The Longevity of Geochemical Differences Between Lavas Erupted Along the Loa and Kea Spatial Trends Defined by the Hawaiian Islands Frederick A. Frey ([email protected]) (MIT), Guangping Xu (Colorado State University), Wafa Abouchami (MPI), Janne Blichert-Toft (Ecole Normale Superieure de Lyon), David A. Clague (MBARI)
Abstract 2. Trace element characteristics 4. Pb isotopic compositions on Molokai Island 5. Hawaiian mantle plume zonation Differences in Pb isotope ratios between lavas from Loa- and 38.3 15.51 25 East Molokai West Molokai A 15.50 Kea-trend volcanoes have been used to infer geochemical zonation a D10-R1 Fig. 2 Primitive mantle-normalized 38.2 E K D10-R2 Penguin Bank T307R6,7,8 in the Hawaiian plume. However, the significance of Kea D10-R4 abundances of incompatible elements MVD10 15.49 a random heterogeneity D10-R5 38.1 partly ordered b D10-R6 with the element incompatibility HIG1972-D4 geochemical signatures in Loa-trend volcanoes and vice versa is 20 15.48 b P5-255-1 b 38.0 P5-255 P P 4 P5-255-2 decreasing from right to left; The two 4 0 0 2 2 / / 15.47 debated. The longevity of these differences is also uncertain; Weis et P5-255-3 HIG1972-D4 b b P P5-255-5 HIG dredge samples have markedly P 37.9 7 8 9, 10 0 0 2 al. [1] claim that geochemical zonation has persisted “for at least 5 P5-255-6 lower abundances. Relative to primitive 2 and 11 15.46 T307R6,7,8 P5-255-7 37.8 M 15 P5-255-8 T307R2 15.45 million years”, but Abouchami et al. [2] suggested that the P mantle notable anomalies are high / P5-255-9
e HIG1972-D4 l rejuvenated p P5-255-10 Ba/Th, low Nb/Ta, low La/Ce and high 37.7 A geochemical differences “terminate at the Molokai Fracture Zone” O stage lavas 15.44 m P5-255-11 L
a a T307R2 b
s P5-255-14 Sr/Nd. (~2 Ma). We have determined the geochemical characteristics of 37.6 15.43 P5-255-15 17.8 17.9 18.0 18.1 18.2 18.3 18.4 18.5 18.6 18.7 10 P5-255-16 17.8 17.9 18.0 18.1 18.2 18.3 18.4 18.5 18.6 18.7 206 204 three contemporaneous volcanoes whose strike crosses that of the HIG-1972-4-1 206Pb/204Pb Pb/ Pb HIG-1972-4-2 Loa- and Kea-trends, specifically late shield and postshield lavas, 15.51 15.70 c KEA LOA East Molokai d ~1.5 to 1.75 Ma, from East Molokai volcano, the northernmost Kea 15.50 West Molokai 5 Mauna Kea 15.65 15.49 (mid8) MVD10 trend volcano recognized by Jackson et al. [3], shield and postshield HIG1972-D4 Kea Trend 15.48 P5-255 15.60 Kea Trend lavas, ~1.7 to 1.9 Ma, from West Molokai volcano, a Loa trend Mk Loihi b -h P b i8
4 Loa Trend
P 15.47 Loa Trend 4 0
0 0 2 2 / 15.55 volcano [4,5], and shield lavas, undated but inferred to be 1.8 to 2.0 / Ba Th Nb Ta La Ce Pb Nd Sr Sm Zr Hf Eu Ti Tb Dy Ho Y Er Tm Yb Lu b b
P 15.46 7 P 0 2 Ma, from the submarine Penguin Bank volcano which we infer to be 7 15.45 KSDP 0 2 15.50 17 Loa Shield a distinct Hawaiian volcano west of the Loa trend. These Loa- and Fig. 3 Zr/Nb vs. Sr/Nb showing that Loa-like 15.44 Kea Shield Mauna Loa rejuvenated Fig. 8 Two models for explaining Loa and Kea trend geochemical differences (from Fig. 2 of Xu P5-255 stage 15.45 Samoan lavas Kea-trend volcanoes are distinguished by incompatible element high ratios are typical of all Penguin Bank 15.43 T307R2 et al., 2007). (a) The partly ordered zonation model (e.g., Herzberg, 2005) has the proportion of MVD10 Icelandic lavas with Koolau unradiogenic Pb Icelandic lavas samples except for samples P5-255-9, -10 and - (Makapuu) abundance ratios (Zr/Nb) as well as isotopic ratios of Sr, Nd, Hf and 15 HIG1972-D4 15.42 15.40 randomly distributed heterogeneities (in red) varying systematically from the Loa to Kea parts of 11. Fields for lavas from Mauna Kea and < 100 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99 0.91 0.94 0.97 1.00 1.03 1.06 Pb. East Molokai lavas are clearly Kea-like whereas