Origin of Mesozoic Porphyritic Rocks and Regional Magmatic Evolution in the Zijinshan Ore Field of Fujian Province, China: Hf-O Isotope Characteristics of Magmatic Zircons

: Mesozoic porphyritic rocks from the Zijinshan area, southwestern Fujian Province, China, are andesitic to rhyolitic in composition. The whole-rock SiO 2 contents of these rocks are between 62.5% and 78.1%. Magmatic zircon from the Mesozoic porphyritic rocks was determined via secondary-ionization mass spectrometry (SIMS) for the U-Pb age and Hf and O isotopes. The zircon U-Pb ages could be mainly divided into three age groups: Group 1: ~138.8 Ma; Group 2: 109.2~107.4 Ma; and Group 3: 99.7~98.2 Ma. The ε Hf(t) and δ 18 O values of the porphyritic zircons showed that the porphyritic rocks in Group 2 were more a ﬀ ected by mantle-derived magma. Combined with previous research results, the medium-acidic magmatism in the southwestern Fujian Province can be divided into eight periods: Paleoproterozoic, Mesoproterozoic, Middle Neoproterozoic, Silurian to Lower Devonian, Permian to Triassic, Middle Jurassic to early Lower Cretaceous, late Lower Cretaceous, and late Lower Cretaceous to early Upper Cretaceous. The Paleoproterozoic crust was the predominant magmatic source for the subsequent Mesoproterozoic to Jurassic magmatism, but the only melts that were closely related to mineralization were derived from partial melting of the Mesoproterozoic crust and a more depleted upper mantle. ) granite porphyry with plagioclase and quartz phenocrysts; ( c ) felsic minerals matrix of the granitic porphyry developed devitriﬁcation phenomenon; ( d ) chloritization was developed along the cleavage fracture of biotite; ( e ) quartz phenocrysts of granitic porphyry were featured frequently with melting corrosion structure; ( f ) amphiboles were strongly altered with carbonation and chloritization; ( g – k ) granodioritic porphyry had a porphyritic structure, and plagioclase microcrystals were developed in the matrix; ( i ) quartz phenocrysts set in a fne- to micro-grained holocrystalline matrix; m quartz phenocryst with a typical recrystallized rim with micro-inclusions; matrix had a microcrystalline structure, plagioclase was partially sericized; plagioclase phenocrysts developed chloritization and sericization, plagioclase obvious sericization; quartz;


