field mapping of magmatic foliations in the southern half of the pluton has been combined with the work of Rutledge (1950) and Oertel (1955) to produce an updated structural map of the pluton (Fig. 2). Several significant trends are present within the mapped magmatic foliations: Foliations trend in a roughly concentric manner about the centre of the pluton. Importantly however, they consistently show a range of angular discordance with the pluton-host rock contact (Fig. 2). There is particularly high discordance between magmatic foliations and the host rock contact in the diorite and quartz diorites in the northwest and south of the pluton. There is also moderate discordance in the granodiorite and tonalite at the ENE host rock contact. Foliations overprint all of the main internal compositional contacts at a range of oblique angles (Fig. 2). This overprinting relationship indicates the magmatic foliations must post-date the juxtaposition of the compositional units and that significant viscosity contrasts could not have existed between the main internal plutonic units at the time of fabric formation. There are two exceptions: (i) Foliations refract through the granodiorite–tonalite–quartz diorite contact zone in the southern part of the pluton, to become sub-parallel to the quartz diorite–tonalite contact. This indicates more contrasting viscosities across these units, consistent with the sharper compositional contacts and the higher degree of veining and stoping of cognate inclusions observed in this area. (ii) Foliations in the granodiorite and granite tend to deflect around the central microgranitic unit. The intensity of the foliations varies across multiple scales (intensity classification based on Miller & Paterson ). (i) On a plutonic scale, foliation intensity decreases inwards, from very strong to moderate crystal alignment at the margins of the pluton to weak in the central granite and microgranite (Fig. 3a–e). The central granite and microgranite in particular show weak foliations that commonly define two foliations with different strikes at single exposures. (ii) On an inter-exposure scale, foliation intensity varies considerably across zones 100s of m wide. The contacts between these strong–weak intensity zones are parallel to magmatic foliations and thus, overprint the internal compositional contact. No textural or compositional relationships that correlate with intensity were identified. Inter-exposure variability is most pronounced in the granodiorite and granite north of Loch Dee. (iii) On an intra-exposure scale, foliation intensity varies considerably between weak–intermediate on a dm- to m-scale. Leucoganitic dykes exhibit a range of relationships with the host rock magmatic foliation. For example, the fine-grained dykes typically overprint and, therefore, must post-date the formation of the host plutonic rock foliation, whereas the coarser grained dykes with more gradational contacts typically have their contacts overprinted by host plutonic rock foliations. Schlieren structures commonly, but not always, have their long axes parallel to the magmatic foliations and the schlieren-host magma contacts are always overprinted by magmatic foliations, even where their contacts are parallel to their host foliation. Whilst xenolith long axes commonly, but not exclusively, rotate parallel to the vertical foliation trends, their internal structures are completely discordant to structures in their host plutonic rocks. Significantly, the local host plutonic rock foliations are discordant to the granular hornfelsic xenolith textures (e.g. Fig. 3h). This further supports a magmatic, rather than subsolidus origin of the host foliation, and that high viscosity contrasts must have prevented the host magmatic foliation from overprinting the xenolith contacts. Enclave–host plutonic rock foliation relationships are complex, mainly because enclaves have a range of compositional contrasts to their host plutonic rocks, even within individual exposures. As such, they have a range of appearances from highly discordant to highly concordant enclaves. The enclaves with strong compositional contrast to host plutonic rocks have discordant host plutonic rock relationships comparable with the xenoliths and cognate inclusions, whereas those with low compositional contrast show increasingly concordant structures. For example, some enclaves have grain sizes similar to their host plutonic rock, commonly have long axes parallel to the host foliation (e.g. Fig. 3i), and show concordant internal-host plutonic rock foliations that overprint the enclave-host plutonic rock contacts. The key implication here is that the magmatic foliation post-dates the final trapping of these enclaves, and that concordant foliations could only form in enclaves with low viscosity contrasts. Magnetic susceptibility analysis Point magnetic susceptibility measurements were determined for plutonic rock hand samples from the southern half of the LDP, using a Bartington MS2E high resolution point sensor following the procedures outlined in the Supplementary Data A1 (all supplementary data are available for download at http://www.petrology.oxfordjournals.org). All magnetic susceptibility results are also presented in the Supplementary Data A2. The quartz diorite and tonalite facies show average point susceptibilities of 139·6x10-5 SI (n = 60) indicating ferrimagnetic (magnetite) mineral assemblages. Granodiorite, granite and microgranite facies show average susceptibilities of 22·5x10-5 SI (n = 90), 13·4x10-5 SI (n = 171) and 2·8x10-5 SI (n = 15), respectively, indicating paramagnetic (ilmenite, biotite, amphibole) mineral assemblages. Shinnel, Portpatrick, Glenwarghen, Kirkcolm and Glenlee formation greywackes record average point susceptibilities of 35·0x10-5 SI (Floyd, 1996). Our new data are consistent with those of Kafafy & Tarling (1985) and Piper (2007) showing a marked susceptibility decrease towards the centre of the pluton, mirroring both the concentric petrographic and geochemical zonation trends and the decrease in fabric intensities. Limited anisotropy of magnetic susceptibility (AMS) analyses show that plutonic rocks have mostly oblate susceptibility ellipsoids and low degrees of magnetic anisotropy (Kafafy & Tarling, 1985; Piper, 2007). The magnetic foliations strike and dip parallel to the magmatic foliations described earlier. The inner aureole rocks also have strongly oblate susceptibility ellipsoids and show higher degrees of magnetic anisotropy compared to the plutonic units. These AMS foliations tend to strike parallel with the pluton-host rock contact. In the outer aureole, the susceptibility ellipsoids become more varied, from oblate to prolate, and have much higher degrees of magnetic anisotropy compared to the inner aureole and the plutonic units. LA-ICP-MS U–Pb zircon geochronology Zircons from samples of one quartz diorite (LD-01), one granite (LD-50), one microgranite (LD-21) and one cross-cutting lamprophyre dyke (LD-51) were separated for dating. Zircons from all samples were picked from a population of colourless to yellow, euhedral, stubby to acicular crystals, free of both cracks and inclusions. Cathodoluminesence (CL) imaging of zircons from all samples shows complex magmatic, oscillatory and concentric zoning from core to rim (see Supplementary Data A3). Commonly, grains contain oscillatory zoned cores with subtle irregular external margins rimmed by younger oscillatory and concentric zones. Some zircon crystals show relatively small, dark, well-rounded, CL-homogenous cores rimmed by younger oscillatory and concentric zones. Zircons were analysed from each sample using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS; see Supplementary Data A1 for instrument parameters) at the Arizona LaserChron Center, USA. We undertook two sets of LA-ICP-MS dating, one using Faraday Cup collection with a beam size of 40 µm only on samples LD-01 and LD-21, the other using Ion Counter collection with a smaller beam size of 20 µm on all four samples LD-01, LD-50, LD-21, and LD-51. The Ion Counter results gave younger ages in both samples (Fig. 4). It must also be noted that the Faraday Cup and Ion Counter analyses used different subsets of zircon from the same rock samples and, therefore, careful age comparisons must be made between the two analyses. We present only those results that are between 90% to 110% concordance (Fig. 4), but raw data for all analyses are provided in Supplementary Data A2. The southern quartz diorite (LD-01) yielded 206Pb/238 U zircon ages between 426 ± 6 Ma and 412 ± 8 Ma (n = 21) using the Faraday Cups and between 423 ± 8 Ma and 393 ± 6 Ma (n = 12) using the Ion Counter. The granite (LD-50) yielded 206Pb/238 U zircon ages between 439 ± 37 Ma and 412 ± 5 Ma (n = 16) using the Ion Counter. The cordierite microgranite (LD-21) yielded 206Pb/238 U zircon ages between 444 ± 9 Ma and 422 ± 2 Ma (n = 7) using the Faraday Cups and between 433 ± 5 Ma and 390 ± 3 Ma (n = 5) using the Ion Counter, excluding two discordant grains which yielded ages of ∼510 Ma. The lamprophyre dyke (LD-51) yielded 206Pb/238 U zircon ages between 499 ± 7 Ma and 389 ± 10 Ma (n = 13) using the Ion Counter, excluding two discordant grains which yielded ages of 1098 ± 7 Ma and 1079 ± 12 Ma. The smaller Ion Counter spots consistently gave younger ages as zircon rims were approached. A combined age distribution plot is presented in Figure 4 and we interpret these analyses in the Discussion section. Fig. 4. View largeDownload slide Combined zircon U–Pb age distribution plot for four compositional zones of the Loch Doon pluton. Weighted mean ages of the youngest analyses for individual rock types are shown by green lines with the 2σ analytical errors highlighted in grey. Faraday Cup analyses have black borders and Ion Counter analyses have no borders. Fig. 4. View largeDownload slide Combined zircon U–Pb age distribution plot for four compositional zones of the Loch Doon pluton. Weighted mean ages of the youngest analyses for individual rock types are shown by green lines with the 2σ analytical errors highlighted in grey. Faraday Cup analyses have black borders and Ion Counter analyses have no borders. Whole-rock geochemistry In addition to eight new whole-rock major and trace element analyses of rocks from the southern half of the LDP, we present a synthesis of published geochemical data available for the pluton (163 samples; see Supplementary Data A2 for data and list of references). The Loch Doon plutonic rocks show a broad continuum from high-K to shoshonitic series compositions on an SiO2–K2O plot (Peccerillo & Taylor, 1976), plot within the sub-alkaline series on a TAS plot (Cox et al., 1979), within the calc-alkaline series on an AFM plot (Irvine & Baragar, 1971), and have metaluminous to peraluminous compositions on an A/CNK–A/NK plot (Shand, 1943), with SiO2 varying between 50–76% for the dioritic to granitic rocks (Fig. 5a). Fig. 5. View largeDownload slide View largeDownload slide Summary geochemical plots for the Loch Doon pluton, Southern Uplands Terrane plutons, volcanic rocks, and regional dyke swarm. (a) Selected classification diagrams. (b) Selected major element v. SiO2. (c) Selected trace element v. SiO2 and chondrite-normalised REE patterns. (d) Selected element ratio diagrams. Fig. 5. View largeDownload slide View largeDownload slide Summary geochemical plots for the Loch Doon pluton, Southern Uplands Terrane plutons, volcanic rocks, and regional dyke swarm. (a) Selected classification diagrams. (b) Selected major element v. SiO2. (c) Selected trace element v. SiO2 and chondrite-normalised REE patterns. (d) Selected element ratio diagrams. Whole-rock major elements in the LDP (Fig. 5b) define weakly curved decreasing trends with SiO2 (e.g. TiO2, MgO, CaO, P2O5, Al2O3 and FeOt), whereas K2O increases and Na2O shows a generally increasing, but scattered trend. These trends reflect the relative decrease in the abundance of mafic minerals and increase in felsic minerals from the margin to the interior of the LDP, mimicking the normal zonation pattern of the pluton. In general, the rocks show greater spread in major element concentrations below 55–57% SiO2 and the scatter decreases markedly with increasing SiO2. Whole-rock plutonic trace element signatures are complex, but show the same general trends exhibited by the major elements, with a reduction in scatter with increasing SiO2 and a particular spread in rocks with <57–60% SiO2. A selection of trace elements plotted v. SiO2 are presented in Figure 5c; Sr, Ba, Ni, Cr, Cu and Sc show a negative correlation with SiO2. A strong positive correlation is apparent for Rb and Pb with increasing SiO2. A somewhat weaker positive correlation is seen with U, Th, and Ta. Concentrations of Zr, Ba, Nb, Hf and Y show curved relationships with an initial increase with increasing SiO2, followed by a gradual or rapid decrease, indicating the saturation and thus crystallization of characteristic minerals (e.g. Zr in zircon, Ba in K-feldspar and biotite). The overall decrease in Cr, Y, Sc, Ni, Mn, Fe, and Ca is indicative of the depletion of pyroxene with increased SiO2, the decrease in Cr and Ti in the felsic rocks is related to the increasing removal of pyroxene and ilmenite respectively. Cr, Sc, Rb, Ba, Ni, Mn, Y, and Nb are compatible in biotite, which like the other mafic minerals decreases towards the interior of the LDP. Decreasing trends for Sr and Ba and increasing Rb are compatible with the relative decrease of plagioclase and increase of K-feldspar. The whole-rock REE profiles for the LDP rocks (Fig. 5c) are consistent and show fairly flat and narrow ranges with 20–200x enrichment of light REEs and 2–20x enrichment of heavier REEs relative to chondritic