West Molokai 208Pb*/206Pb* the plume. (b) The randomly distributed heterogeneity model has no systematic zonation (Huang ka lavas from Mauna Loa are shown for 208Pb*/206Pb* and Penguin Bank lavas range from Loa-like to Kea-like. This result and Frey, 2005a, Ren et al., 2005, 2009). As drawn, Loa volcanoes sample a higher proportion of 13 9,10,11 comparison. Only lavas with major element data Fig. 5 (a) 208Pb/204Pb vs. 206Pb/204Pb for East Molokai, West Molokai and Penguin Bank lavas. heterogeneities with low solidii, perhaps recycled oceanic crust and sediment (red color), is consistent with more geochemical heterogeneity in Loa volcanoes b N and MgO >6.5% are included to minimize the The heavy diagonal line in panel is the Loa-Kea dividing line proposed by Abouchami et al. / whereas Kea volcanoes, which are closer to the plume center, sample a higher proportion of r than in Kea volcanoes [1,2]. Z effect of mineral fractionation on Sr/Nb and (2005). All East Molokai lavas are in the Kea field, but West Molokai range from the Kea-field 11 components (blue background) with high solidii such as peridotites. Zr/Nb ratios. to the Loa-field, and Penguin Bank lavas are in the Loa field except for samples P5-255-9, -10 These results are consistent with a randomly distributed and -11, which are in the Kea field as shown in Fig. 3; hence both trace element abundance and heterogeneity model with Loa volcanoes sampling a higher 9 Mauna Kea isotopic ratios distinguish Loa- and Kea-trend magmas. 6. Summary proportion of components with low solidii, perhaps recycled oceanic (b) 207Pb/204Pb vs. 206Pb/204Pb for East Molokai, West Molokai and Penguin Bank lavas. Three crust and sediment, whereas Kea volcanoes, which are closer to the samples from West Molokai and the two HIG samples from Penguin Bank are offset from a (1) Although Pb isotopic ratios have been emphasized in documenting the geo- 7 linear trend to low 207Pb/204Pb at a given 206Pb/204Pb. plume center, sample a higher proportion of components with high 12 16 20 24 28 32 36 40 44 chemical difference between volcanoes forming the Loa- and Kea- spatial trends Sr/Nb (c) 208Pb*/206Pb* vs. 207Pb/204Pb for East Molokai, West Molokai and Penguin Bank lavas. West solidii such as peridotites. Molokai and Penguin Bank lavas define a negative trend with Loa lavas at the high (Fig. 1a), we find that isotopic ratios of Sr, Nd and Hf and incompatible element 208Pb*/206Pb* and low 207Pb/204Pb end of the trend reflecting long term high Th/U and low abundance ratios such as Zr/Nb and La/Nb also distinguish Loa- from Kea-trend 1. Sample location 3. Sr-Nd-Hf isotopic ratios U/Pb. 208Pb*/206Pb* is defined as (208Pb/204Pb-29.475)/(206Pb/204Pb-9.3066) (Galer and volcanoes. O‘Nions, 1985). 158° 156° (d) In contrast to the negative trend formed by Hawaiian shield lavas, Samoan lavas define a (2) The transect from East Molokai to West Molokai to Penguin Bank includes 22° 0.51310 0.28320 W. Molokai E. Molokai Rejuvenated stage East Molokai Rejuvenated stage i b positive trend. Icelandic lavas do not define a trend (Thirlwall et al., 2004). the oldest volcano (East Molokai) that is obviously on the Kea trend, West Koolau loka ai uin West Molokai Mo lok ng est Mo Pe W ast Kea Trend nk E Ba Penguin Bank East Molokai Molokai on the Loa trend and Penguin Bank west of West Molokai (Fig. 1c). k West Maui 0.51305 0.28315 0.51310 an W. Maui West Molokai Kea W Maui B endmember 143 144 208 206 uin Haleakala Lanai Penguin Bank 0.51305 Fig. 6 Nd/ Nd vs. Pb*/ Pb* showing the extreme (a) The late- and post-shield East Molokai lavas are exclusively Kea-like in eng Lanai P Haleakala HIG1972-D4 Mauna Kea 143 144 208 206 Kahoolawe 0.51300 Loihi Kea component at high Nd/ Nd and low Pb*/ Pb* MVD-10 9,10,11 geochemical characteristics.