Introduction
The Zijinshan ore field, the largest porphyry-epithermal mineralization system in southeast China, is located in the southwestern Fujian Province ( Figure 1) [1,2]. The major economic deposits in the region include: (1) the Zijinshan high-sulfidation epithermal deposit [3][4][5], (2) the Luoboling Cu-Mo-Au porphyritic deposit [6], (3) the Yueyang low-sulfidation epithermal Ag polymetallic deposit [7], and (4) the Wuziqilong transitional Cu deposit [8,9]. Zijinshan mineralization was closely related to two large-scale intermediate-acid magmatic activities in the Mesozoic: late Jurassic (165-149 Ma) and Cretaceous (118-92 Ma) [10][11][12][13][14]. The late Jurassic intrusive rocks were mainly composed of medium-to coarse-grained monzogranite and granite, and the early Cretaceous intrusive-volcanic rocks were mainly hypabyssal intrusive rock or subvolcanic porphyries such as granodioritic porphyry, granitic porphyry, quartz porphyry, and dacitic porphyry. The dacitic and/or granodioritic porphyries were strongly associated with the Zijinshan Cu-Au deposit and the Luoboling porphyry Cu-Mo mineralization [15,16]. subjected to intense alteration, it is difficult to completely describe the characteristics of the original magma from a whole-rock geochemical study. Even though zircon Hf isotopes have been conventionally used to study the origin of these rocks [4,10], the large range of Hf isotopic compositions from various mantle sources sometimes prevents the effective identification of the magmatic source and, thus, the genesis of the economic ore deposits. In addition to U, Pb, and Hf isotopes, zircon has a high closure temperature for oxygen. Oxygen isotopes retain their original basic oxygen isotope characteristics even during high-grade eclogite-facies metamorphism and, therefore, work as more effective indicators for magma genesis [18]. Studies of oxygen isotopes in magmatic zircons have shown that mantle-derived zircons have almost uniform δ 18 O values (5.3 ± 0.6‰) [19,20]. Increasing δ 18 O values generally indicate the involvement of crustal materials in a magmatic source. Therefore, the oxygen isotopic signatures of magmatic zircons can provide robust indicators of the interaction between crust and mantle in a given region [21][22][23][24]. As Cretaceous porphyritic rocks play an important role in mineralization, much attention has been paid to their genesis [14,17]. However, as the magmatic rocks in the Zijinshan area were usually subjected to intense alteration, it is difficult to completely describe the characteristics of the original magma from a whole-rock geochemical study. Even though zircon Hf isotopes have been conventionally used to study the origin of these rocks [4,10], the large range of Hf isotopic compositions from various mantle sources sometimes prevents the effective identification of the magmatic source and, thus, the genesis of the economic ore deposits.
In addition to U, Pb, and Hf isotopes, zircon has a high closure temperature for oxygen. Oxygen isotopes retain their original basic oxygen isotope characteristics even during high-grade eclogite-facies metamorphism and, therefore, work as more effective indicators for magma genesis [18]. Studies of oxygen isotopes in magmatic zircons have shown that mantle-derived zircons have almost uniform δ 18 O values (5.3 ± 0.6‰) [19,20]. Increasing δ 18 O values generally indicate the involvement of crustal materials in a magmatic source. Therefore, the oxygen isotopic signatures of magmatic zircons can provide robust indicators of the interaction between crust and mantle in a given region [21][22][23][24].
In this study, we focus on Mesozoic porphyritic rocks in the Zijinshan area, as these rocks are closely related to the formation of economic ore deposits. The zircon Hf-O isotopic data of the porphyritic rocks demonstrated that the mantle-derived materials contributed more to the early Cretaceous melts than the late Cretaceous ones. The Jurassic porphyritic rocks showed more involvement in crustal materials. We also discuss the magmatic evolution of the crustal source from Proterozoic to Mesozoic.

Geological Background
The Zijinshan area is located in the southwestern Fujian Provence (Figure 1b), China (Figure 1a). The northwest-striking Yunxiao-Shanghang fault zone and several northeast-orienting Xuanhe anticlinoria are the chief ore-controlling structures. Flysch sediments metamorphosed by the Sinian and early Paleozoic periods are outcropped in the Zijinshan area ( Figure 1c). The Indosinian orogeny, which occurred in the eastern part of China during the early Triassic, resulted in a change in the tectonic framework in Southeast China from the predominantly Tethys tectonic domain to the circum Pacific active continental margin [25][26][27][28]. The orogeny formed a series of intermediate-acid igneous rocks accompanied by large-scale mineralization [25].
The Cretaceous (105 Ma) granodioritic porphyry invaded and was extensively buried at the depth of this area, which was not only associated with the Zhongliao porphyry copper (molybdenum) deposit but also provided a heat source for the epigenetic hydrothermal Cu-Au deposits that resulted in large-scale Cu-Au mineralization [29][30][31][32][33]. However, the dacitic porphyry contributed to the Zijinshan Cu-Au deposit, while the granodioritic porphyry contributed to the Luoboling Cu-Mo ore [34].