values (Nakamura, 1974). Only minor negative Eu anomalies are observed. The P2O5/K2O v. MgO (or SiO2) plot (Fig. 5d) is often used to discriminate between dominantly mantle derived (>0·4 P2O5/K2O) versus crust-mantle mixed magmas (Miller et al., 2000). With only a few exceptions, all the LDP samples fall below the 0·4 ratio, indicating that the magmas are derived from a mantle source mixed with various proportions of crustal-derived magmas. According to Profeta et al. (2015), the Sr/Y v. SiO2 plot can be used to determine crustal thickness during magmatism. The data shown in Figure 5d have been filtered to include only samples with SiO2 contents of 55–68 wt %, MgO contents <4 wt %, Rb/Sr <0·2 and Rb/Sr >0·05 to discard mantle-derived rocks or rocks formed by melting of pre-existing metasedimentary rocks. The few LDP samples that match these criteria all plot at a low Sr/Y ratio of about 20, suggesting Moho depths during magma generation of ∼30 km. Whole-rock Sri, εNd, δ18O and Pb isotopes We present a synthesis of published isotopic analyses for the LDP (83 samples; see Supplementary Data A2 for data and list of references). Like the petrological variation in the LDP, the isotopic data also display a crude concentric zonation in map view as they become increasingly more evolved towards the LDP interior. Average Sri has been calculated at 400 Ma for diorites (0·70476), quartz diorites (0·70487), tonalites (0·70519), granodiorites (0·70549), granites (0·70581) and microgranites (0·70748). Initial 87Sr/86Sr v. 1/Sr concentration diagrams (Fig. 6a) show linear relationships within the diorite, quartz diorite, tonalite and granodiorite facies of the pluton, whereas the granite and microgranite facies have a distinctly different linear relationship and show Sr depletion relative to the more mafic facies. With increased SiO2 contents toward the interior plutonic units, the range of isotopic compositions is also increased. Average εNd has been calculated at 400 Ma (Fig. 6b) for the diorites (-2·1), quartz diorites (-1·9), tonalites (-1·3), granodiorites (-2·6), granites (-3·0) and microgranites (-2·8). Average δ18O (‰ SMOW, Fig. 6c) has been calculated for quartz diorites (8·3), tonalites (8·1) and granites (10·3). Limited 207Pb/204Pb and 206Pb/204Pb analyses for plutonic rocks range between 15·52–15·57 and 18·0–18·8 respectively (see Supplementary Data A4). Fig. 6. View largeDownload slide Summary isotopic data for the Loch Doon pluton, Southern Uplands Terrane plutons, volcanic rocks, and regional dyke swarm (calculated at interpreted final crystallization age) and Southern Uplands Terrane host rocks (calculated at 400 Ma). (a) Whole-rock Sriv. 1/Sr. (b) εNd v. whole-rock Sri. (c) δ18O v. whole-rock Sri. Isotopic data recalculated after authors listed in Supplementary Data A2. Fig. 6. View largeDownload slide Summary isotopic data for the Loch Doon pluton, Southern Uplands Terrane plutons, volcanic rocks, and regional dyke swarm (calculated at interpreted final crystallization age) and Southern Uplands Terrane host rocks (calculated at 400 Ma). (a) Whole-rock Sriv. 1/Sr. (b) εNd v. whole-rock Sri. (c) δ18O v. whole-rock Sri. Isotopic data recalculated after authors listed in Supplementary Data A2. OTHER ZONED PLUTONS IN THE SOUTHERN UPLANDS TERRANE As noted in the ‘Previous Work’ section, there are several other volumetrically significant, zoned late Caledonian plutons exposed within the SUT (Fig. 1) and whilst our research has focused on the LDP, we provide a brief analysis of its neighboring plutons, given their apparent spatio-temporal correlation. Maps showing both internal and host rock characteristics of other SUT plutons are shown in Figure 7a–f and we draw attention to key characteristics that are shared with the LDP: Fig. 7. View largeDownload slide View largeDownload slide View largeDownload slide Summary geological maps of zoned plutons in the Southern Uplands Terrane. See Figure 2 for extended key, and Figure 4 and Supplementary Data A2 for age data. All maps based upon BGS DiGMap-50, with permission from the British Geological Survey. (a) The Fleet pluton (map after: Parslow, 1968; Cook & Weir, 1979). (b) The Criffel and Bengairn plutons (map after: Phillips, 1956; Leake & Cooper, 1983; Stephens et al., 1985; Courrioux, 1987). (c) The Newry Igneous Complex (map after: Reynolds, 1934, 1936, 1943; Neeson, 1984; Anderson et al., 2016; Cooper et al., 2016). (d) The Carsphairn pluton (map after: Deer, 1934; Tindle et al., 1988). (e) The Cheviot pluton (map after: Jhingran & Tomkeieff, 1942; Lee, 1982; Al-Hafdh, 1985). (f) The Portencorkrie pluton (map after: Holgate, 1943; Gaskarth, 1995). Fig. 7. View largeDownload slide View largeDownload slide View largeDownload slide Summary geological maps of zoned plutons in the Southern Uplands Terrane. See Figure 2 for extended key, and Figure 4 and Supplementary Data A2 for age data. All maps based upon BGS DiGMap-50, with permission from the British Geological Survey. (a) The Fleet pluton (map after: Parslow, 1968; Cook & Weir, 1979). (b) The Criffel and Bengairn plutons (map after: Phillips, 1956; Leake & Cooper, 1983; Stephens et al., 1985; Courrioux, 1987). (c) The Newry Igneous Complex (map after: Reynolds, 1934, 1936, 1943; Neeson, 1984; Anderson et al., 2016; Cooper et al., 2016). (d) The Carsphairn pluton (map after: Deer, 1934; Tindle et al., 1988). (e) The Cheviot pluton (map after: Jhingran & Tomkeieff, 1942; Lee, 1982; Al-Hafdh, 1985). (f) The Portencorkrie pluton (map after: Holgate, 1943; Gaskarth, 1995). All plutons were emplaced into the NE–SW striking metasediments of the Southern Uplands accretionary prism and commonly have overlapping Rb–Sr and, or, U–Pb ages between ∼430–382 Ma (see Discussion section for age interpretation, and Supplementary Data A2 for a collection of published age determinations and a list of references). The size of individual plutons ranges from the 189 km2 Criffel pluton to the 5 km2 Portencorkrie pluton. The plutons have predominantly irregular elliptical shapes in map view, with only the Cheviot, Bengairn and Carsphairn plutons displaying non-elliptical shapes. The long axes of the elliptical plutons are generally parallel to the NE–SW regional host rock fabric. The LDP is the exception and strikes discordant to the regional host rock fabric. Pluton-host rock contacts are steep, sharp, largely discordant and commonly cross-cut host rock structures, including bedding, folds, and regional tectonostratigraphic packages and their package-bounding faults. The pluton contacts are almost entirely intrusive, although local examples in the Cairnsmore of Fleet (hereafter referred to as Fleet), Criffel, Bengairn, Portencorkrie and the Cheviot show examples of local faulted contacts. In the few areas where the host rocks are concordant to pluton contacts, the structural aureoles show only narrow zones of host rock deflection that only extend for short distances along pluton margins. Thermal aureoles are generally ∼1–2 km wide, consistent with the relatively steep pluton–host rock contact dips seen in the field and inferred from residual gravity surveys (Powell, 1970; Dawson et al., 1977; Stone et al., 2012). The extension of the thermal aureole to the SW of the Fleet pluton suggests a minor lateral subsurface extension of this (or a buried) pluton in this region (Fig. 7a). Concentric geochemical zonation is remarkable in each pluton. In general, plutons are normally zoned from more basic margins (diorite to quartz diorites) to more evolved centres (granodiorite to granite). The central part of the Newry Igneous Complex, the Newry pluton (Fig. 7c), is the exception, showing minor reverse zonation (bio–granodiorite to hbl–granodiorite; Cooper & Johnston, 2004). Internal compositional contacts are almost always gradational and usually inferred on maps based on laboratory mineral analyses. There are exceptions where sharp internal contacts occur, in the Carsphairn (Fig. 7d) and Newry Igneous Complex (Fig. 7c), but these are very localized in comparison to the pervasive gradational contacts. Magmatic foliations are observed in all of the plutons and have steeply dipping foliations that decrease in intensity towards the pluton centres. Foliations show a range of discordance and concordance with the pluton–host rock contact and typically overprint internal compositional contacts. Although no foliation maps exist for the Carsphairn pluton, magmatic foliations were recorded by Deer (1934). Also, although foliations have not been documented for the Cheviot pluton, we believe thin section and hand specimen samples collected by Al-Hafdh (1985) show magmatic textures and foliations very similar to those observed in all of the plutons we present here. The Cheviot pluton has some intricacies not exhibited in the other plutons. Most notably is that the pluton intrudes into andesitic–rhyolitic lavas of the same age (396 Ma; Thirlwall, 1988). The pluton thermally overprints the lavas to varying distances, and shows both highly discordant (faulted) and highly concordant pluton–host rock contact relationships (Fig. 7e). We have examined a collection of all the available geochemical data for the SUT plutons and regional lamprophyre dyke swarm (>750 samples, see Supplementary Data A2 for data and list of references). The key trends are: (i) All SUT plutons show overlapping compositions from high-K to shoshonitic, calc-alkaline, metaluminous to peraluminous (Fig. 5a) with SiO2 varying between 41–80% SiO2; (ii) whole-rock major and trace element trends in the SUT plutons show a reduction in scatter with increasing SiO2 and a particular spread in rocks with <57–60% SiO2, consistent with trends in the LDP (Fig. 5b–c); (iii) plutonic whole-rock major oxides (Fig. 5b) define compatible linear trends with SiO2 and TiO2, MgO, CaO, P2O5 and FeOt, whereas K2O defines an incompatible trend with SiO2. There is marked scatter in Na2O and K2O at all SiO2 concentrations and both Al2O3 and Na2O show curved (but relatively flat) trends with SiO2; (iv) plutonic whole-rock trace elements (Fig. 5c) show a strong negative correlation with SiO2 and Sr, Ba, Y, and Sc, with a somewhat weaker negative correlation with Cr and Ni. A strong positive correlation is observed with SiO2 and Rb, and a curved relationships with Zr; (v) the REE distribution for all SUT plutons (Fig. 5c) shows a wide-ranging pattern from 5–450x enrichment of the light REE relative to chondrite and chondritic to 450 x chondritic values for the heavy REE; (vi) on a P2O5/K2O ratio plot (Fig. 5d), the SUT plutons are similar to the LDP at ratios <0·4, however, the Sr/Y v. SiO2 plot indicates much higher Sr/Y values at up to 55–60 compared to the LDP; (vii) the regional lamprophyre dykes are broadly comparable with the LDP and SUT plutons (Fig. 5a–d); however, although the composition of the dykes overlaps the range of the SUT pluton data, there is particularly wide spread in whole-rock dyke compositions at the primitive end of plutonic data array. We have also examined a collection of all available isotopic data for SUT plutons (273 samples, see Supplementary Data A2 for data and list of references). The key trends are: (i) whole-rock Sri ratios range between 0·7030–0·8058 and all plutons show a cluster of relatively low Sri ratios with an associated increase in 1/Sr (Fig. 6a); (ii) the LDP, Fleet and Criffel plutons also show progression to higher whole-rock Sri ratios with an associated decrease in 1/Sr; (iii) the felsic plutonic units have higher whole-rock Sri ratios than the more mafic units, independent of which pluton they are in; (iv) plutons show relatively tightly clustered whole-rock εNd values between +0·33 and -5·75 (Fig. 6b), and the evolved plutonic units have greater εNd values than the more primitive units, independent of which pluton they are in; (v) the whole-rock δ18O analyses show approximately linear positive correlation with Sri, between 6·90–11·89‰ SMOW (Fig. 6c); (vi) the Cheviot pluton falls outside of this correlation with δ18O between 4·71–4·78‰ SMOW; and (vii) the plutons show overlapping whole-rock 207Pb/204Pb and 206Pb/204Pb ratios and whilst they also overlap to some extent with host rock analyses, they are on average distinctly more radiogenic (see Supplementary Data A4). SUT host rock compositions To better constrain potential contamination of magmas by host rock assimilation, we synthesized all available host rock chemical data: we discuss these using igneous terms for ease of comparison. Host rock compositions (204 samples, see Supplementary Data A2 for data and list of references) have more scattered, broadly calc-alkaline to shoshonitic, calc-alkaline series, metaluminous to mostly peraluminous compositions with SiO2 between 51–79% (see plots in Supplementary Data A5). Major elements in the host rocks show more scatter, but broadly reflect the SUT plutonic rocks in loosely linear (e.g. K2O, TiO2, MgO, CaO, P2O5 and FeOt) and curved (e.g. Al2O3 and Na2O) trends v. SiO2. Whole-rock trace element signatures in the host rocks are similarly scattered in comparison with the SUT plutons. There is a weak negative correlation between SiO2 and Sr, Ni and Ce, and a weak positive correlation between SiO2 and Rb and Zr. Concentrations of Ba and Cr show weakly curved relationships and there are no discernable relationships between SiO2 and La and Y. Whole-rock REE profiles for host rocks are consistent with the LDP showing relative enrichment of light REE compared to heavy REE, and minor Eu anomalies. Host rock isotope analyses (98 samples, see Supplementary Data A2) show a wide range of whole-rock Sri with an average of 0·70860, and average whole-rock εNd of -6·3, δ18O of 12·5‰ SMOW, and 207Pb/204Pb and 206Pb/204Pb ratios between 15·45–15·58 and 17·7–18·4 