0.51300 f
d Kea Shield K 20° 0.51295 S Kohala H 0.28310 P5-255 Kilauea
Mahukona D with Kea shields trending to the Loihi field and the Loa 7 P
Kea Trend N (Kilauea, Mauna Kea, 7
Kahoo 4 lawe Mauna Kea Mahukona
d Mauna Loa 1 Mauna Kea 4 0.51290 143 144 208 206
/ (b) The shield lavas of West Molokai include lavas with both Loa-and Kea-like
1 f Haleakala, West Maui) N component at low Nd/ Nd and high Pb*/ Pb* with / Hualalai 4 Loa Trend H Mauna Loa 4
Loa Trend d
1 6 0.51285 /
Kilauea c 7 d
N Loa shields trending to the Loihi field. As in previous geochemical characteristics including an extreme Loa sample which is similar to Mauna Loa 0.51295 9,10,11 1 3 N
3 0.51280
4 0.28305 4 1 1 East Molokai figures, East Molokai lavas plot within the Kea field, West 0.51275 Koolau (Makapuu), Lanai, and Kahoolawe lavas. In contrast post-shield West 200 km Loihi Fig. 1 (a) Young, < 3 Ma, Hawaiian West Molokai Penguin Bank 0.51270 Molokai ranges from the Kea to Loa fields and Penguin a 18° volcanoes showing the spatially defined 0.51290 HIG1972-D4 Lanai, Kahoolawe, Molokai lavas have Kea geochemical characteristics. 0.28300 MVD-10 Koolau-Makapuu Bank lavas are all within the Mauna Loa field with samples Loa Shield 0.51265 P5-255 Koolau endmember Loa and Kea trends. East Molokai. West (Mauna Loa, Koolau, a T307R2 0.51260 P5-255-9, -10 and -11 extending in to the Kilauea field. (3) We emphasize that the occurrence of Kea-like lavas in Loa trend volcanoes b Molokai and Penguin Bank form a trend Hualalai,Kahoolawe, Lanai) 0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99 208 206 0.51285 T307R2 Pb*/ Pb* perpendicular to the Loa & Kea trend. 0.28295 (e.g., West Molokai, Penguin Bank and Mahukona) and vice versa (e.g., Mauna 0.7030 0.7032 0.7034 0.7036 0.7038 0.7040 0.98 0.51270 0.51280 0.51290 0.51300 0.51310 Makapuu a 208 208 i
i Loa & Kea shield lavas on the BIG 87 86 o Fig. 7 Pb*/ Pb* vs. distance from Kilauea
a 208 206
a l Sr/ Sr 143 144 LOA L Kea) is a common observation (note the overlap of Pb*/ Pb* for Loa and e n a Nd/ Nd a l a
a i w n a n u
L a a ISLAND are geochemically distinct o showing a boundary between Loa and Kea lavas l u u k a i k l a o a o H
h Kea volcanoes in Fig. 7). This result requires that Loa-like and Kea-like sources u l o e i o h o M o o K h
(e.g., Kilauea & Mauna Kea compared a at 0.947, that is slightly lower than the 0.95 L K a M a M t K
* 0.96 n
s are present in the sources of all Hawaiian shields (Fig. 8). b 143 144 87 86 u e
to Mauna Loa). The objective of this a proposed by Abouchami et al. (2005). Regardless Fig. 4 (a) Nd/ Nd vs. Sr/ Sr showing data for East and West Molokai and Penguin Bank lavas. P W 6 M i KSDP 0 a k 2 research was to determine if the lavas i of where the boundary is placed the important
Notably East Molokai shield lavas are entirely within the Kea field; in contrast, West Molokai shield / o u l * a o b
hi8 k M M P forming the three Molokai shields are lavas range from Kea- to Loa-like, Penguin Bank samples are within the field of Loa shield lavas and n result is that lavas with Kea and Loa geochemical t t 8 a s s a 0 l 0.94 B e a 2 mid8 a n a also geochemically distinct. E characteristics occur in most shields. MFZ W h References l
samples P5-255-9, -10 and -11 are in the overlap fields of Loa and Kea. i o a u k KEA a K g a e
(b)(c) Submarine sampling locations on n designates location and width of Molokai e 143 144 176 177 u lo8 l (b) Nd/ Nd vs. Hf/ Hf for East and West Molokai and Penguin Bank lavas. As in (a) East e a a l Z P Z i
H 1. Weis et al., 2011 Nature Geoscience; 2. Abouchami et al., 2005 Nature; F F Penguin Bank: Melville dredge 10 K Fracture Zone. Filled circles are the average for M Molokai lavas are exclusively Kea-like, West Molokai lavas range from Kea- to Loa-like, whereas M (MVD10) (5 samples), HIG dredge (2 0.92 each shield with sub-fields shown for Mauna Kea 3. Jackson et al., 1972 Geol. Soc. Am Bull.; Penguin Bank samples are within the Mauna Loa field overlapping with the Loa-like West Molokai 0 100 200 300 400 samples) and Pisces V dive samples P5- lavas except that three Pisces V samples -9, -10 and -11 are in the field of Mauna Kea. Distance from Kilauea (km) (Eisele et al. 2003) and Koolau Kalihi- stage 4. Xu et al., 2005 G3 doi:10.1029/2004GC000830; 255 (13 samples). All samples are Only acid-leached Sr and Nd isotopic data were used to define fields for Kea and Loa volcanoes (Fekiacova et al. 2007) and Makupuu- stage. Figure is modified from Abouchami et al., 2005. 5. Xu et al., 2007 G3 doi:10.1029/2006GC0001554. tholeiitic basalt. Only high-precision MC-ICP-MS and triple spike Pb data are used.