Petrography
Nine samples were collected for studies of bulk-rock chemistry and U-Pb, Hf, and O isotopes of zircon. Microscopic photographs of representative samples are shown in Figure 2. The porphyritic rocks in the region were found to be mainly granitic porphyry, with a minor amount of granodioritic porphyry (sampe CX-5). The granitic porphyry had a porphyritic texture (Figure 2a,b), with the matrix of a microcrystalline-cryptocrystalline texture ( Figure 2d). We found that some materials of the matrix were devitrified (Figure 2c). The content of phenocrysts was found to be about 20-30%, mainly composed of alkali feldspar, plagioclase, quartz, and a small amount of biotite and amphibole. The alkali feldspar was in the shape of a euhedral-subhedral plate, exhibiting common kaolinization and, to a lesser degree, sericitization. The quartz was granular and normally eroded. Some quartz grains underwent late hydrothermal carbonate alteration (Figure 2e). The biotite was generally chloritized with an abnormal interference color (Figure 2d,h). The amphiboles were also strongly altered by carbonation and chloritization (Figure 2f). Sericitization was common in the plagioclase (Figure 2n). The granodioritic porphyry had a porphyritic structure with some phenocrysts fragmented and cut through by the late quartz veins (Figure 2k). The fine-grained matrix consisted of long strips of plagioclase that were arranged roughly in orientation. (a,b) granite porphyry with plagioclase and quartz phenocrysts; (c) felsic minerals matrix of the granitic porphyry developed devitrification phenomenon; (d) chloritization was developed along the cleavage fracture of biotite; (e) quartz phenocrysts of granitic porphyry were featured frequently with melting corrosion structure; (f) amphiboles were strongly altered with carbonation and chloritization; (g-k) granodioritic porphyry had a porphyritic structure, and plagioclase microcrystals were developed in the matrix; (i) quartz phenocrysts set in a fne-to micro-grained holocrystalline matrix; (m) quartz phenocryst with a typical recrystallized rim with micro-inclusions; (g-k) the matrix had a microcrystalline structure, and plagioclase was partially sericized; (n) biotite and plagioclase phenocrysts developed chloritization and sericization, respectively; and (o) plagioclase in the matrix had obvious sericization; abbreviations: Q: quartz; Pl: plagioclase; Kfs: K-feldspar; Bi: biotite; Amp: amphibole; Ser: Sericite; and Chl: chlorite.

Major and Trace Element Analysis
Whole-rock samples were analyzed for major and trace elements at the Institute of Regional Geology and Mineral Resources Survey in Hebei Province, China. The major elements were analyzed using the X-ray fluorescence (XRF) method with an analytical error of <2%. Rare earth and other trace elements were analyzed by using plasma mass spectrometry (ICP-MS) with an analytical error of <5%. (a,b) granite porphyry with plagioclase and quartz phenocrysts; (c) felsic minerals matrix of the granitic porphyry developed devitrification phenomenon; (d) chloritization was developed along the cleavage fracture of biotite; (e) quartz phenocrysts of granitic porphyry were featured frequently with melting corrosion structure; (f) amphiboles were strongly altered with carbonation and chloritization; (g-k) granodioritic porphyry had a porphyritic structure, and plagioclase microcrystals were developed in the matrix; (i) quartz phenocrysts set in a fne-to micro-grained holocrystalline matrix; (m) quartz phenocryst with a typical recrystallized rim with micro-inclusions; (g-k) the matrix had a microcrystalline structure, and plagioclase was partially sericized; (n) biotite and plagioclase phenocrysts developed chloritization and sericization, respectively; and (o) plagioclase in the matrix had obvious sericization; abbreviations: Q: quartz; Pl: plagioclase; Kfs: K-feldspar; Bi: biotite; Amp: amphibole; Ser: Sericite; and Chl: chlorite.

Major and Trace Element Analysis
Whole-rock samples were analyzed for major and trace elements at the Institute of Regional Geology and Mineral Resources Survey in Hebei Province, China. The major elements were analyzed using the X-ray fluorescence (XRF) method with an analytical error of <2%. Rare earth and other trace elements were analyzed by using plasma mass spectrometry (ICP-MS) with an analytical error of <5%.