respectively (Fig. 6 and Supplementary Data A4). DISCUSSION We now synthesize the new and published data and propose an updated model for the origin, emplacement and evolution of the LDP. We also explore and test our model against the other nearby plutons in the Southern Uplands Terrane. Age of the Loch Doon and Southern Uplands Terrane plutons The relative ages of the main LDP units from field relationships are constrained by field evidence of cross-cutting relationships. Field evidence indicates the diorites and quartz diorites were the initial magmas emplaced. This is best observed by the common occurrence of dioritic-quartz dioritic cognate inclusions hosted within the tonalite-granodiorite units, particularly around the contact zones, suggesting stoping of the former into the latter units. The diorites and quartz diorites also have much more irregular pluton–host rock contacts, are typically finer grained near their host rock contacts, suggesting more rapid cooling, and their magmatic foliations are frequently discordant to the magmatic foliations in the neighboring units. There is no field evidence to constrain the relative timing of the diorites–quartz diorites in the northwest part of the pluton to those in the southern part. It is difficult to separate the relative timing of the tonalite, granodiorite, and granitic units in the field, given their gradational contacts and concordant magmatic fabrics, both within units and across the geochemical contacts. The occasional occurrence of granite within the granodiorite (e.g. Fig. 3g) suggests the granite may have intruded later; however these examples are isolated and not connected with the central granitic unit. The central microgranite unit has a narrower (but still gradational) contact with its host granite, and also has a finer grain size and discordant magmatic foliations to its surrounding host, and thus was likely the youngest major unit to form. The leucogranitic dykes cross cut all of the main plutonic units with a variety of contact relationships, indicating that they intruded progressively throughout crystallization and cooling of the pluton. Regionally, the lamprophyre dykes are pre-, syn-, and post-emplacement of the LDP; however the dated dyke cross-cuts all of the units indicating it post-dates the main magmatic activity of the LDP. A maximum age for the emplacement of the LDP can be constrained by the ages of the host rock and structures in which it intrudes. For example, deposition of the Shinnel Formation and Moffat Shale Group host-rock around the LDP occurred during the Ashgillian (449–444 Ma), delineated by anceps graptolite fauna (Floyd & Rushton, 1993; Floyd, 1996). Also, the pluton contact discordantly cross cuts the Leadhills and Fardingmullach faults and the thermal aureole overprints regional metamorphic assemblages. Whilst the exact timings of regional fault kinematics and metamorphism are poorly constrained, they must have occurred sometime after original Ashgillian deposition and final closure of the Iapetus Ocean at ∼424 Ma (Dewey et al., 2015). Determining robust absolute ages for the LDP units has been problematic in the past and remains challenging with our new results. Even assignment of a specific age for each rock sample is difficult because of the wide zircon age distributions recorded. We interpret the geochronology results (Fig. 4) with the following cautions: (1) ICP analyses do not pre-treat for Pb loss. This may be a particularly large issue due to U-enrichment in the more evolved units, which is known to increase the potential for crystal lattice damage and Pb loss (e.g. Silver & Deutsch, 1963; Cherniak & Watson, 2001); (2) it is now well known that zircons in plutonic rocks often show a range of U–Pb ages that reflect autocrystic zircons grown in the melt, antecrystic zircons recycled from other parts of the plutonic system, and xenocrystic zircons inherited from older host rocks (e.g. Matzel et al., 2006; Lipman, 2007; Miller et al., 2007; Miller, 2008; Memeti et al., 2010; Schoene et al., 2012; Barboni et al., 2015); (3) there is a systematic analytical difference in ages between the 20 µm and 40 µm laser ablation beam sizes for the Ion Counter and Faraday Cup datasets and thus the chance of multiple zircon age domains being analysed and averaged; and (4) there are only a limited number of crystals analysed for the Ion Counter dataset. We interpret ages using the following steps: (1) only the zircons that are ±10% concordant have been plotted in Figure 4. The number of total zircons analysed per unit (see Supplementary Data A2) v. only those included in the age calculation in Figure 4 are as follows: granite (24/16), southern quartz diorite (43/33), microgranite (18/14), and lamprophyre (21/15); (2) we recognize that Faraday Cup analyses statistically give older ages compared to the Ion Counter ages, presumably because of the larger ablation spot at 40 µm versus only 20 µm for the Ion Counter spot sizes. We interpret this to imply that most zircons have older cores and that zircons grew over prolonged times, maybe in a deeper batholith that must have fed the SUT plutons. Thus, the crystallization age for the LDP is only represented by the rim ages of the zircons. Note that only the southern quartz diorite and the microgranite have both Faraday Cup analyses and Ion Counter analyses. The other two samples just have Ion Counter analyses; (3) Given the lack of any recognized post-emplacement thermal metamorphism in the LDP and hence no subsequent zircon growth or resorption, only the youngest ages were considered in calculations of the weighted mean ages that we think best represent the final crystallization age of each sample. The determined ages with errors and number of zircons used are as follows: (a) 396·7 ± 4·3 Ma for the southern quartz diorite (n = 5; sample LD-01); only Ion Counter analyses were used as the Faraday Cup ages were all older; (b) 415·5 ± 5·4 Ma for the granite (n = 6; sample LD-50); (c) for microgranite (sample LD-21) the youngest weighted mean age we can calculate for a group of zircons is much older at 424·3 ± 3·5 Ma (n = 6), which includes two Ion Counter and four Faraday Cup zircon analyses. We interpret this to be a mixed core–rim age. The two youngest zircons are 390·2 ± 3·4 Ma and 397·6 ± 3·0 Ma and potentially approximate final crystallization. The two zircons which yielded ages at >500 Ma are interpreted as xenocrysts; (d) the lamprophyre dyke (sample LD-51) yielded a weighted mean age for the youngest zircons of 397·6 ± 6·9 Ma (n = 4). Two xenocrysts yielded ages at 1078·5 ± 11·7 Ma and 1097·9 ± 7·1 Ma. Three other xenocrysts are ∼470–500 Ma old. The dyke, therefore, picked up xenocrysts from the SUT host rocks at some unknown level in the crustal column; and (4) we interpret the discordant ages from all samples to be mixed ages between in situ grown rims and much older inherited xenocrystic cores. The southern quartz diorite and the lamprophyre dyke both yielded weighted mean ages of the youngest zircons at ∼397 Ma. The younger zircon age of 390 Ma in the microgranite might have undergone some Pb loss, given that we do not find any other ages in any of the samples that are this young. The other young zircon age in the microgranite agrees well with the dyke and quartz diorite age and suggests that the best final crystallization age for the LDP is ∼397 Ma. We interpret the range of concordant zircon ages from 397 Ma to ∼430 Ma to be mixed ages of antecrystic zircons from deeper parts of the magmatic system and xenocrystic zircons from the host rock. The new U–Pb ages suggest that the LDP is younger than previously thought. However, we do not believe our final crystallization ages are inconsistent with those published previously and instead suggest that the continuum of ages between ∼430–397 Ma most likely records the complexity of the tectonomagmatic history prior to and post emplacement. We also suggest that the previously published U–Pb ages for the southern quartz diorite (414 ± 1 Ma; Aftalion et al., 1984) and granitic facies (422 Ma; 422 Ma and 439 Ma; Halliday et al., 1979) are mixed crystallization ages of rims and xenocrystic and/or antecrystic zircon crystals formed at depth. Given that amphibole largely occurs as a subsolidus replacement of primary pyroxene, we interpret the amphibole K–Ar age in the tonalite (387 ± 4 Ma; Halliday et al., 1980) as a cooling age, consistent with our final pluton crystallization age of ∼397 Ma. The older K–Ar biotite age (408 ± 4 Ma) for the tonalite may indicate localized open system issues with some minerals in this area (Halliday et al., 1980). The other plutons in the SUT have a range of published Rb–Sr whole-rock ages (410–392 Ma) and U–Pb zircon ages (414–387 Ma) (Table 1). We re-interpret the U–Pb SIMS ages of Miles et al. (2014) for the Fleet and Criffel plutons for which they frequently analysed internal zircon zones in complex crystals and chose to provide an average zircon age for each geochemical unit, despite distinctly different analysed ages (of up to 21 Ma) within individual units. Instead, we believe it is more reasonable to interpret their data as showing xenocrystic, antecrystic and autocrystic zircon ages falling between ∼410–382 Ma in the Fleet pluton and ∼425–400 Ma in the Criffel pluton. We interpret their minimum peak ages, ∼382 Ma and ∼400 Ma respectively, as more closely reflecting final crystallization ages, given the lack of post-emplacement thermal metamorphism in the SUT. Similarly, we interpret the zircon age ranges between ∼408–395 Ma for the Portencorkrie pluton (Oliver et al., 2008) and between ∼414–392 Ma for the Newry Igneous Complex (Cooper et al., 2016) as a mix of xenocrystic, antecrystic and autocrystic zircon ages. If only the minimum ages for all of the above, are considered, the range for final crystallization of the SUT plutons falls between 414 and 382 Ma at presently exposed levels and possibly ranging back to ∼430 Ma for plutons at deeper levels. Alternatively, these older ages are simply mixed ages between 414 and 382 Ma for in situ grown rims and much older xenocrystic cores. Constraints on pluton emplacement and magma chamber construction Traditionally, roughly elliptical, compositionally zoned plutons like the LDP have been assumed to be ‘ballooning’ plutons during which pluton margins and surrounding host rock expanded laterally as new magma batches arrived in the pluton centre. If correct, one would expect the following features to form: (1) intense flattening strains in the external host rock that for thermal–mechanical reasons should be focused in hot aureoles, thus forming mylonites in these aureoles; (2) large flattening strains in pluton margins, much of which should be sub-solidus, since these early margins would have cooled below their solidi before internal magma batches arrived; (3) the external host rock strains in the aureoles should be much larger than the strains in the pluton margin because of the smaller aureole radii versus the pluton radii; (4) constrictional to plane strain fabrics in pluton interiors, where newly arrived magma pulses were ascending; and (5) internal fabrics that pre-dated or were synchronous to internal contacts. Alternative models of nested diapirs (Paterson & Vernon, 1995), punch laccoliths (Corry, 1988; McCarthy et al., 2015), and sheeted and magmatic folded laccoliths (Coleman et al., 2012) can be evaluated as alternatives. In many previous studies of plutons, ascent, emplacement, and internal growth was inferred from either the internal characteristics of the pluton or the external characteristics of the host rock, but rarely both. Figures 2 and 7a–f illustrate the mapped characteristics of both the plutons and their surrounding host rocks; here we evaluate both sets of features with the goal of determining the coupled growth of the SUT plutons and displacement of their host rocks. In the presently exposed surface, individual pluton areas range from ∼5 km2 to ∼189 km2, representing the minimum amount (see below) of host rock that must be locally displaced out of this surface. The total area of the plutons makes up around 5% of the exposed SUT area in the UK (Fig. 1). Geophysical results (Stone et al., 2012) have concluded that the amount of plutonic material increases with depth, indicating that these plutons may simply represent upper crustal magmatic ‘fingers’ or columns rising off much larger plutonic systems at depth. If correct, then the amount of host rock displacement needed during magma emplacement also dramatically increases with depth (e.g. Cao et al., 2016). Similar ‘fingers’ or columns rising off larger intrusive complexes, of various scales, have been described or proposed in other studies worldwide (e.g. Pollard et al., 1975; Gerya et al., 2003; Perugini & Poli, 2005; Paterson et al., 2011). Map patterns, ages, and geochemistry indicate that all of these plutons grew incrementally by the amalgamation of smaller magma batches, thus indicating that displacement of the host rocks was also incremental. This implies that as younger pulses arrived, the previous magma pulses increasingly became the ‘host rock’, herein called internal host rock in comparison to the original external host rock. Recycling of external or internal host rock into newly arriving magma batches reduces the magnitude of required host rock displacement (Paterson et al., 2016); however, if additional magmatism passed through these plutons to higher crustal levels, actual internal host rock displacement magnitudes may be larger. A wealth of external host rock markers, including: (1) thicknesses and orientations of tectonostratigraphic packages; (2) package bounding and within package faults; and (3) bedding orientation and thicknesses, provide