Growth of plutons by floor subsidence: implications for rates of emplacement, intrusion spacing and melt-extraction mechanisms

Geophysical and field-based studies indicate that granitic plutons occur as either tabular (disk) or wedge (funnel) shapes whose length (L) to thickness (T) ratio is controlled by the empirical power law, T = 0.6(±0.15)L^0.6(±0.1). The dimensions of plutons are self-similar to other natural subsidence phenomena (calderas, ice cauldrons, sinkholes, ice pits) and it is proposed that they grow in a similar fashion by withdrawal of material (melt) from an underlying source, which is then transferred to the growing pluton within the crust. Experimental studies show that growth of subsidence structures occurs by vertical inflation ⪢ horizontal elongation of an initial depression with L ≈ width of the source region. If pluton growth is modelled in the same way, the empirical power law relating T and L defines limits for pluton growth that are imposed by the width, thickness and degree of partial melting from a lower crustal source. Several growth modes that predict testable internal structural patterns are identified for plutons, depending on whether they are tabular or wedge-shaped, grow by continuous or pulsed magma delivery and whether magma is accreted from bottom to top, or vice versa. Rates of pluton growth are geologically fast (hundreds to hundreds of thousands of years) if magma supply is effectively continuous, but can also take millions of years if the time between magma delivery events is much longer than magma injection events. Plutons formed by melt extraction from an area directly beneath require large degrees of partial melting and or very thick sources. Lower degrees of partial melting and thinner sources are permitted when melt extraction occurs over a larger region, which can lead to the formation of spaced plutons. Tabular pluton growth will tend to favour widely spaced plutons, unless degrees of partial melting in the source are high. Wedge-shaped plutons can form much closer together and require lower degrees of partial melting. These results are in general agreement with current geophysical, petrological and experimental estimates of partial melting in the lower continental crust.