Zircon U-Pb Dating
Zircons from about 20 kg of the sample were first separated using magnetic and heavy liquid separation methods, and then it was hand-picked under a binocular microscope. Zircon grains and a standard sample (Qinghu, Plésovice and 91500) were stuck onto the target using epoxy resin, and gold was sprayed on the surface after polishing the internal zircon. The morphology and internal structure of the zircons were studied using the cathodoluminescence (CL) technique with a JXA-8100 Electron Probe Microanalyzer for high-resolution imaging and spectroscopy at the State Key Laboratory of Continental Tectonics and Dynamics in Chinese Academy of Geological Sciences. The zircon in-situ O isotopic ratios were analyzed first, and the U-Pb and Hf isotope were subsequently analyzed at the same point.
Zircon U-Pb and Pb-Pb dating were conducted using a Cameca IMS 1280 ion microprobe (secondary-ionization mass spectrometry (SIMS)) at the Institute of Geology and Geophysics, Chinese Academy of Sciences (CAS) in Beijing. The detailed method of the analysis can be found in [18]. A single O 2 − ion beam (20 × 30 µm in size with 10 nA of strength) bombarded the sample surface with an acceleration voltage of −13 kV. Oxygen flow increased the sample chamber pressure to approximately 5 × 10 −6 T in order to improve the sensitivity of the Pb + ion in zircon. Seven sets of data were analyzed for each sample spot, and the measuring time was approximately 12 min. The obtained standard deviation of 1.5% (1σ) through long-term monitoring of the standard sample [34] and the internal test accuracy of single sample points together determined the single point error. Standard zircon 91500 (1065 Ma) was used to correct the Pb/U ratios and calculate the Th and U contents of the zircon samples (Th = 29 µg/g; U = 81 µg/g) [35]. Ordinary Pb was corrected by measuring 204 Pb. The analysis error of the isotope ratio and age for a single point was 1σ. The data were processed using the ISOPLOT software [36].

O Isotope Analysis
Zircon in-situ O isotopes were also analyzed by using the Cameca IMS 1280 ion microprobe (CASIMS) at the Institute of Geology and Geophysics at the CAS. A detailed analysis method can be found in [37]. A single 133 Cs + ion beam with a strength of 2 nA bombarded the sample surface and then received both 16

Hf Isotope Analysis
In-situ zircon Lu-Hf isotopic measurements were performed on a Neptune Plus Multi Collector -ICP-MS, coupled with a 193 nm resolution laser ablation system at the Institute of Geology and Geophysics at the CAS. A detailed description of analytical methods can be found in [38]. The time of laser ablation was 26 s. The laser ablation pulse frequency was 10 Hz, and the pulse power was 100 mJ, with a beam spot size of approximately 60 µm. The isobaric interference of 176 Lu and 176 Yb to 176 Hf was rectified by monitoring the signal strength of 175 Lu and 172 Yb using the 175 Lu/ 176 Lu ratio (0.02655) and the 176 Yb/ 172 Yb ratio (0.5886), respectively [39]. Standard zircon 91500 crossing with zircon sample analysis was used for the external monitoring of instrument drift. In the process of the analysis, the average 176 Hf/ 177 Hf ratio for the zircon standard Mud Tank was 0.282491 ± 27 (2σ), and it was 0.281971 ± 26 (2σ) for GJ, which was consistent with the results of the literature within the error range [40]. The

Zircon SIMS U/Pb Age
The analyzed zircons were mostly colorless and transparent crystals, ranging from semi-automorphic to idiomorphic, with a length of 100-400 µm and varying length/width ratio from 1:1 to 4:1. The CL images clearly showed that the internal structure of the zircons was oscillatory zoning (Figure 3). No influence of alteration on these zircon grains was observed. The Th/U ratios of the zircons were generally >0.1 (Supplementary Table S1), which is typical of magmatic zircons [41].

Zircon SIMS U/Pb Age
The analyzed zircons were mostly colorless and transparent crystals, ranging from semi-automorphic to idiomorphic, with a length of 100-400 µm and varying length/width ratio from 1:1 to 4:1. The CL images clearly showed that the internal structure of the zircons was oscillatory zoning (Figure 3). No influence of alteration on these zircon grains was observed. The Th/U ratios of the zircons were generally >0.1 (Supplementary Table S1), which is typical of magmatic zircons [41]. The detailed dating results are listed in Supplementary Table S1. Because of errors in counting statistics during the analysis, the 207 Pb/ 206 Pb ages were more accurate for older (>1 Ga) zircons, whereas the 206 Pb/ 238 U ages were more accurate for younger zircons [42]. Therefore, we use the 206 Pb/ 238 U ages for younger zircons in the following discussion.
Fifteen analyses on zircon grains from sample CX-9 yielded concordant 206 Pb/ 238 U dates ranging from 136.9 ± 2.0 to 141.  The detailed dating results are listed in Supplementary Table S1. Because of errors in counting statistics during the analysis, the 207 Pb/ 206 Pb ages were more accurate for older (>1 Ga) zircons, whereas the 206 Pb/ 238 U ages were more accurate for younger zircons [42]. Therefore, we use the 206 Pb/ 238 U ages for younger zircons in the following discussion.