information about host rock displacements during emplacement. An examination of these markers in Figures 2 and 7a–f provides strong constraints, indicating that little to no lateral displacement of external host-rock occurred in the preserved sections. Package and bedding thicknesses do not systematically increase or decrease towards pluton boundaries, and orientations of almost all markers do not show significant deflections towards pluton margins, even well within thermal aureoles. Some apparent deflections and local folding, such as those around the Criffel pluton (Fig. 7b), have been argued to pre-date emplacement (Paterson & Vernon, 1995). Locally within narrow aureoles some deflection of external host rock markers occurs, as noted above for the LDP. Retro-deformation (see Supplementary Data A6) of these local deflections cannot account for more than 1–15% of the areal dimensions of the adjacent plutons. Furthermore, these local deflections are not propagated laterally outside of the narrow pluton aureoles, indicating that the lateral displacement must be accommodated by equal vertical displacements. The continuity of these same markers precludes lateral displacements on faults during pluton growth. Sometimes, local, weak, margin parallel structures occur in aureoles, but the implied strains are minor: emplacement-related mylonites in these aureoles are completely absent (mylonitic host rocks in the northern aureole of the Fleet pluton are associated with regional strike-slip deformation along the Moniaive Shear Zone and Orlock Bridge Fault [Phillips et al., 1995] and the pluton discordantly cross-cuts the hostrock foliation [Barnes et al., 1996], indicating the mylonites are unrelated to the emplacement of the Fleet pluton). The common undeflected, steeply dipping markers, and lack of deflection of markers with different dips, and the common steep dips of pluton contacts also precludes growth of sub-horizontal laccolith bodies. The above observations cannot rule out some early, lateral, external host rock displacement in the region now occupied by the plutons. But they require that the great majority of the external host rock displacement must occur by either upward (e.g. Corry, 1988; McCarthy et al., 2015) or downward displacement (e.g. Cruden, 1998; Paterson & Farris, 2006) within the region now occupied by the pluton and to a lesser degree in the preserved, narrow structural aureoles. Figure 8 displays published examples of how upward or downward movement of external host rock may occur without significantly deflecting external host rock markers outside of narrow aureoles. We believe that depending on depth, the mechanisms of host rock displacement changed from upward (shallow levels) to downward (deeper levels) during growth of the magmatic ‘fingers’. At shallow depths where magma pressures overcame the overlying rock strength and weight, upward host rock movement dominated and laccoliths formed. At deeper levels, downward host rock movements increasingly dominated and fairly discordant stocks or plutons formed. Published examples of downward host rock movement include: (1) formation of rim synclines and inward dipping monoclines within aureoles; (2) sinking of large host rock segments or ‘flaps’ within the pluton; (3) stoping; (4) downward displacement on brittle or ductile faults; and (5) ductile flow within aureoles resulting in steeply plunging lineations and strain intensities that increase with depth (e.g. Cruden, 1998; Paterson & Farris, 2006; Cao et al., 2016). Fig. 8. View largeDownload slide Models for the emplacement of the Loch Doon and other late Caledonian plutons in the upper crust of the Southern Uplands Terrane. (a) Upward doming and faulting. (b) Downward host rock flow, stoping and floor subsidence. Fig. 8. View largeDownload slide Models for the emplacement of the Loch Doon and other late Caledonian plutons in the upper crust of the Southern Uplands Terrane. (a) Upward doming and faulting. (b) Downward host rock flow, stoping and floor subsidence. In detail, external SUT pluton contacts are irregular and stepped rather than smoothly curving (e.g. Figs 2 and 7; Pitcher & Berger, 1972; Stevenson, 2009). A few steps are associated with post-emplacement faults, but most are not. Host rock markers are typically truncated along these stepped margins. Stoped host rock blocks often occur short distances within the pluton. These observations suggest that mechanisms causing either upward or downward movement are often masked or modified by rare faulting, more common stoping and widespread hornfels metamorphism in the late stages of intrusive complex growth. At depth within the SUT plutons, we argue that downward flow mechanisms dominated, resulting in ‘punch laccoliths’ or fairly discordant stocks with irregular margins and approximate elliptical shapes. We would predict that upward movement accompanying growth of stocks or sheeted laccoliths occurred immediately above these plutons, potentially reflected by the Cheviot and Carsphairn plutons, and downward flow resulting in larger plutonic bodies occurred below them (e.g. Stone et al., 2012). Internal growth of SUT plutons The irregular elliptical shapes, normal compositional zoning (with local reversals), inward younging, steeply dipping, sharp to more commonly gradational internal contacts, and approximately concentric magmatic fabrics that overprint many (but not all) internal contacts are widespread in SUT plutons and in normally zoned plutons worldwide (e.g. Paterson & Vernon, 1995). The internal contacts, magmatic structures, and geochemistry discussed subsequently all support a model of incremental growth of these intrusive complexes, implying that the final body formed from the amalgamation of a few to many trapped magma batches. The marginal distribution of early, more mafic units raises several possibilities: (1) early mafic bodies formed near the centre of the complexes and were laterally displaced during arrival of younger bodies; (2) early mafic batches were small and spaced out around the centre of the future plutonic complex and, or, formed ring-like, steeply dipping sheets; or (3) early mafic pulse(s) formed a large magma body close to the size of the present pluton. We discard #1 for the SUT plutons because these mafic complexes show no strong, solid-state deformation or other structures required if large lateral displacement occurred. Although we have seen convincing examples of #2 in other arcs (e.g. Paterson et al., 2018), we prefer #3 for the SUT plutons for the following reasons: (i) in some SUT plutons mafic outer margins are fairly continuous and these zones show no evidence of separate internal bodies or strong deformation that would support expansion; (ii) internal contacts with the next innermost unit suggest either mingling or mixing with, or erosion of, a formerly more extensive outer unit; and (iii) the presence in inner units of cognate blocks and antecrystic zircons likely derived from the outer units, supports erosion and ‘recycling’ of these formerly more extensive outer units (e.g. Gaschnig et al., 2016; Paterson et al., 2016). If either model #2 or #3 is correct, this implies that very little lateral growth of the evolving intrusive complex occurred after the initial mafic units were emplaced. Instead emplacement of younger pulses caused either upward or downward flow of older units and, or, erosion ± recycling of parts of the earlier pulses (Paterson et al., 2016); active crystal–liquid fractionation and filter pressing also likely transferred melt from older pulses into younger (e.g. Bachmann & Bergantz, 2004). We suggest that the younger pulses tended to ‘nest’ in the earlier pulses largely due to a rheological control; they ascended where the former intrusive material was hottest and, or, still had melt present. Thus, the SUT plutons can be viewed as horizontal slices through inward freezing and narrowing vertical magma conduits. For most SUT plutons and particularly the LDP, we see no evidence for the growth of gently dipping laccoliths and the expected host rock displacements. Nor do we see evidence for magmatic folding and sinking or rising of earlier, sub-horizontal magmatic sheets. We suggest that the magmatic fabrics reflect late increments of strain in crystal mushes as they approach their solidi, with the strain reflecting complex crystal interactions (e.g. Bergantz et al., 2015) in response to a roughly concentric stress field around the evolving complex. Karlstrom et al. (2010, 2017) have modeled such concentric stress fields and established that they can easily swamp regional ambient stress, both in the pluton and surrounding host rock. Examples in the SUT plutons where magmatic fabrics are discordant to external pluton-host contacts across which they are continuous with metamorphic foliations in structural aureoles, but discordant to regional host rock foliations support the Karlstrom models. Summary A combination of regional geophysical studies and our evaluation of the external and internal features of the SUT plutons suggest a model of magmatic ‘fingers’ or columns rising off deeper, bigger intrusive complexes (Figs 8 and 9). At shallow levels these fingers may form laccolith complexes or more irregular stocks. At slightly deeper levels they form fairly discordant stocks, during which downward host rock displacement largely occurs early when a magma finger first reaches a new crustal level. Subsequently arriving, typically more evolved batches tend to ‘nest’ within the rheologically weakest portions of earlier batches and largely displace these earlier intrusive batches rather than the external host-rock. The active parts of these nesting systems shrank with time, consistent with a freezing magma feeder system. Magmatic fabrics formed during ascent are largely reset just above the solidus for any magma batch and reflect the strain driven by complex crystal interactions and local stress fields caused by the evolving intrusive complex. Fig. 9. View largeDownload slide A model for the evolution of the Southern Uplands Terrane crustal column during late Caledonian times. (a) Early magmatism at ∼430 Ma was in response to break-off of subducted Iapetan lithosphere, subsequent upwelling of hot asthenosphere, and melting of slightly depleted lithospheric mantle. Magmas partially fused, mixed and mingled with Laurentian and Iapetan (± young primitive Avalonian) accretionary and collisional complexes in the lower crust. (b) Post-convergence transtension at ∼415–400 Ma renewed melting of lithospheric mantle and lower crustal complexes, adding new and recharging existing magma reservoirs, leading to the formation of a mid-lower crustal batholith. (c) Magmas in the mid–lower crustal batholith evolved by complex open system dynamics, including fractionation, mixing and contamination, before being emplaced in the upper crust as compositionally zoned nested plutons at ∼400–380 Ma. Fig. 9. View largeDownload slide A model for the evolution of the Southern Uplands Terrane crustal column during late Caledonian times. (a) Early magmatism at ∼430 Ma was in response to break-off of subducted Iapetan lithosphere, subsequent upwelling of hot asthenosphere, and melting of slightly depleted lithospheric mantle. Magmas partially fused, mixed and mingled with Laurentian and Iapetan (± young primitive Avalonian) accretionary and collisional complexes in the lower crust. (b) Post-convergence transtension at ∼415–400 Ma renewed melting of lithospheric mantle and lower crustal complexes, adding new and recharging existing magma reservoirs, leading to the formation of a mid-lower crustal batholith. (c) Magmas in the mid–lower crustal batholith evolved by complex open system dynamics, including fractionation, mixing and contamination, before being emplaced in the upper crust as compositionally zoned nested plutons at ∼400–380 Ma. Geochemical evolution of Southern Uplands Terrane plutons and crustal column As multiple magma batches ascended and coalesced into upper crustal complexes, it is appropriate to ask if temporal changes occurred in the crustal column and the magma source(s), and the degree to which ascending magma batches fractionated, mixed or were contaminated. We explore these questions below. Most SUT plutons have average Sr/Y ratios of around 40, suggesting depths to the Moho during magma generation of around 40-50 km (Profeta et al., 2015). The LDP samples plot at lower Sr/Y ratios of about 20, suggesting Moho depths of ∼30 km. If future, more precise, dating indicates that the LDP is relatively young compared to other SUT plutons, this would suggest that crustal thinning occurred from ∼45 km to ∼30 km during SUT magmatism. Magma sources and mixing We have identified two key trends in the isotopic data for both the LDP and SUT plutons that shed light on potential sources and the relative roles of fractionation, magma mixing, and host rock contamination in the LDP (Fig. 6a): (1) defined by data from diorite, quartz diorite, tonalite, and granodiorite, and (2) by data from the granite and microgranite. In Trend #1 progressively more radiogenic Sri signatures (from 0·7043 to 0·7061) occur over a small variation in 1/Sr: thus fractionation was not the major control on the isotopic compositions. Given that SUT metasedimentary host rock Sri, εNd, and δ18O values are higher than those observed in Trend #1, contamination of mantle-derived melts by host rocks at some crustal depth is a likely option. We thus argue that Trend #1 reflects a number of magma pulses produced by the mixing of mantle and crustal magmas, representing at least two, if not more, end-member compositions. This is also reflected in the <0·4 P2O5/K2O ratio v. MgO plot (Fig. 5d) and plots of εNd and δ18O (Fig. 