More than 19,000 analytical data mainly from the literature were used to study statistically the distribution patterns of F and the oxides of minor and trace elements (Ti, Sn, Sc, V, Cr, Ga, Mn, Co, Ni, Zn, Sr, Ba, Rb, Cs) in trioctahedral micas of the system phlogopite-annite/siderophyllite-polylithionite (PASP), which is divided here into seven varieties, whose compositional ranges are defined by the parameter mgli (= octahedral Mg minus Li). Plots of trace-element contents vs. mgli reveal that the elements form distinct groups according to the configuration of their distribution patterns. Substitution of most of these elements was established as a function of mgli. Micas incorporate the elements in different abundances of up to four orders of magnitude between the concentration highs and lows in micas of ‘normal’ composition. Only Zn, Sr and Sc are poorly correlated to mgli. In compositional extremes, some elements (Zn, Mn, Ba, Sr, Cs, Rb) may be enriched by up to 2–3 orders of magnitude relative to their mean abundance in the respective mica variety. Mica/melt partition coefficients calculated for Variscan granites of the German Erzgebirge demonstrate that trace-element partitioning is strongly dependent on the position of the mica in the PASP system, which has to be considered in petrogenetic modelling.

This review indicates that for a number of trace elements, the concentration ranges are poorly known for some of the mica varieties, as they are for particular host rocks (i.e. igneous rocks of A-type affiliation). The study should help to develop optimal analytical strategies and to provide a tool to distinguish between micas of ‘normal’ and ‘abnormal’ trace-element composition.

In order to limit the range of possible differentiation mechanisms and the impact of these processes on the trace element signatures of igneous suites, it is important to be able to predict or model the compositional evolution of the primary and/or parental magmas. Part of the problem in understanding these relationships in basalts from ocean islands is the paucity of reliable trace element mineral/melt partition coefficients, particularly for undersaturated magmas. Consequently, we have measured mineral/groundmass partition coefficients for Rb, Sr, Y, Zr, Nb, Hf, Ta, Th and REE, in situ, in clinopyroxene, Fe-Ti oxides and olivine in primitive basalts from Rarotonga, Cook Islands using LAM-ICP-MS.

Analyses of these mafic rocks show high concentrations of most incompatible trace elements (e.g. Sr, Th, Y, REE, etc.) in pyroxenes relative to the other phases, suggesting that fractionation of pyroxenes was the dominant factor in the distribution of these trace elements during crystal fractionation. In such cases, the highly incompatible elements (D <0.01) are Ba, Cs and Nb. Elements that could also be classified as strongly incompatible (D <0.1) are Rb, Ta and Th. The remaining trace elements have bulk partition coefficient values that range from ∼.0.1 up to ∼.0.8 (Sr, Hf, Zr, Y and REE). Magnetites incorporate greater amounts of Nb and Ta than the titanaugites, and any significant fractionation of magnetite would have affected the bulk distribution of Nb and Ta. The LREE have lower Kd values than other REE, with the HREE having Kd values close to unity. Consequently, with the separation of titanaugite an overall enrichment of REE in the residual liquid with an increase in the La/Yb ratio is produced.

The behaviour of trace elements during partial melting depends primarily on their mode of occurrence. For elements occurring as trace constituents of major phases (e.g. Li, Rb, Cs, Eu, Sr, Ba, Ga, etc.), slow intracrystalline diffusion (D ≍ 10−16 cm2 s−1) at the temperature range of crustal anatexis causes all effective crystal-melt partition coefficients to have a value close to unity and impedes further melt-restite re-equilibration. Usually, therefore, the trace element composition of crustal melts simply depends on the mass balance between the proportion and composition of phases that melt and the proportion and composition of newly formed phases. The behaviour of trace elements occurring as essential structural components in accessory phases (e.g. P, La-Sm, Gd-Lu, Y, Th, U, Zr, Hf, etc.) depends on the solubility, solution kinetics, grain size and the textural position of accessory phases. In common crustal protoliths a significant mass fraction of monazite, zircon, xenotime, Th-orthosilicates, uraninite; etc.—but not apatite—is included within other major and accessory phases. During low melt fraction anatexis the amount of accessory phases available for the melt is not sufficient for saturation, thus producing leucosomes with concentrations of La-Sm, Gd-Lu, Y, Th, U and Zr lower than expected from solubility equations. Low concentrations of these elements may also occur if the melt is prevented from reaching equilibrium with the accessories due to fast segregation. However, the first mechanism seems more feasible as leucosomes that are undersaturated with respect to monazite and zircon are frequently saturated, even oversaturated, with respect to apatite.

Melting experiments with and without added H2O on a model metagreywacke and a natural metapelite demonstrate how pressure and H2O content control the compositions of melts and residual assemblages. Several effects are observed under isothermal conditions. Firstly, the stability field of biotite shrinks with decreasing pressure and with increasing H2O content, whereas that of plagioclase shrinks with increasing pressure and H2O content. Secondly, the ferromagnesian content of melts at the source (i.e. coexisting with their residual assemblages) decreases with decreasing H2O activity. Thirdly, with increasing pressure the Ca/Mg and Ca/Fe ratios of melts decrease relative to those of coexisting garnet. As a consequence, a wide spectrum of melts and crystalline residues can be generated from the same source material. For example, H2O-starved dehydration melting of metagreywacke at low pressure (≤10 kbar) generates K-rich (granitic) melts that coexist with pyroxene- and plagioclase-rich residues, whereas melting of the same material at high pressure (≍15 kbar) and with minor H2O infiltration can generate leucocratic Na-rich and Ca-poor (trondhjemitic) melts that coexist with biotite- and garnet-rich residues. An increased H2O content stabilises orthopyroxene at the expense of garnet + biotite + plagioclase, causing melts to shift towards granodioritic or perhaps tonalitic compositions.

Granites within suites share compositional properties that reflect features of their source rocks. Variation within suites results dominantly from crystal fractionation, either of restite crystals entrained from the source, or by the fractional crystallisation of precipitated crystals. At least in the Lachlan Fold Belt, the processes of magma mixing, assimilation or hydrothermal alteration were insignificant in producing the major compositional variations within suites. Fractional crystallisation produced the complete variation in only one significant group of rocks of that area, the relatively high temperature Boggy Plain Supersuite. Modelling of Sr, Ba and Rb variations in the I-type Glenbog and Moruya suites and the S-type Bullenbalong Suite shows that variation within those suites cannot be the result of fractional crystallisation, but can be readily accounted for by restite fractionation. Direct evidence for the dominance of restite fractionation includes the close chemical equivalence of some plutonic and volcanic rocks, the presence of plagioclase cores that were not derived from a mingled mafic component, and the occurrence of older cores in many zircon crystals. In the Lachlan Fold Belt, granite suites typically evolved through a protracted phase of restite fractionation, with a brief episode of fractional crystallisation sometimes evident in the most felsic rocks. Evolution of the S-type Koetong Suite passed at about 69% SiO2 from a stage dominated by restite separation to one of fractional crystallisation. Other suites exist where felsic rocks evolved in the same way, but the more mafic rocks are absent. In terranes in which tonalitic rocks formed at high temperatures are more common, fractional crystallisation would be a more important process than was the case for the Lachlan Fold Belt.