Characteristics of Major and Trace Elements
The whole-rock SiO2 contents of the sample ranged from 62.46 to 78.09%. The relatively high SiO2 contents are usually combined with low Fe2O3 T , MgO, and CaO (Supplementary Table S2). The porphyry were characterized by low-to-high concentrations of Na2O between 0.09% and 4.27%, high K2O and Na2O between 4.24% and 8.98%, MgO of 0.1-3.6 % with most lower than 1%, and Al2O3 between 11.63 and 15.92%. In the SiO2 versus K2O and Na2O diagram, most samples plotted in the granite and granodiorite fields (Figure 5a). They had high K2O contents (3.27-8.85%), and all samples were located in the subalkaline and high-K calc-alkaline fields (Figure 5b

Characteristics of Major and Trace Elements
The whole-rock SiO 2 contents of the sample ranged from 62.46 to 78.09%. The relatively high SiO 2 contents are usually combined with low Fe 2 O 3 T , MgO, and CaO (Supplementary Table S2).  showing the highest contents of Yb and Lu (7.13 and 1.15 ppm, respectively) and the CX-2 (age = 99.56 ± 0.86 Ma) showing the lowest (1.04 and 0.24 ppm, respectively). Strongly negative Eu anomalies could be observed in samples CX-9 and SMS-1, indicating that they may have been formed in low oxygen fugacity and consequently experienced the significant crystallization of plagioclase in the magma source area.
The trace element abundance diagrams of the porphyritic rocks display roughly similar patterns with negative anomalies of Ba, Nb, Sr, P, and Ti anomalies combined with small positive Rb, Th, U, La, Ce, Nd, and Hf anomalies (Figure 5d). The enrichment in LILEs and depletion in HFSEs in the porphyry samples indicate the influence of fluids in subduction conditions. The significant negative Ba anomalies are thought to be the result of alkali feldspar-dominated fractionation. Positive Rb and negative Sr anomalies, both of which are associated with negative Eu anomalies, are consistent with dominant plagioclase fractionation. The significant Ti depletions were due to magnetite and ilmenite fractionation.
As the alteration of the Mesozoic granite samples in the Zijinshan area is strong, its major and trace elements may not be reliably used to describe the characteristics of the source region.

O-Isotope Composition of Zircons
The zircon oxygen isotopic data are listed in Supplementary Table S3 and shown in Figure 6.

Hf-Isotope Compositions of Zircons
Lu-Hf isotopic ratios were determined on the same or similar domains as the zircon grains analyzed for U-Pb and O isotopes. The analytical results are listed in Supplementary Table S3

Hf-Isotope Compositions of Zircons
Lu-Hf isotopic ratios were determined on the same or similar domains as the zircon grains analyzed for U-Pb and O isotopes. The analytical results are listed in Supplementary Table S3