6b–c) and Pb isotopes (see Supplementary Data A4), all of which show that most, if not all, of these magmas had at least some crustal input. Isotopic trends suggest that the most primitive magmas project towards Sri of ∼0·704, εNd of 0, and δ18O of < 6, indicating a mantle source between Bulk Earth and Depleted Mantle. At least one of the most evolved end-members projects towards Sri ∼0·712, εNd of -6, and δ18O of ∼12, indicating a crustal source that is easily represented by the analysed host rock compositions in the SUT. Other evidence that supports this trend reflecting mixing of multiple pulses of variably contaminated magmas includes: (1) the common occurrence of mafic enclaves in more felsic magmas (e.g. Fig. 3i); (2) the numerous xenoliths or ‘autoliths’ (inclusions) in SUT plutons that isotopically and geochemically overlap with analysed host rock (Fig. 6a–c); (3) the presence of older xenocrystic zircons (Fig. 4; see Supplementary Data A2 and A3); and (4) hybrid magma zones between pulses that are overprinted by magmatic foliations indicating magma mixing (Fig. 2). In Trend #2, the granite and microgranite show progressively more evolved Sri signatures (∼0·7053 to ∼0·7080) relative to Trend #1 and have a different slope on the Sriv. 1/Sr plot, indicating that fractionation played a more important role. We suggest that it is likely that the rocks in Trend #2 represent a number of magma pulses produced from the increased fractionation of the previously mixed magmas of Trend #1, along with some continued crustal contamination. Additional evidence to support mixing and crustal contamination of the magma pulses in this trend is the same as described above for Trend #1. The evidence supporting an increase in fractionation is discussed in more detail below. The isotopic compositions in Trend #2 project towards the quartz diorite–granodiorite transition in Trend #1 (Fig. 6b), suggesting that these represent the primitive source of magmas for this trend. Previous authors have raised the issue of whether old crust at depth was involved in generating the SUT plutons (e.g. Blaxland et al., 1979; Brown et al., 1979; Halliday et al., 1979, 1980; Harmon & Halliday, 1980; Stephens & Halliday, 1984; Stephens et al., 1985; Thirlwall, 1988, 1989; Stone et al., 1997; Miles et al., 2014; Stone, 2014; Dewey et al., 2015). Trend #1, which can be fully explained by mixing depleted mantle and exposed crust, along with the moderately radiogenic Pb isotope compositions comparable to SUT metasediments, indicates that no older crust is needed, but these results do not rule out the involvement of older crust. Ideally the ‘contaminating crustal source’ should be low in εNd and high in Sri (greater spread in Sriv. εNd) and contribute zircons of the age of the crust to ascending melts, although the latter could have been dissolved at high temperatures. Two Mesoproterozoic and a few lower Palaeozoic zircons did occur in our dating of a dyke in the LDP (Fig. 4, sample LD-51), but it is possible these zircons were inherited from detritus in SUT sediments sourced from the north, either from the Grenvillian-aged Laurentian basement or from the Midland Valley Arc basement, a signature commonly seen in SUT metasediments (e.g. Waldron et al., 2008; Phillips et al., 2009). With the present dataset, we cannot rule this out, but we note that there is no substantial evidence that there is a considerable thickness of old or evolved basement rocks below the LDP that was involved in the generation of its magmas. Instead, it is likely that the most primitive LDP units were sourced from a slightly depleted mantle modified by partial melts of lower crustal material, which may include a complex array of young accretionary and collisional lithologies such as: (1) marine sediments derived from the overriding Laurentian plate; (2) oceanic sediments, volcanic rocks, pillow lavas and ophiolite slivers accreted from the subducting Iapetus oceanic plate; and (3) oceanic sediments or young primitive basement from the imbricated, under-thrusted Avalonian continental margin. In contrast, the central units of the Fleet pluton have much higher average Sri ratios than SUT host rocks (Fig. 6), which led Halliday et al. (1980) to suggest they were sourced from a different parent to the LDP. Halliday et al. (1980) modeled the source region for these plutons to be the allochthonous underplated margin of Avalonia, whereas the LDP magmas were sourced from the autochthonous Laurentian margin. Miles et al. (2014) agreed with this model and extended it to the Criffel pluton, based on analyses of 207Pb/204Pb, zircon εHfi, and δ18O isotopes. This model is thought to be consistent with seismic profiles (Soper et al., 1992) and potential field modelling (Stone et al., 2012) indicating a northward dipping reflector in the mid-lower crust beneath the Criffel and Fleet plutons, interpreted as the plate boundary zone between Avalonia (Lakesman-Leinster Terrane) and Laurentia (SUT; Fig. 1). Miles et al. (2014) modelled the elevated Sri and δ18O, and radiogenic Pb ratios, and initial εHfi in zircon between approximately -1·2 to +4·5, of these two plutons to show an original depleted mantle melt contaminated by either Avalonian basement, or supracrustal rocks deposited on the Avalonian basement, equivalent to the Skiddaw Group sediments in the Lakesman-Leinster Terrane. To test whether these isotopic signatures could have been sourced from Avalonian basement underthrusted below the SUT, more robust age and isotopic analyses of the poorly exposed basement in the Lakesman-Leinster Terrane would be necessary. Fractionation The weakly curved trends in many of the whole-rock major and trace-element relationships for the LDP (Fig. 5) have led the most recent studies to conclude that the concentric compositional zoning within the LDP is a result of either a two- or three-stage in situ fractional crystallization process (Brown et al., 1979; Halliday et al., 1980; Tindle & Pearce, 1981; Stephens & Halliday, 1984). Whilst we certainly believe that in situ fractionation did occur during progressive crystallization of the LDP, we explore further whether fractionation was the dominant driver of the LDP compositional zonation and whether this or other processes occurred in situ or at deeper crustal levels. Scattered, but on average gently curved patterns of whole-rock major and trace elements, combined with the REE plots, suggest that ongoing fractionation occurred either in magma chambers feeding the LDP and, or, in the LDP magma chamber. The decrease in whole-rock TiO2, MgO, CaO, P2O5, FeOt, Rb, Ba, Sr, and Nd with increasing SiO2 is consistent with the growth and continuous removal of pyroxene, amphibole, Ca-plagioclase, biotite, ilmenite, apatite, allanite and titanite and increase in Na-plagioclase, K-feldspar and quartz. These minerals are all observed in thin section with approximated orders of crystallization in agreement with a fractionation trend. We interpret this crystal–liquid fractionation to have occurred in already mixed mantle–crustal magmas as indicated by the isotopic data. Other evidence to support strong fractionation of the least evolved granite magma to more evolved granite includes: (1) the narrowing range of major and trace element concentrations with increasing SiO2 (Fig. 5b–c); (2) the gradational compositional zonation (Fig. 2); (3) high concentrations of incompatible elements in the most evolved granites (Fig. 5b–c); and (4) the dramatic scatter at lower SiO2, which in part resulted from the formation of cumulates. The Sri and δ18O isotope compositions of the microgranite approach similar levels in the host rocks and, therefore, it is likely that partial melting of SUT metasediments may have contributed to the source of the innermost microgranite unit. Simple mass balance calculations suggest that the least evolved microgranites (Sri ∼0·7080) could have evolved to the most radiogenic granite magmas (Sri ∼0·7061) by up to ∼80% partial melting of SUT metasedimentary host rocks (average Sri of ∼0·7086). Other evidence to support continued host rock contamination in the innermost microgranite units includes the presence of Ordovician and Cambrian U–Pb zircon ages in sample LD-21 (see Supplementary Data A2). We argue that a prolonged open magma reservoir system is necessary to satisfy all the geologic datasets from both plutonic rock and host rocks. We propose that after the initial construction of the magma chamber(s), it is likely that the magma reservoir was replenished by the injection of additional magmas from a source undergoing varying degrees of mixing and fractionation at depth. Mixing of two magma sources was the likely dominant control on evolution of the diorite to granodiorite units, whereas simultaneous fractionation and mixing of different magmas was probably the dominant control on the evolution of the granite to microgranite units. Whilst the LDP magma chamber likely underwent ongoing in situ fractionation, new hotter magma batches probably reheated the older magma mush causing further mixing and formation of gradational compositional contacts. We believe that high-resolution, high precision zircon geochronology would show crystallization did not occur rapidly by simple in situ thermal cooling of two or three magma batches (as suggested by previous models) but was a complex process extended by continual chamber rejuvenation and thermal buffering from magma recharge sourced from below. This would be consistent with numerical modeling by Paterson et al. (2011), Gelman et al. (2013), and Karakas et al. (2017) suggesting that plutonic-scale upper crustal magma reservoirs can reside above their solidus for >105 years provided they are fed with frequent magma inputs. The closely related isotopic and whole-rock geochemistry of all the SUT plutons indicates that the overall processes controlling their evolution were similar. Whilst the isotopic data (Fig. 6) suggest that mixing of different source magmas occurred to various extents within individual plutons, the whole-rock major oxides, trace elements and REE (Fig. 5) in every pluton follow similar patterns, including focusing into a narrow range at higher SiO2 concentrations indicating that fractionation became increasingly important at higher degrees of evolution. The scattered whole-rock chemistry and isotope compositions of both mafic plutons (<65% SiO2) and the contemporaneous regional lamprophyre dykes (Figs 5 and 6), which are typically at the primitive end of the plutonic data, is likely due to three factors: (1) the heterogeneous compositions of the lower crust where the mantle-derived melts forming dykes and plutons were mixed with partially melted crust; (2) the subtle differences in the composition of the source magmas of the plutons versus the dykes; and (3) the formation of cumulates left behind from fractionated, filter-pressed melts. Given that all SUT plutons show contemporaneous antecryst and autocryst U–Pb zircon ages between ∼430–380 Ma, overlapping isotopic and geochemical affinities (Figs 5 and 6) and concentric zonation patterns (Figs 2 and 7a–f), we consider that all the SUT plutons discussed in this paper have exceptionally similar petrogenetic affinities and were connected to related, deeper magma reservoirs in the mid-lower crust. We envisage the currently exposed plutons to be the upper crustal ‘fingers’ extending from much bigger magmatic systems in the SUT mid-lower crust (Fig. 9). The foliated and non-foliated cognate inclusions found in lamprophyre dykes in the SUT of Scotland and Ireland (Smith et al., 1991; Floyd & Phillips, 1999) may record further evidence of the proposed batholithic magmatism at depth. The integration of pluton geochemistry and isotopic data with a range of other geologic datasets from both plutonic rocks and host rocks is something rarely done in studies of zoned plutons. We think the above discussion demonstrates the value in, and importance of, evaluating both plutonic rock and host rock datasets towards fully understanding the origin, emplacement and evolution of plutons. Constraints on tectonic setting Whilst the primary focus of this study was to record emplacement-related and magma chamber-related processes within zoned plutons, our new results and alternative interpretations of previously published data, provide additional constraints on geodynamic models for the LDP and the surrounding plutons in the SUT. Many recent studies of plutons north and south of the Iapetus Suture zone, the so called ‘Trans-Suture Suite’ containing the Fleet, Criffel, Bengairn, Cheviot, Skiddaw, Shap and Leinster plutons (Brown et al., 2008; Dewey et al., 2015), exclude the LDP when providing geodynamic models for magma genesis. Significantly, our new work, including re-interpretation of the other SUT plutons, clearly illustrates that the geochemistry, isotopic composition, geochronology and plutonic- host rock relationships of the LDP are analogous to the Trans-Suture Suite granitoids, as are the Carsphairn, Portencorkrie and Newry Igneous Complex plutons. Our data support the proposal of Miles et al. (2016) in that any geodynamic models for the Trans-Suture Suite must include and be consistent with data from the LDP, Carsphairn, Portencorkrie and the Newry Igneous Complex. A geodynamic model that satisfactorily integrates all of the available tectono-magmatic data for the SUT and nearby Caledonian terranes remains elusive (Dewey et al., 2015). The LDP and other SUT plutons record magmatic activity between ∼430–382 Ma. Significantly, our new interpretations of the antecrystic zircon ages of the SUT plutons indicate prolonged magmatic activity at depth, starting sometime during final Laurentian–Avalonian convergence (∼430 Ma) and culminating