Attribution: You must attribute the work in the manner specified by the author or licensor (but no in any way that suggests that they endorse you or your use of the work).Noncommercial ‒ you may not use this work for commercial purpose.No Derivative works ‒ You may not alter, transform, or build upon this work.Sharing ‒ Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in other subsequent works and to make unlimited photocopies of items in this journal for noncommercial use in classrooms to further education and science.

Most volcanism on Earth is related to plate tectonics. Hidden at the bottom of the oceans, mid-ocean ridges form where oceanic plates drift apart. More visible, and explosively dangerous, is volcanism related to the subduction of oceanic plates back into the Earth's mantle, e.g., the Pacific Ring of Fire. Less clear is the origin of the third major type of volcanism, the so-called intraplate volcanism, favorably known in creating popular holiday destinations such as Hawaii. In the early days of plate tectonics, a model of narrow, thermally driven upwellings independent of plate tectonics was proposed, so-called mantle plumes (Morgan, 1971), which explained the formation of intraplate ocean islands and seamount chains in the direction of plate motion, sometimes connected to flood basalt provinces. However, the existence of mantle plumes has been hotly debated in recent years, fueled by the lack of unequivocal evidence for their existence (e.g., DePaolo and Manga, 2003; Foulger and Natland, 2003; Campbell, 2007). What could cool off this debate?

Some hot evidence could help! If it could be shown that mantle plumes are hotter than ambient mantle, this would prove their existence as thermal upwellings. In this issue, Putirka (2008, p. 283) argues that temperatures based on olivine/melt thermometers for 28 oceanic hotspots indicate excess temperatures above that of the ambient mantle, and therefore thermal mantle plumes exist, and in fact are rather common. Olivine thermometry is not new, and has been used by many to argue for and against thermal mantle plumes (e.g., reviews by Falloon et al., 2007; Putirka et al., 2007). Such opposite conclusions, much to the confusion of the broad science community, are drawn from mantle temperature estimates based on only slightly different olivine-liquid-Fe-Mg partition coefficients, FeO or MgO contents of magmas, and olivine forsterite contents. So what is different about Putirka's approach relative to others? In short, studies other than Putirka's try to constrain the MgO content of the magmas that last equilibrated with the mantle. When combined with olivine compositions that equilibrated with such parental magmas (or assumed mantle olivine compositions), a minimum temperature for the mantle source can be derived. However, MgO contents are highly variable, depending on amounts of olivine fractionation or accumulation, and lavas frequently represent mixtures of more evolved lavas with olivines possibly remobilized from different magma batches. More or less sophisticated approaches to get around this problem have been suggested (e.g., Courtier et al., 2007; Herzberg et al., 2007), but many uncertainties remain. Putirka (2008 and 2005) elegantly avoided the uncertainties related to the MgO content by using the FeO content of lavas. This is a clever approach because along olivine control lines, which dominate the early fractionation of basaltic magmas, FeO is basically constant, and therefore we know the FeO content of parental magmas much better than their MgO content. Putirka combines the FeO content with maximum forsterite content of olivines, and can demonstrate that melts related to mantle plumes are significantly hotter than melts from the ambient mantle. This first-order result is independent of source composition, which is important because recent work has confirmed the long-suspected significance of lithologic variability in mantle plumes (Sobolev et al., 2005). It is significant that Putirka's work, as well as studies from two other groups (Courtier et al., 2007; Herzberg et al., 2007), converge to similar excess temperatures for mantle plumes. The translation of excess temperatures into mantle temperatures requires additional thermodynamic modeling, where some agreement within the petrology community on the best approach would be helpful. Significantly, Putirka's excess temperatures correlate with buoyancy flux estimated from hotspot swell topography (Sleep, 1990). Although the correlation is dominated by the two points for Hawaii, it demonstrates consistency between petrologic and geophysical constraints on excess temperature. Thus, plumes are hotter than ambient mantle; but will this solid conclusion cool off the plume debate? Probably not, because inferred plume temperatures are at best lukewarm relative to the much larger temperature gradient constrained for the core-mantle boundary and given the significant temperature variations of the ambient mantle. Probably only improved seismic imaging will be able to provide the ultimate proof for the existence of mantle plumes.

Based on the constraint that thermal upwellings exist, Putirka (2008) suggests that the mantle plume excess temperatures, combined with helium isotope evidence, place additional constraints on the controversial question of whole versus layered mantle convection. Despite many contributions on the subject (including our own advocating whole mantle convection; Class and Goldstein, 2005), the question of geodynamic regimes in the mantle is far from being resolved. Helium is an important tracer for mantle dynamics. Still persistent in the literature (e.g., Courtillot et al., 2003), high 3He/4He isotope ratios are considered a prerequisite for the presence of a mantle plume from the lower mantle, despite the early notion by Kurz et al. (1982) that so-called “low 3He/4He” plumes exist (“low” indicates that 3He/4He ratios are lower than in mid-oceanic ridge basalts [MORB]). Ocean island basalts (OIB) usually show a range in He isotope ratios, from MORB values to either higher or lower 3He/4He than MORB, which suggests the contribution of upper mantle helium to most OIB, where only the extreme He isotope ratios are the ones representative of the related plume source. Based on this interpretation, Class and Goldstein (2005) grouped ocean islands by their “extreme 3He/4He” isotope ratios relative to MORB, and found that OIB form a continuum in “extreme” 3He/4He-206Pb/204Pb-208Pb/204Pb-Th space. Low 3He/4He plumes are more enriched in Th+U, thus the range in “extreme 3He/4He” isotope ratios most likely reflects the variable production rate of radiogenic 4He (α-particle) from Th and U decay. Thus, helium isotopes alone are not diagnostic for plumes, and no correlation between mantle plume excess temperature (Tex) and 3He/4 Hemax should be expected as long as plume sources are externally heated, e.g., in a boundary layer.

Nevertheless, Putirka finds a positive correlation between Tex and 3He/4Hemax and suggests that this supports a layered mantle with high 3He/4He most likely stored within a lower thermal boundary layer. Unfortunately, Putirka's analysis of global data compiled in GeoRoc (http://georoc.mpch-mainz.gwdg.de/georoc/) suffers from the general problem that integrated data sets are incomplete, especially when combining information on the standard Sr-Nd-Pb isotope systems with noble gases, since traditionally noble gas studies were performed on separate samples. Maximum values for 3He/4He measured for Hawaii, Iceland, Galapagos, and Samoa are ~35 RA (RA is the atmospheric ratio of 1.4 × 10−6) (compare recent compilation by Jackson et al., 2007), and with Tex ranging from 130 to 290 K this eliminates the proposed positive correlation between Tex and 3He/4Hemax. Does this lack of a correlation exclude the single hetero geneous layer model with a low-T, high 3He/4He component (Ito and Mahoney, 2006) and thus support a layered mantle? Putirka argues. that such a single heterogeneous layer should produce a negative correlation between Tex and 3He/4Hemax. However, because enriched components contribute radiogenic 4He, their presence in variable proportions in such Tex and 3He/4 Hemax, and this is a model will allow for any combination of exactly what we observe. Maybe there is information on the convective Tex and 3He/4 Hemax variation, but prior regimes of the mantle hidden in the to its extraction, first the dilution by radiogenic 4He needs to be unraveled. Nevertheless, studies such as the one of Putirka are a start in the right direction, as neither petrologic temperature estimates nor geochemistry alone seem capable of resolving the question of the geodynamic regimes of the mantle, but their combination with geophysical evidence are most promising in providing future answers.