Magmatic Source
Zircon has high closure temperatures for U, Pb, O, and Hf; thus, it is less susceptible to mixing the overlying crustal material during magma ascent. Therefore, these isotopic signatures could represent the geochemical characteristics of the original magma.
As discussed before, the ages of the Zijinshan porphyritic rocks were divided into three groups: Group 1:~138.8 Ma; Group 2: 109.2-107.4 Ma; and Group 3: 99.7-98.2 Ma. In Group 1, the zircon δ 18 O values of sample CX-9 ranged from 6.96 to 8.17‰ (Figure 6), which was higher than the primitive mantle value (5.3 ± 0.3‰) [19,20].   (Figure 8). Therefore, it is possible that the Group 3 rocks had a mixed source composed of Group 1 and Group 2 rocks. In Group 3, the influence of mantle-derived magma was less than that in Group 2, and there was an increasing tendency indication of the crustal source. The δ 18 O and εHf(t) values also showed the mixing characteristic of Group 1 and Group 2 magma.
In summary, Group 2 was more affected by mantle-derived magmas. This conclusion is consistent with previous studies arguing that the late Early Cretaceous was an important ore-forming period, and some ore-forming elements were derived directly from the mantle. It has been widely accepted that Cretaceous magmatic rocks formed in an extensional tectonic setting and are probably related to subduction of the Western Pacific plate [4,14].

Regional Magmatic Evolution
Combined with the previous data of the SIMS U-Pb age of magmatic zircons, it can be summarized that the magmatic activities in the southwestern Fujian Province can be divided into eight main periods. The zircon O and Hf data indicated that the different episodes of magmatic rocks had different δ 18 O and εHf(t) signatures, reflecting different magma sources and tectonic environments [43,44]. Following is a summary of the magmatic zircons categorized by period. Some of them were captured by granite and mafic dikes, some samples were directly crystallized from intermediate-acid magmatic rocks, and the rest were sampled from the metamorphose basement of Precambrian.

1.
Paleoproterozoic (2467~1950 Ma) Zircons of Early Proterozoic ages were collected from basaltic dikes [41]. The measured δ 18 O values (4.42-8.09‰; average 5.81‰) of these zircons were close to the δ 18 O value of the zircons crystallized from mantle-derived magmas [19]. The εHf(t) values of these zircons were mainly between 2.2 and −7.9, falling within the depleted mantle and average crust. The Hf two-stage depleted mantle model ages (T DM2 ) were mainly in the Archaeozoic (Figure 8). This indicated that the magmatism was derived from the partial melting of an ancient depleted mantle and an Archaeozoic crust.

2.
Mesoproterozoic (1514~933 Ma) The Mesoproterozoic magmatic zircons from our study displayed a large variation both in Hf isotopic compositions (εHf(t) = from 4.5 to −15.5) and in oxygen isotopic compositions ranging from a lower than the normal mantle value of 3.79‰ to supracrustal component of 9.14‰. Most of the Mesoproterozoic zircons yielded Paleoproterozoic T DM2 ages (1.7-2.5 Ga). These characteristics implied that the Mesoproterozoic magmatism was generated mainly by the partial melting of the Paleoproterozoic crustal materials and mixed with depleted mantle-derived magmas.

3.
Middle Neoproterozoic (705~647 Ma) Only five captured magmatic zircon grains were dated during this period. Their values of δ 18 O and εHf(t) varied in a limited range. This suggests that, from the Late Proterozoic to Pre-Silurian, there was probably no significant magmatism in this area.

4.
Silurian to Lower Devonian (443~395 Ma) Zircons analyzed from this period showed relatively low εHf(t) values and plotted along the mafic crustal evolution line (Figure 8). The δ 18 O values of the zircons were higher than those of the others (Figure 7). The calculated Hf T DM2 ages were within the Paleoproterozoic (Figure 8). This suggested that magma mainly came from the partial melting of the Paleoproterozoic mafic crust and mantle materials had a smaller contribution.

5.
Permian to Triassic (277~211 Ma) In this period, the εHf(t) and δ 18 O values of zircons varied over a wide range. This suggested magmatic mixing between the depleted mantle and mafic crust. However, the contribution of mantle material to magma was more significant than that of the previous Silurian magmas. The age of the crustal source was mainly Paleoproterozoic, with a minor contribution from the Middle Proterozoic crust.

6.
Middle Jurassic to Early Lower Cretaceous (174~137 Ma) Late Jurassic granites were found to be widely distributed in the Zijinshan area. The magmatic zircons generally showed εHf(t) values of <0 and δ 18 O values of >6. Zircon Hf T DM2 ages were mainly from the Paleoproterozoic. This was the result of mixing of the mantle source material and the molten magma of the Paleoproterozoic crust.