in early- to mid-Devonian upper crustal plutonism. This is contemporaneous with widespread intrusion of calc-alkaline lamprophyre dykes throughout the SUT metasediments dated at ∼424-385 Ma (Rock et al., 1986, 1988; Shand et al., 1994; Vaughan, 1996; Soper & Woodcock, 2003; Oliver et al., 2008; this study). Final convergence and suturing of Laurentia and Avalonia is constrained by sedimentology, biostratigraphy, and palaeomagnetic data to have occurred around ∼424 Ma (Kemp, 1987; Soper et al., 1992; Mac Niocaill, 2000; Mange et al., 2005; Dewey et al., 2015). Thus, an unusual feature of the SUT plutons is that they record an overall high-K to shoshonitic calc-alkaline signature (Fig. 5a), but magmatism is constrained by geochronology to be syn- to post-collisional (Fig. 4 and Supplementary Data A2). Tectonic discrimination plots (Supplementary Data A7) also indicate the SUT plutons have volcanic-arc to syn-collisional affinities, further supporting the apparent discrepancies between geochemical and geochronological datasets. Models of post-collisional slab break-off subsequent to Iapetus suturing favour decompression of the asthenosphere, and subsequent pervasive melting of, and addition of lithospheric mantle and lamprophyric melts of calc-alkaline affinity to the lower crust of the SUT and Lakesman-Leinster Terrane (e.g. Atherton & Ghani, 2002; Oliver et al., 2008; Chew & Stillman, 2009). Whilst melting of lithospheric mantle and lower crustal material after slab break-off is consistent with the available geochemical and isotopic evidence from the SUT plutons (Figs 5 and 6), our interpretation of antecrystic zircon with ages between ∼430–424 Ma suggests that slab break-off may have occurred prior to the cessation of Laurentian–Avalonian convergence as recorded by sedimentology, biostratigraphy and palaeomagnetic evidence. This interpretation supports other recent geochronological studies of Caledonian magmatism which suggest slab break-off occurred at ∼430-425 Ma (Neilson et al., 2009; Cooper et al., 2013; Miles et al., 2016). Miles et al. (2016) suggested that break-off of the Iapetan slab was accompanied by widespread delamination of Avalonian lithospheric mantle below the Iapetus Suture zone and Avalonia to the south and that hot dry asthenospheric mantle flowed up to impinge directly on the based on the lower crust or on thin, delaminated lithospheric mantle in this area. Whilst partial delamination of lithospheric mantle below the suture zone may have occurred, geochemical and isotopic data (Figs 5 and 6) from the SUT plutons, volcanic rocks and lamprophyre dykes do not indicate any contribution of asthenospheric mantle melt. As such, it is likely that an appreciable thickness of lithospheric mantle remained below the SUT crust during and after Iapetan slab break-off. Current geodynamic models that attempt to explain all the geological relationships struggle to resolve the issue of prolonged magmatism at depth within the SUT (Dewey, 2002; Woodcock et al., 2007; Brown et al., 2008; Miles et al., 2014, 2016; Dewey et al., 2015). Whilst slab break-off models suitably explain the origin of zircons with ages of ∼430–415 Ma within SUT intrusions, it is unlikely that break-off could have led to such prolonged magmatism to explain the origin of antecrystic and autocrytsic zircon ages between ∼415 and 382 Ma. Brown et al. (2008) modeled the generation of the Trans-Suture Suite plutons to be driven entirely by post-convergence transtension and lithospheric thinning at ∼410–400 Ma, which led to widespread melting of subducted, enriched, hydrated Avalonian lithospheric mantle and subsequent intrusion of lamprophyric magmas into and melting of the crust. However, this does not explain the absence of enriched mantle isotopic signatures (Fig. 6b), nor the origin of zircons with analysed ages of ∼430–410 Ma in these plutons. In their model of slab break-off and lithospheric delamination beneath the Iapetus Suture zone, Miles et al. (2014, 2016) argue that post-convergence transtension simply facilitated emplacement of the SUT plutons rather than enhanced and prolonged melt generation and magmatism at depth. However, it is difficult to explain how delamination of the Iapetan slab and Avalonian lithosphere could have led to such prolonged magmatism, as shown by the range of antecrystic and autocrystic zircon ages, without subsequent additional magma or heat input from other tectonic controls. A single tectonic control for the origin of the prolonged magmatism in the SUT is difficult to reconcile with all the available data and it is likely that the plutons formed as a response to a combination of tectonic controls through time. Therefore, we provide an adaptation of previously published late Caledonian geodynamic models that best fits our interpretation of the SUT plutons and other available tectono-magmatic and stratigraphic data. We suggest early magmatism at depth in the SUT was initiated at ∼430 Ma in response to break-off of the subducted Iapetan lithosphere under the Laurentian margin. Slab break-off led to the upwelling of hot asthenosphere and subsequent widespread melting of slightly depleted lithospheric mantle beneath the Iapetus Suture zone. Lithospheric mantle melts likely intruded the lower crust in the SUT where they partially fused, mixed and mingled with Laurentian and Iapetan (± young primitive Avalonian) accretionary and collisional complexes (Fig. 9a). Subsequent post-convergence transtension between ∼415–400 Ma likely renewed melting of lithospheric mantle and lower crustal complexes, adding new and recharging existing magma reservoirs at depth, and leading to the formation of a mid–lower crustal batholith (Fig. 9b). Evidence for transtension at the current exposure level in the SUT is limited (Dewey et al., 2015) but transtension may have been facilitated by reactivation of inherited collisional-related structures at depth. The Sr/Y ratios of the SUT plutons (Fig. 5d) suggest that crustal thicknesses in the SUT during magma generation may have varied in space and, or, time, which may be a result of regional transtension. Magmas in the mid-lower crustal batholith subsequently evolved by the complex open system dynamics discussed earlier, before being emplaced as compositionally zoned nested plutons in the upper crust between ∼414–382 Ma (Fig. 9c). CONCLUSIONS Previously published models of the emplacement and evolution of the Loch Doon pluton do not correspond to the entire field, structural, geochronological, whole-rock geochemical and isotopic, magnetic susceptibility, and geophysical observations available from both the pluton and its host rock. Through examination of new data and reinterpretation of previous studies, we propose a new model that records the type and relative magnitudes of processes controlling magma chamber growth and evolution in this pluton. New U–Pb zircon ages from the Loch Doon pluton indicate that it was emplaced into the upper crust at ∼397 Ma. Through detailed field mapping of plutonic and host rock structures and consideration of regional geophysical studies, we suggest that trapping of rising magma and construction of the pluton was dominated by vertical host rock displacement: likely a combination of upward doming and faulting within a few km of the Earth’s surface, and downward host rock flow, stoping and floor subsidence at deeper levels. At the level of current exposure, we interpret that the majority of hostrock displacement was downward, forming ‘punch laccoliths’ or discordant stocks and that pluton growth likely occurred early when the initial magma ‘finger’ first reached its new crustal level. Detailed analysis of Sri, εNd, δ18O and Pb isotopes and whole-rock major and trace element geochemistry indicates that in situ crystallization was a complex process, extended by continual chamber rejuvenation and thermal buffering by recharge from magmas undergoing varying degrees of mixing, fractionation and contamination at deeper crustal levels. Early magma batches in the Loch Doon magma chamber were mainly related by mixing between depleted mantle-derived magmas with two or more deep crustal sources, whereas later batches underwent increasing degrees of fractionation, mixing with other fractionates and late contamination by metasedimentary host rocks. These subsequently arriving, typically more evolved magma batches tended to ‘nest’ within the rheologically weakest portions of earlier batches and largely displaced these earlier intrusive batches rather than the external host rock. Analysis of the pluton’s ubiquitous magmatic foliations and their relationship to internal compositional boundaries, host rock structures and stoped blocks, indicate that internal magma readjustments in the Loch Doon magma chamber were dominated by incremental vertical flow of partially crystallized marginal magmas during centralized nesting of magma batches. The active parts of the nesting systems shrank with time, consistent with a freezing magma feeder system. A collection of equivalent data from many other similarly zoned late Caledonian Southern Uplands Terrane plutons of the UK shows that our model for the Loch Doon pluton could be applicable elsewhere. A combination of regional geophysical studies and our evaluation of the external and internal features of the Southern Uplands Terrane plutons, support a model of magmatic ‘fingers’ rising upwards from larger intrusive complexes at depth. Our interpretation of antecrystic and autocrystic zircon in the Southern Uplands Terrane plutons between ∼430–382 Ma allows us to provide further constraints on the likely tectonic controls on the generation of these magmas. Using zircon ages and whole-rock isotopic data, we envisage a model of slab break-off at ∼430 Ma initiating partial melting of slightly depleted mantle material which intruded the lower crust and partially fused, mixed and mingled with Laurentian, Iapetan, and Avalonian accretionary and collisional complexes. Subsequent lithospheric transtension between ∼415–400 Ma added more mantle partial melts and heat to the mid-lower crust. Sr/Y ratios indicate crustal thicknesses during magma generation varied in space and, or, time between ∼30 and 45 km, perhaps as a result of regional transtension. These tectonic processes resulted in widespread genesis of magmas of high-K to shoshonitic calc-alkaline affinity which subsequently evolved by complex open-system mixing, fractionation, and contamination in a mid-lower crustal batholith before being emplaced as compositionally zoned nested plutons in the upper crust between ∼414–382 Ma. This study demonstrates the importance of integrating a range of geologic datasets from both plutonic rocks and their host rocks towards fully understanding the origin, emplacement and evolution of zoned plutons. ACKNOWLEDGEMENTS RH would like to thank N. & G. Bromilow for invaluable help with fieldwork logistics. Thanks are extended to A. Calder, D. Herd, A. Mackie, and T. Raub for help with laboratory analyses and L. Medina for data digitization. We gratefully acknowledge reviews by R. Strachan, M. Cooper, and A. Yoshinobu, and the editorial handling of John Gamble. FUNDING This work was supported by the National Science Foundation (EAR-0073943 to S.R.P.) for geochronological analysis and the University of St Andrews (Irving Fund to R.H.) towards fieldwork activities. Arizona LaserChron Center is supported by the National Science Foundation (EAR-1338583). SUPPLEMENTARY DATA Supplementary data for this paper are available at Journal of Petrology online. REFERENCES Aftalion M. , van Breeman O. , Bowes D. R. ( 1984 ). Age constraints on basement of the Midland Valley of Scotland . Transactions of the Royal Society of Edinburgh: Earth Sciences 75 , 53 – 64 . Google Scholar CrossRef Search ADS Al-Hafdh N. M. ( 1985 ). The alteration petrology of the Cheviot granite. Ph.D. thesis, University of Newcastle upon Tyne, Newcastle upon Tyne. Anderson P. E. , Cooper M. R. , Stevenson C. T. , Hastie A. R. , Hoggett M. , Inman J. , Meighan I. G. , Hurley C. , Reavy R. J. , Ellam R. M. ( 2016 ). Zonation of the Newry Igenous Complex, Northern Ireland, based on geochemical and geophysical data . Lithos 260 , 95 – 106 . Google Scholar CrossRef Search ADS Atherton M. P. , Ghani A. A. ( 2002 ). Slab breakoff: a model for Caledonian, Late Granite syn-collisional magmatism in the orthotectonic (metamorphic) zone of Scotland and Donegal, Ireland . Lithos 62 , 65 – 85 . Google Scholar CrossRef Search ADS Bachmann O. , Bergantz G. W. ( 2004 ). On the origin of crystal-poor rhyolites: extracted from batholithic crystal mushes . Journal of Petrology 45 , 1565 – 1582 . Google Scholar CrossRef Search ADS Barboni M. , Annen C. , Schoene B. ( 2015 ). Evaluating the construction and evolution of upper crustal magma reservoirs with coupled U/Pb zircon geochronology and thermal modelling: a case study from the Mt. Capanne pluton (Elba, Italy) . Earth and Planetary Science Letters 432 , 436 – 448 . Google Scholar CrossRef Search ADS Barnes R. P. , Boland M. P. , Phillips E. R. , Stone P. ( 1996 ). Orlock Bridge Fault and the Moniaive Shear Zone- an excursion. In: Stone P. (ed.) Geology in South-West Scotland: An Excursion Guide . Nottingham, UK : British Geological Survey . Bergantz G. W. , Schleicher J. M. , Burgisser A. ( 2015 ). Open-system dynamics and mixing in magma mushes . Nature Geoscience 8 , 793 – 796 . Google Scholar CrossRef Search ADS Bird A. F. , Thirlwall M. F. , Strachan R. A. , Manning C. J. ( 2013 ). Lu–Hf and Sm–Nd dating of metamorphic garnet: evidence for multiple accretion events during the Caledonian orogeny in Scotland . Journal of Geological Society, London 170 , 301 – 317 . Google Scholar CrossRef Search ADS Blaxland A. B. , Aftalion M. , van Breem O. ( 1979 ). Pb isotopic compositions of the feldspars from the Scottish Caledonian granites, and the nature of the underlying crust . Scottish Journal of Geology 15 , 139 – 151 . Google Scholar CrossRef Search ADS Bluck B. J. ( 2013 ). Geotectonic evolution of Midland Scotland from Cambrian to Silurian: a review . Scottish Journal of Geology 49 , 105 – 116 . Google Scholar CrossRef Search ADS British Geological Survey ( 1994 ). Loch Doon, Scotland Sheet 8E, Solid Geology 1:50000. Keyworth, Nottingham : The Stationery Office . Brown G. C. , Cassidy J. , Tindle A. G. , Hughes D. J. ( 1979 ). The Loch Doon Granite: an example of granite petrogenesis in the British Caledonides . Journal of the Geological Society, London 136 , 745 – 753 . Google Scholar CrossRef Search ADS Brown P. E. , Ryan P. D. , Soper N. J. , Woodcock N. H. ( 2008 ). The Newer Granite problem revisited: a transtensional origin for the Early Devonian Trans-Suture Suite . Geological Magazine 145 , 235 – 256 . Google Scholar CrossRef Search ADS Cao W. , Paterson S. R. , Saleeby J. , Zalunardo S. ( 2016 ). Bulk arc strain, crustal thickening, magma emplacement, and mass balance in the Mesozoic Sierra Nevada arc . Journal of Structural Geology 84 , 14 – 30 . Google Scholar CrossRef Search ADS Chapman N. A. , Cameron I. B. ( 1978 ). Initial results of the investigation of the north-central part of the Loch Doon complex. Institute of Geological Sciences, Hydrological Report 17K, WE/EN/78/13. Chappell B. W. , White A. J. R. ( 1974 ). Two contrasting granite types . Pacific Geology 8 , 173 – 174 . Cherniak D. J. , Watson E. B. ( 2001 ). Pb diffusion in zircon . Chemical Geology 172 , 5 – 24 . Google Scholar CrossRef Search ADS Chew D. M. , Stillman C. J. ( 2009 ). Late Caledonian orogeny and magmatism. In: Holland C. H. , Sanders I. S. (eds) The Geology of Ireland , 2nd edn. Edinburgh : Dunedin Academic Press , pp. 143 – 173 . Coleman D. S. , Bartley J. M. , Glazner A. F. , Pardue M. S. ( 2012 ). Is chemical zonation in plutonic rocks driven by changes in source magma composition or shallow-crustal differentiation? Geosphere 8 , 1568 – 1587 . Google Scholar CrossRef Search ADS Cook D. R. , Weir J. A. ( 1979 ). Structure of the Lower Palaeozoic rocks around Cairnsmore of Fleet, Galloway . Scottish Journal of Geology 3 , 187 – 202 . Google Scholar CrossRef Search ADS Cooper M. R. , Johnston T. P. ( 2004 ). Late Palaeozoic intrusives. In: Mitchell W. I. (ed.) The Geology of Northern Ireland: Our Natural Foundation , 2nd edn. Belfast : Geological Survey of Northern Ireland , pp. 61 – 68 . Cooper M. R. , Crowley Q. G. , Hollis S. P. , Noble S. R. , Henney P. J. ( 2013 ). A U–Pb age for the Late Caledonian Sperrin Mountains minor intrusions suite in the north of Ireland: timing of slab break-off in the Grampian terrane and the significance of deep-seated, crustal lineaments . Journal of the Geological Society, London 170 , 603 – 614 . Google Scholar CrossRef Search ADS Cooper M. R. , Anderson P. , Condon D. J. , Stevenson C. , Ellam R. , Meighan I. G. , Crowley Q. G. ( 2016 ). Shape and intrusion history of the Late Caledonian, Newry Igneous Complex, Northern Ireland. In: Young M. E. (ed.) Unearthed: Impacts of the Tellus Surveys of North of Ireland . Dublin : Royal Irish Academy , pp. 145 – 155 . Corry C. E. ( 1988 ). Laccoliths: mechanics of emplacement and growth . Geological Society of America, Special Paper 220 , 1 – 114 . Google Scholar CrossRef Search ADS Courrioux G. ( 1987 ). Oblique diapirism: the criffel granodiorite/granite zoned pluton (southwest Scotland) . Journal of Structural Geology 9 , 313 – 330 . Google Scholar CrossRef Search ADS Cox K. G. , Bell J. D. , Pankhurst R. J. ( 1979 ). The Interpretation of Igneous Rocks . London : Allen & Unwin . Google Scholar CrossRef Search ADS Cruden A. R. ( 1998 ). On the emplacement of tabular granites . Journal of the Geological Society, London 155 , 853 – 862 . Google Scholar CrossRef Search ADS Dawson J. , Floyd J. D. , Philip P. R. , Burley A. J. , Allsop J. M. , Bennet J. R. P. , Marsden G. R. , Leake R. C. , Brown M. J. ( 1977 ). A mineral reconnaissance survey of the Doon-Glenkens area, south-west Scotland . Institute of Geological Sciences, Mineral Reconnaissance Programme Report 18 . Deer W. A. ( 1934 ). The Cairnsmore of Carsphairn igneous complex . Quarterly Journal of the Geological Society, London 91 , 47 – 76 . Google Scholar CrossRef Search ADS Dewey J. F. ( 2002 ). Transtension in arcs and orogens . International Geology Review 44 , 402 – 438 . Google Scholar CrossRef Search ADS Dewey J. F. , Dalziel I. W. D. , Reavy R. J. , Strachan R. A. ( 2015 ). The Neoproterozoic to Mid-Devonian evolution of Scotland: a review and unresolved issues . Scottish Journal of Geology 51 , 5 – 30 . Google Scholar CrossRef Search ADS Evans J. A. ( 1995 ). Geochronology. In: Stone P. (ed.) Geology of the Rhins of Galloway District. Memoir of the British Geological Survey, Sheets 1 and 3 (Scotland) . London, UK : HMSO , p. 96 . Fettes D. J. , Henney P. J. ( 1999 ). Intrusive rocks. In: Floyd J. D. (ed.) Geology of the Carrick-Loch Doon District. Memoir of the British Geological Survey, Sheets 8W and 8E (Scotland) . London, UK : HMSO , pp. 56 – 76 . Floyd J. D. ( 1996 ). Lithostratigraphy of the Ordovician rocks in the Southern Uplands: Crawford Group, Moffat Shale Group, Leadhills Supergroup . Transactions of the Royal Society of Edinburgh: Earth Sciences 86 , 153 – 165 . Google Scholar CrossRef Search ADS Floyd J. D. ( 1999 ). Ordovician and Silurian turbidite succession of the Southern Uplands Terrane and structure. In: Floyd J. D. (ed.) Geology of the Carrick-Loch Doon District. Memoir of the British Geological Survey, Sheets 8W and 8E (Scotland) . London, UK : HMSO , pp. 13 – 53 . Floyd J. D. , Phillips E. R. ( 1999 ). Xenoliths of Southern Uplands “Basement” in a lamprophyre dyke, Central Belt, Glenluce, SW Scotland . Scottish Journal of Geology 35 , 57 – 62 . Google Scholar CrossRef Search ADS Floyd J. D. , Rushton A. W. A. ( 1993 ). Ashgill greywackes in the Southern Uplands of Scotland: an extension of the Ordovician succession in the Northern Belt . Transactions of the Royal Society of Edinburgh: Earth Sciences 84 , 79 – 85 . Google Scholar CrossRef Search ADS Gardiner C. I. , Reynolds S. H. ( 1932 ). The Loch Doon granite area, Galloway . Quarterly Journal of the Geological Society, London 83 , 1 – 34 . Google Scholar CrossRef Search ADS Gaschnig R. M. , Vervoort J. D. , Tikoff B. , Lewis R. S. ( 2016 ). Construction and preservation of batholiths in the northern U.S. Cordillera . Lithosphere 9 , 315 – 324 . Google Scholar CrossRef Search ADS Gaskarth J. W. ( 1995 ). Intrusive rocks. In: Stone P. (ed.) Geology of the Rhins of Galloway District. Memoir of the British Geological Survey, Sheets 1 and 3 (Scotland) . London, UK : HMSO , pp. 55 – 65 . Gelman S. E. , Gutiérrez F. J. , Bachmann O. ( 2013 ). On the longevity of large upper crustal silicic magma reservoirs . Geology 41 , 759 – 762 . Google Scholar CrossRef Search ADS Gerya T. V. , Uken R. , Reinhardt J. , Watkeys M. K. , Maresch W. V. , Clarke B. M. ( 2003 ). Cold fingers in a hot magma: numerical modeling of country-rock diapirs in the Bushveld Complex, South Africa . Geology 31 , 753 – 756 . Google Scholar CrossRef Search ADS Halliday A. N. , Stephens W. E. , Harmon R. S. ( 1980 ). Rb-Sr and O isotopic relationships in 3 zoned Caledonian granitic plutons, Southern Uplands, Scotland: evidence for varied sources and hybridization of magmas . Journal of the Geological Society, London 137 , 329 – 348 . Google Scholar CrossRef Search ADS Halliday A. N. , Aftalion M. , van Breeman O. , Jocelyn J. ( 1979 ). Petrogenetic significance of Rb–Sr and U–Pb isotopic systems in the 400 Ma old British Isles granitoids and their hosts . The Geological Society, London, Special Publications 8 , 653 – 661 . Google Scholar CrossRef Search ADS Harmon R. S. , Halliday A. N. ( 1980 ). Oxygen and strontium isotope relationships in the British late Caledonian granites . Nature 283 , 21 – 25 . Google Scholar CrossRef Search ADS Holgate N. ( 1943 ). The Portencorkrie complex of Wigtownshire . Geological Magazine 80 , 171 – 195 . Google Scholar CrossRef Search ADS Irvine T. N. , Baragar W. R. A. ( 1971 ). A guide to the chemical classification of the common volcanic rocks . Canadian Journal of Earth Science 8 , 523 – 548 . Google Scholar CrossRef Search ADS Jhingran A. G. , Tomkeieff S. I. ( 1942 ). The Cheviot granite . Quarterly Journal of the Geological Society, London 98 , 241 – 254 . Google Scholar CrossRef Search ADS Kafafy A. M. , Tarling D. H. ( 1985 ). Magnetic fabric of some granite aureoles, Southern Uplands, Scotland . Journal of the Geological Society, London 142 , 1007 – 1014 . Google Scholar CrossRef Search ADS Karakas O. , Degruyter W. , Bachmann O. , Dufek J. ( 2017 ). Lifetime and size of shallow magma bodies controlled by crustal-scale magmatism . Nature Geoscience 10 , 446 – 450 . Google Scholar CrossRef Search ADS Karlstrom L. , Dufek J. , Manga M. ( 2010 ). Magma chamber stability in arc and continental crust . Journal of Volcanology and Geothermal Research 190 , 249 – 270 . Google Scholar CrossRef Search ADS Karlstrom L. , Paterson S. R. , Jellinek A. M. ( 2017 ). A reverse energy cascade for crustal magma transport . Nature Geoscience 10 , 604 – 608 . Google Scholar CrossRef Search ADS Kemp A. E. S. ( 1987 ). Tectonic development of the Southern Belt of the Southern Uplands accretionary complex . Journal of the Geological Society, London 144 , 827 – 838 . Google Scholar CrossRef Search ADS Leake R. C. , Auld H. A. , Stone P. , Johnson C. E. ( 1981 ). Gold mineralisation at the southern margin of the Loch Doon granitoid complex, south-west Scotland . Institute of Geological Sciences, Mineral Reconnaissance Programme Report 46 , 8 – 9 . Leake R. C. , Cooper C. ( 1983 ). The Black Stockarton Moor subvolcanic complex, Galloway . Journal of the Geological Society, London 140 , 665 – 676 . Google Scholar CrossRef Search ADS Lee M. K. ( 1982 ). Regional Geophysics of the Cheviot Area . Institute of Geological Science, Radioactive Waste Disposal Research Series 82 , WE/EN/82/002. Leggett J. K. , McKerrow W. S. , Eales M. H. ( 1979 ). The Southern Uplands of Scotland: a Lower Palaeozoic accretionary prism . Journal of the Geological Society, London 136 , 755 – 770 . Google Scholar CrossRef Search ADS Lipman P. W. ( 2007 ). Incremental assembly and prolonged consolidation of Cordilleran magma chambers: evidence from the Southern Rocky Mountain volcanic field . Geosphere 3 , 42 – 70 . Google Scholar CrossRef Search ADS Mac Niocaill C. ( 2000 ). A new Silurian palaeolatitude for eastern Avalonia and evidence for crustal rotations in the Avalonian margin of southwestern Ireland . Geophysics Journal International 141 , 661 – 671 . Google Scholar CrossRef Search ADS Mange M. A. , Dewey J. F. , Floyd J. D. ( 2005 ). The origin, evolution and provenance of the Northern Belt (Ordovician) of the Southern Uplands Terrane, Scotland: a heavy mineral perspective . Proceedings of the Geologists’ Association 116 , 251 – 280 . Google Scholar CrossRef Search ADS Matzel J. E. P. , Bowring S. M. , Miller R. B. ( 2006 ). Timescale of pluton construction at differing crustal levels: examples from the Mount Stuart and Tenpeak intrusions, North Cascades, Washington . GSA Bulletin 118 , 1412 – 1430 . Google Scholar CrossRef Search ADS McCarthy W. J. ( 2013 ). An evaluation of orogenic kinematic evolution utilizing crystalline and magnetic anisotropy in granitoids. Ph.D. thesis, National University of Ireland, Cork. McCarthy W. J. , Petronis M. S. , Reavy R. J. , Stevenson C. T. ( 2015 ). Distinguishing diapirs from inflated plutons: an integrated rock magnetic fabric and structural study on the Roundstone Pluton, western Ireland . Journal of the Geological Society, London 172 , 550 – 565 . Google Scholar CrossRef Search ADS McKerrow W. S. , Mac Niocaill C. , Dewey J. F. ( 2000 ). The Caledonian Orogeny redefined . Journal of the Geological Society, London 157 , 1149 – 1154 . Google Scholar CrossRef Search ADS Memeti V. , Paterson S. R. , Matzel J. , Mundil R. , Okaya D. ( 2010 ). Magmatic lobes as “snapshots” of magma chamber growth and evolution in large, composite batholiths: an example from the Tuolumne intrusion, Sierra Nevada, California . GSA Bulletin 122 , 1912 – 1931 . Google Scholar CrossRef Search ADS Merriman R. J. , Roberts B. ( 2001 ). Low-grade metamorphism in the Scottish Southern Uplands terrane: deciphering the patterns of accretionary burial, shearing and cryptic aureoles . Transactions of the Royal Society of Edinburgh: Earth Sciences 91 , 521 – 537 . Google Scholar CrossRef Search ADS Miles A. J. , Woodcock N. H. , Hawkesworth C. J. ( 2016 ). Tectonic controls on post-subduction granite genesis and emplacement: the late Caledonian suite of Britain and Ireland . Gondwana Research 39 , 250 – 260 . Google Scholar CrossRef Search ADS Miles A. J. , Graham C. M. , Hawkesworth C. J. , Gillespie M. R. , Dhuime B. , Hinton R. W. ( 2014 ). Using zircon isotope compositions to constrain crustal structure and pluton evolution: the iapetus suture zone granites in Northern Britain . Journal of Petrology 55 , 181 – 207 . Google Scholar CrossRef Search ADS Miller J. S. ( 2008 ). Assembling a pluton… one increment at a time . Geology 36 , 511 – 512 . Google Scholar CrossRef Search ADS Miller J. S. , Glazner A. F. , Farmer G. L. , Suayah I. B. , Keith L. A. ( 2000 ). A Sr, Nd, and Pb isotopic study of mantle domains and crustal structure from Miocene volcanic rocks in the Mojave Desert, California . GSA Bulletin 112 , 1264 – 1279 . Google Scholar CrossRef Search ADS Miller J. S. , Matzel J. E. P. , Miller C. F. , Burgess S. D. , Miller