7.
Late Lower Cretaceous (118~102 Ma) Cretaceous shallow intrusive and volcanic rocks were found to be widely distributed in the Zijinshan area, with limited variations in εHf(t) and δ 18 O values. This indicated that the composition of the source region was relatively homogeneous. The low δ 18 O and high εHf(t) values suggested that the strongest mantle-derived magma intrusion occurred in this area in the Paleozoic. For all the magmatic activity in the region, only zircons from this period showed that the crustal component of the magma source was from the partial melting of the Middle Proterozoic strata, as shown by their Hf T DM2 ages ( Figure 8). This period was an important Cu-Au-Ag mineralization period in the Zijinshan region. Both mantle-derived materials and materials derived from the Middle Proterozoic crust probably contributed to large-scale mineralization.

8.
Late Lower Cretaceous to Early Upper Cretaceous (101~92 Ma) The compositional ranges of εHf(t) and δ 18 O values were also relatively limited similar to those observed in the late Early Cretaceous zircons. The limited variation of zircon εHf(t) and δ 18 O values in the Cretaceous felsic magmatism implied that the lower crust had been transformed into a more uniform lower crust. The zircon Hf T DM2 in this period, as well as εHf(t) and δ 18 O values, fell into the range between the Late Jurassic and Earliest Late Cretaceous.
In summary, the Paleoproterozoic crust was the main magmatic source for Mesoproterozoic to Jurassic magmatism ( Figure 8). For pre-Mesoproterozoic magmatism (in period 1), crustal materials came from the partial melting of both the Archaeozoic and Paleoproterozoic crusts. Mantle-derived magmas played an important role in the formation of the Mesoproterozoic and earlier magmatism. From Mesoproterozoic to lower Devonian, the role of crustal melting in magmatic genesis gradually increased at the expense of the mantle component (Figures 7 and 8). However, the contribution of mantle-derived materials increased in the magmatism from the Middle Jurassic to the late Lower Cretaceous (from period 6 to period 7). Intensive crust-mantle interactions may have caused a higher degree of the partial melting of the Mesoproterozoic crust. Mantle and Mesoproterozoic crustal materials provided sufficient ore-forming components and eventually led to the formation of large-scale deposits in the region. Period 8 was a continuation of Late Early Cretaceous magmatism. At this time, the mantle source invasion was reduced, the crust source became more homogeneous, and some mineralization continued.

Conclusions
(1) The ages of Mesozoic porphyritic rocks in the Zijinshan ore field region could be divided into three groups: Group 1:~138.8 Ma; Group 2:~109.2-107.4 Ma; and Group 3:~99.7-98.2 Ma. Mantle-derived magma made a significant contribution to the Group 2 porphyries.
(2) Based on all the available SIMS U-Pb age data of the zircons, the medium-acidic magmatism in the southwestern Fujian province could be divided into eight periods: Paleoproterozoic, Mesoproterozoic, Middle Neoproterozoic, Silurian to Lower Devonian, Permian to Triassic, Middle Jurassic to Early Lower Cretaceous, Late Lower Cretaceous, and Late Lower Cretaceous to Early Upper Cretaceous.
(3) Paleoproterozoic crust was the main magmatic source for Mesoproterozoic to Jurassic magmatism, but only magmatism that was closely related to mineralization was derived from partial melting of the Mesoproterozoic crust and a more depleted mantle.
(4) Intensive crust-mantle interactions may have caused a higher degree of partial melting of the Mesoproterozoic crust. Mantle and Mesoproterozoic crustal materials provided sufficient ore-forming components and eventually led to the formation of large-scale deposits in the region.
Supplementary Materials: The following are available online at http://www.mdpi.com/2075-163X/10/12/1143/s1, Supplementary Table S1: U-Pb zircon dating results of the porphyritic rocks from the Zijinshan area; Table S2: Whole-rock geochemical data of the porphyritic rocks from the Zijinshan area; and Table S3: Hafnium and oxygen isotopes of the porphyritic rocks from the Zijinshan area.