R. B. ( 2007 ). Zircon growth and recycling during the assembly of large, composite arc plutons . Journal of Volcanology and Geothermal Research 167 , 282 – 299 . Google Scholar CrossRef Search ADS Miller R. B. , Paterson S. R. ( 2001 ). Construction of mid-crustal sheeted plutons: examples from the North Cascades, Washington . GSA Bulletin 113 , 1423 – 1442 . Google Scholar CrossRef Search ADS Mitchell A. H. G. , McKerrow W. S. ( 1975 ). Analogous evolution of the Burma Orogen and the Scottish Caledonides . GSA Bulletin 86 , 305 – 315 . Google Scholar CrossRef Search ADS Nakamura N. ( 1974 ). Determination of REE, Ba, Fe, Mg, Na, and K in carbonaceous and ordinary chondrites . Geochimica et Cosmochimica Acta 38 , 757 – 775 . Google Scholar CrossRef Search ADS Needham D. T. ( 2004 ). Deformation in Moffat Shale detachment zones in the western part of the Scottish Southern Uplands . Geological Magazine 141 , 441 – 453 . Google Scholar CrossRef Search ADS Neeson J. C. ( 1984 ). The geology and geochemistry of the Newry igneous complex, Northern Ireland. Ph.D. thesis, Queen’s University, Belfast. Neilson J. C. , Kokelaar B. P. , Crowley Q. G. ( 2009 ). Timing, relations and cause of plutonic and volcanic activity of the Siluro-Devonian post-collision magmatic episode in the Grampian Terrane, Scotland . Journal of the Geological Society, London 166 , 545 – 561 . Google Scholar CrossRef Search ADS O’Connor P. J. ( 1975 ). Rb-Sr whole-rock isochron for the Newry granodiorite, NE Ireland . Science Proceedings of the Royal Society of Dublin A5 , 407 – 413 . Oertel G. ( 1955 ). Der pluton von Loch Doon in Sudschottland . Geotektonische Forschungen 11 , 1 – 83 . Oliver G. J. H. , Wilde S. A. , Wan Y. ( 2008 ). Geochronology and geodynamics of Scottish granitoids from the late Neoproterozoic break-up of Rodinia to Palaeozoic collision . Journal of the Geological Society, London 165 , 661 – 674 . Google Scholar CrossRef Search ADS Parslow G. R. ( 1968 ). The physical and structural features of the Cairnsmore of Fleet granite and its aureole . Scottish Journal of Geology 4 , 91 – 108 . Google Scholar CrossRef Search ADS Paterson S. R. , Farris D. W. ( 2006 ). Downward host rock transport and the formation of rim monoclines during the emplacement of Cordilleran batholiths . Transactions of the Royal Society of Edinburgh: Earth Sciences 97 , 397 – 413 . Google Scholar CrossRef Search ADS Paterson S. R. , Vernon R. H. ( 1995 ). Bursting the bubble of ballooning plutons: a return to nested diapirs emplaced by multiple processes . GSA Bulletin 107 , 1356 – 1380 . Google Scholar CrossRef Search ADS Paterson S. R. , Clausen B. , Memeti V. , Schwartz J. ( 2018 ). Arc magmatism, tectonism, and tempos in Mesozoic arc crustal sections of the Peninsular and Transverse Ranges, southern California. In: Kraatz B. , Lackley J. S. , Fryxell J. E. (eds) Field Excursions in Southern California: Field Guide to the 2016 GSA Cordilleran Section Meeting: Geological Society of America Field Guide 45 , 1 – 104 . Boulder, USA : The Geological Society of America, Inc . Paterson S. R. , Memeti V. , Mundil R. , Žák J. ( 2016 ). Repeated, multiscale, magmatic erosion and recycling in an upper-crustal pluton: implications for magma chamber dynamics and magma volume estimates . American Mineralogist 101 , 2176 – 2198 . Google Scholar CrossRef Search ADS Paterson S. R. , Okaya D. , Memeti V. , Economos R. , Miller R. B. ( 2011 ). Magma addition and flux calculations of incrementally constructed magma chambers in continental margin arcs: combined field, geochronologic, and thermal modeling studies . Geosphere 7 , 1439 – 1468 . Google Scholar CrossRef Search ADS Paterson S. R. , Yoshinobu A. S. , Yuan E. S. , Miller R. B. , Fowler T. K. Jr , Schmidt K. L. ( 1998 ). Interpreting magmatic fabric patterns in plutons . Lithos 44 , 53 – 82 . Google Scholar CrossRef Search ADS Peccerillo A. , Taylor S. R. ( 1976 ). Geochemistry of Eocene calc-alkaline volcanic rocks of the Kastamonu area, northern Turkey . Contributions to Mineralogy and Petrology 58 , 63 – 81 . Google Scholar CrossRef Search ADS Perugini D. , Poli D. ( 2005 ). Viscous fingering during replenishment of felsic magma chambers by continuous inputs of mafic magmas: field evidence and fluid-mechanics experiements . Geology 33 , 5 – 8 . Google Scholar CrossRef Search ADS Phillips E. R. , Barnes R. P. , Boland M. P. , Fortey N. J. , McMillan A. A. ( 1995 ). The Moniaive Shear Zone: a major zone of sinistral strike-slip deformation in the Southern Uplands of Scotland . Scottish Journal of Geology 31 , 139 – 149 . Google Scholar CrossRef Search ADS Phillips E. R. , Evans J. A. , Stone P. , Horstwood M. S. A. , Floyd J. D. , Smith R. A. , Akhurst M. C. , Barron H. F. ( 2003 ). Detrital Avalonian zircons in the Laurentian Southern Uplands terrane, Scotland . Geology 31 , 625 – 628 . Google Scholar CrossRef Search ADS Phillips E. R. , Smith R. A. , Stone P. , Pashley V. , Horstwood M. ( 2009 ). Zircon age constraints on the provenance of Llandovery to Wenlock sandstones from the Midland Valley Terrane of the Scottish Caledonides . Scottish Journal of Geology 45 , 131 – 146 . Google Scholar CrossRef Search ADS Phillips W. J. ( 1956 ). The Criffel-Dalbeatte granodiorite complex . Quarterly Journal of the Geological Society, London 112 , 221 – 239 . Google Scholar CrossRef Search ADS Pidgeon R. T. , Aftalion M. ( 1978 ). Cogenetic and inherited zircon U–Pb systems in granites: Palaeozoic granites of Scotland and England . The Geological Society, London, Special Publication 10 , 183 – 248 . Piper J. D. A. ( 2007 ). Palaeomagnetism of the Loch Doon Granite Complex, Southern Uplands of Scotland: the Late Caledonian palaeomagnetic record and an Early Devonian episode of True Polar Wander . Tectonophysics 432 , 133 – 157 . Google Scholar CrossRef Search ADS Pitcher W. S. , Berger A. R. ( 1972 ). The Geology of Donegal, a Study of Granite Emplacement and Unroofing . London : Wiley Interscience . Pollard D. D. , Muller O. H. , Dockstader D. R. ( 1975 ). The form and growth of fingered sheet intrusions . GSA Bulletin 86 , 351 – 363 . Google Scholar CrossRef Search ADS Powell D. W. ( 1970 ). Magnetised rocks within the Lewisian of Western Scotland and under the Southern Uplands . Scottish Journal of Geology 6 , 353 – 369 . Google Scholar CrossRef Search ADS Profeta L. , Ducea M. N. , Chapman J. B. , Paterson S. R. , Gonzalez S. M. H. , Kirsch M. , Petrescu L. , DeCelles P. G. ( 2015 ). Quantifying crustal thickness over time in magmatic arcs . Scientific Reports 5 , 17786 . Google Scholar CrossRef Search ADS PubMed Reynolds D. L. ( 1934 ). The eastern end of the Newry igneous complex . Quarterly Journal of the Geological Society, London 90 , 585 – 636 . Google Scholar CrossRef Search ADS Reynolds D. L. ( 1936 ). The two monzonitic series of the Newry Complex . Geological Magazine 73 , 337 – 348 . Google Scholar CrossRef Search ADS Reynolds D. L. ( 1943 ). The south-western end of the Newry igneous complex. A contribution towards the petrogenesis of the granodiorites . Quarterly Journal of the Geological Society, London 99 , 205 – 246 . Google Scholar CrossRef Search ADS Rock N. M. S. , Gaskarth J. W. , Rundle C. C. ( 1986 ). Late Caledonian dyke-swarms in Southern Scotland: a regional zone of primitive K-rich lamprophyres and associated vents . Journal of Geology 94 , 505 – 522 . Google Scholar CrossRef Search ADS Rock N. M. S. , Gaskarth J. W. , Henny P. J. , Shand P. ( 1988 ). Late Caledonian dyke-swarms in northern Britain: some preliminary petrogenetic and tectonic implications of their distribution and chemical variations . Canadian Mineralogist 26 , 3 – 22 . Ruddock I. ( 1969 ). The geochemistry of the southern half of the Loch Dee-Loch Doon pluton, S.W. Scotland. Ph.D. thesis, University of Newcastle upon Tyne, Newcastle upon Tyne. Rutledge H. ( 1950 ). The petrology and structure of the southern part of the Loch Doon Complex. Ph.D. thesis, University of Edinburgh, Edinburgh. Schoene B. , Schaltegger U. , Brack P. , Latkoczy C. , Stracke A. , Günther D. ( 2012 ). Rates of magma differentiation and emplacement in a ballooning pluton recorded by U–Pb TIMS-TEA, Adamello batholith, Italy . Earth and Planetary Science Letters 355–356 , 162 – 173 . Google Scholar CrossRef Search ADS Shand S. J. ( 1943 ). Eruptive Rocks. Their Genesis, Composition, Classification, and Their Relation to Ore-Deposits with a Chapter on Meteorite . New York : John Wiley & Sons . Shand P. , Gaskarth J. W. , Thirlwall M. F. , Rock N. M. S. ( 1994 ). Late Caledonian lamprophyre dyke swarms of South-Eastern Scotland . Contributions to Mineralogy and Petrology 51 , 277–228. Silver L. T. , Deutsch S. ( 1963 ). Uranium-Lead isotopic variations in zircons: a case study . Journal of Geology 71 , 721 – 758 . Google Scholar CrossRef Search ADS Smith R. A. , Johnston T. P. , Legg I. C. ( 1991 ). Geology of the country around Newtownards. Memoir of the Geological Survey of Northern Ireland, Sheet 37 and 38. London : Stationery Office Books . Soper N. J. , Woodcock N. H. ( 2003 ). The lost Lower Old Red Sandstone of England and Wales: a record of post-Iapetan flexure or Early Devonian transtension? Geological Magazine 140 , 627 – 647 . Google Scholar CrossRef Search ADS Soper N. J. , Strachan R. A. , Holdsworth R. E. , Gayer R. A. , Greiling R. O. ( 1992 ). Sinistral transpression and the Silurian closure of Iapetus . Journal of the Geological Society, London 149 , 871 – 880 . Google Scholar CrossRef Search ADS Stephens W. E. , Halliday A. N. ( 1979 ). Compositional variation in the Galloway plutons. In: Atherton M. P. , Tarney J. (eds) Origin of Granite Batholiths . Orpington : Shiva , pp. 9 – 17 . Google Scholar CrossRef Search ADS Stephens W. E. , Halliday A. N. ( 1984 ). Geochemical contrasts between late Caledonian granitoid plutons of northern, central and southern Scotland . Transactions of the Royal Society of Edinburgh: Earth Sciences 75 , 259 – 273 . Google Scholar CrossRef Search ADS Stephens W. E. , Whitely J. E. , Thirlwall M. F. , Halliday A. N. ( 1985 ). The Criffel zoned pluton: correlated behaviour of rare earth element abundances with isotopic systems . Contributions to Mineralogy and Petrology 89 , 226 – 238 . Google Scholar CrossRef Search ADS Stevenson C. T. E. ( 2009 ). The relationship between forceful and passive emplacement: the interplay between tectonic strain and magma supply in the Rosses Granite Complex, NW Ireland . Journal of Structural Geology 31 , 270 – 287 . Google Scholar CrossRef Search ADS Stone P. ( 2014 ). The Southern Uplands Terrane in Scotland—a notional controversy revisited . Scottish Journal of Geology 50 , 97 – 123 . Google Scholar CrossRef Search ADS Stone P. , Merriman R. J. ( 2004 ). Basin thermal history favours an accretionary origin for the Southern Uplands terrane, Scottish Caledonides . Journal of the Geological Society, London 161 , 829 – 836 . Google Scholar CrossRef Search ADS Stone P. , Kimbell G. S. , Henney P. J. ( 1997 ). Basement control on the location of strike-slip shear in the Southern Uplands of Scotland . Journal of the Geological Society, London 154 , 141 – 144 . Google Scholar CrossRef Search ADS Stone P. , McMillan A. A. , Floyd J. D. , Barnes R. P. , Phillips E. R. ( 2012 ). British Regional Geology: South of Scotland , 4th edn. Keyworth, Nottingham : British Geological survey , pp. 78 – 95 . Teall J. J. H. ( 1899 ). The Silurian Rocks of Britain. Memoirs of the Geological Survey, UK. Volume 1: Scotland . Glasgow : HMSO . Thirlwall M. F. ( 1988 ). Geochronology of Late Caledonian magmatism in northern Britain . Journal of the Geological Society, London 145 , 951 – 967 . Google Scholar CrossRef Search ADS Thirlwall M. F. ( 1989 ). Short paper: movement on proposed terrane boundaries in northern Britain: constraints from Ordovician-Devonian igneous rocks . Journal of the Geological Society, London 146 , 373–316. Tindle A. G. ( 1982 ). Petrogenesis of the Loch Doon granitic intrusion, Southern Uplands of Scotland. Ph.D. thesis, Open University, Milton Keynes. Tindle A. G. , Pearce J. A. ( 1981 ). Petrogenetic modelling of in situ fractional crystallization in the zoned Loch Doon pluton, Scotland . Contributions to Mineralogy and Petrology 78 , 196 – 207 . Google Scholar CrossRef Search ADS Tindle A. G. , McGarvie D. W. , Webb P. C. ( 1988 ). The role of hybridization and crystal fractionation in the evolution of the Cairnsmore of Carsphairn Intrusion, Southern Uplands of Scotland . Journal of the Geological Society, London 145 , 11 – 21 . Google Scholar CrossRef Search ADS Vaughan A. P. M. ( 1996 ). A tectonomagmatic model for the genesis and emplacement of Caledonian calc-alkaline lamprophyres . Journal of the Geological Society, London 153 , 613 – 623 . Google Scholar CrossRef Search ADS Waldron J. W. F. , Floyd J. D. , Simonetti A. , Heaman L. M. ( 2008 ). Ancient Laurentian detrital zircon the closing Iapetus Ocean, Southern Uplands terrane, Scotland . Geology 36 , 527 – 530 . Google Scholar CrossRef Search ADS Waldron J. W. F. , Schofield D. I. , Dufrane S. A. , Floyd J. D. , Crowley Q. G. , Simonetti A. , Dokken R. J. , Pothier H. D. ( 2014 ). Ganderia–Laurentia collision in the Caledonides of Great Britain and Ireland . Journal of the Geological Society, London 171 , 555 – 569 . Google Scholar CrossRef Search ADS Woodcock N. H. , Soper N. J. , Strachan R. A. ( 2007 ). A Rheic cause for the Acadian deformation in Europe . Journal of the Geological Society, London 164 , 1023 – 1036 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: firstname.lastname@example.org This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
Journal of Petrology – Oxford University Press
Published: Apr 10, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera