KIMBERLITES IN FINLAND: INFORMATION ABOUT THE MANTLE OF THE KARELIAN CRATON

AND IMPLICATIONS FOR DIAMOND EXPLORATION

by

Marjaleena Lehtonen

Geological Survey of Finland

P.O. Box 96

FIN-02151 Espoo, Finland

ACADEMIC DISSERTATION

To be presented, with the permission of the Faculty of Science of the University of Helsinki, for public

criticism in the auditorium E204 of Physicum, Kumpula, on November 25th, 2005, at 14:00 hours.

Geological Survey of Finland Espoo 2005

Supervisors: Dr. Hugh O’Brien

Geological Survey of Finland

Espoo, Finland

Dr. Jukka Marmo

Geological Survey of Finland

Espoo, Finland

Reviewers: Dr. Sven Monrad Jensen

Geological Survey of Denmark and Greenland

Copenhagen, Denmark Dr. Arto Luttinen

Department of Geology

University of Helsinki, Finland

Opponent: Dr. Alan Woolley

Department of Mineralogy

Natural History Museum, London, UK

ISBN 951-690-940-X

Cover: Blue low-temperature garnet xenocrysts, “CCGE’s”, from the Ryönä kimberlite, Kuopio,

eastern Finland. The largest grains are approximately 1 mm across. Photo: Kari A. Kinnunen.

2

Lehtonen, M.L., 2005. Kimberlites in Finland: Information about the mantle of the Karelian Craton

and implications for diamond exploration. Academic dissertation, University of Helsinki, Finland.

The compositional variability, evolutionary history and diamond potential of the lithospheric mantle

underlying the southwestern margin of the Karelian Craton has been studied using major and trace

element geochemistry of peridotitic xenocrysts from five kimberlite pipes in the Kaavi-Kuopio area,

eastern Finland. Pyrope garnet xenocryst data, combined with petrological constraints provided by

peridotite xenoliths, yield a relatively complete lithologic section through the lithospheric mantle. Ni

thermometry on pyropes gives a temperature range of 650-1350 °C and, based on a geotherm

determined using mantle xenoliths, indicates a wide sampling interval, ca. 80-230 km. By studying

pyrope compositions as a function of temperature, three distinct layers can be recognized in the local

lithospheric mantle: 1. A shallow (<110 km) clinopyroxene-bearing garnet-spinel harzburgite layer

distinguished by Ca-rich but Ti-, Y-, Zr- and REE-depleted pyropes. 2. A variably depleted lherzolitic,

harzburgitic and wehrlitic horizon from 110 to 180 km. 3. A deep layer from 180 to 240 km composed

largely of fertile peridotites. The chondrite-normalized REE profiles of subcalcic harzburgitic garnet

xenocrysts originating from layer 2 preserve memory of an extensive ancient melt extraction event,

similar to observations from lithosphere underlying Archean cratons elsewhere. The REEN signatures

of rare depleted garnets from layer 3 also bear evidence of this event. The majority of pyropes are

lherzolitic and megacryst varieties exhibiting LREEN depletion relative to MREEN and HREEN typical

of Ca-saturated mantle garnets. The enrichment of MREE-HREE probably derives from a melt

metasomatic event, which is also recorded by a Ti-metasomatic overprint on many pyropes. The

subcalcic harzburgitic and a subset of the lherzolitic garnets have apparently remained unaffected by

this metasomatism. The rare depleted pyropes originating from the deepest mantle horizon possibly

represent remnants of depleted peridotites that once existed at depths greater than 180 km. The

peridotitic diamond window at Kaavi-Kuopio stretches from the top of the diamond stability field at 140

km to the base of the harzburgite-bearing mantle at about 180 km, implying a roughly 40 km thick

prospective zone.

A basic diamond exploration method in recently glaciated terrains, such as Fennoscandia, is to track

kimberlitic indicator mineral grains dispersed in Quaternary deposits. In order to develop prospecting

techniques, a target area in Lapland, 20 km by 50 km in size, was selected for a pilot study to test

semi-quantitative extraction of chromite from till. Despite the negative outcome for diamond

exploration in the target area, the main goal was realized by showing the possibility of recognizing

regionally and more locally derived chromite populations in till based on their compositional and

morphological characteristics. As a follow-up study, a detailed heavy mineral and geochemical survey

of basal till was carried out around two known kimberlites, Pipe 7 in Kaavi and Dyke 16 in Kuhmo.

The targets represent two different kimberlite types in terms of shape, size, age and petrology. The

indicator fan derived from Pipe 7 is well defined and extends at least 2 km down-ice with a maximum

concentration at 1.2 km distance from the pipe. Another kimberlitic body discovered during the study

underlines the possibility of even more undiscovered kimberlites in the area. The indicator train from

3

Dyke 16 is shorter (~1 km) and less well defined than that at Kaavi, mainly due to lower indicator

content in the kimberlite itself and subsequently in till, and a strong population of background

chromites in till; the latter were probably derived from the Kuhmo greenstone belt. The indicator

maximum occurs immediately down-ice from the kimberlite, after which the concentration drops

rapidly. Results from till fine fraction geochemistry did not reveal a kimberlitic signature in either of the

study areas; this may reflect the small size of the kimberlitic bodies.

Keywords: kimberlite, lamproite, cratons, mantle, xenocrysts, xenoliths, diamond, mineral

exploration, till, indicators, Kaavi, Kuopio, Kuhmo, Finland

Marjaleena Lehtonen

Geological Survey of Finland

FIN-02150 Espoo

Finland

4

TABLE OF CONTENTS List of original publications 7

Preface 8

1. Introduction 9

1.1. Kimberlites 9

1.2. Lamproites 11

1.3. Kimberlites and lamproites in Finland 12

1.3.1. The Kaavi-Kuopio Kimberlite Province 13

1.3.2. Kimberlites and related rocks in the Kuusamo and Kuhmo areas 14

1.4. Diamond exploration methods in glaciated terrains 15

1.5. The aim of the thesis 16

2. Review of the original publications 17

3. Discussion 19

3.1. Stratification and diamond potential of the lithospheric mantle

underlying Kaavi-Kuopio 19

3.2. Implications for diamond exploration based on glacial dispersal studies 22

4. Main conclusions 26

Acknowledgements 27

References 27

Original publications

5

6

LIST OF ORIGINAL PUBLICATIONS

This thesis consists of the following four publications. Reference to these publications in the text is

made with respect to Roman numerals, as designated below:

Paper I: Lehtonen, M.L., O'Brien, H.E., Peltonen, P., Johanson, B.S., Pakkanen, L.K., 2004. Layered mantle at the Karelian Craton margin: P-T of mantle xenocrysts and xenoliths from the Kaavi-

Kuopio kimberlites, Finland. In: Mitchell, R.H., Grütter, H.S., Heaman, L.M., Scott Smith, B.H.,

Stachel, T. (Eds.), Proceedings volume of the 8th International Kimberlite Conference, Victoria, BC,

Canada, June 22-27th, 2003. Lithos 77, 593-608.

Paper II: Lehtonen, M.L., 2005. Rare-earth element characteristics of pyrope garnets from the Kaavi-

Kuopio kimberlites – implications for mantle metasomatism. Bulletin of the Geological Society of

Finland 77, 31-47.

Paper III: Lehtonen, M.L., Marmo, J.S., 2002. Exploring for kimberlites in glaciated terrains using

chromite in quaternary till - a regional case study from northern Finland. Journal of Geochemical

Exploration 76 (3), 155-174.

Paper IV: Lehtonen, M.L., Marmo, J.S., Nissinen, A.J., Johanson, B.S., Pakkanen, L.K., 2005. Glacial dispersal studies using indicator minerals and till geochemistry around two eastern Finland

kimberlites. Journal of Geochemical Exploration 87 (1), 19-43.

M.L. Lehtonen’s contribution to the multi-authored papers was, as follows: In Paper I, she was

responsible for the microscopy of mantle xenocrysts, their LA-ICP-MS analyses, processing and

interpretation of the analytical data, and writing the manuscript for the most part. In Paper III, M.L.

Lehtonen participated in the fieldwork and sampling. Her main contribution included the laboratory

studies and microscopy of the till samples, SEM imaging of the chromite grains, processing and

interpretation of the field observations and the analytical data, and writing the manuscript for the most

part. In Paper IV, M.L. Lehtonen was involved in the fieldwork and microscopy of the till samples. She

was also responsible for the mineralogical study of the kimberlite samples, processing and

interpretation of the field observations and analytical data and, and writing the manuscript.

7

PREFACE

Kimberlitic rocks occur in the eastern and northern parts of Finland, in the regions of Kaavi-Kuopio,

Kuhmo and Kuusamo. Many of the bodies are diamondiferous. Active diamond exploration has been

ongoing in the country since the 1980’s (Tyni, 1997). Public scientific research on the Finnish

kimberlites started in the mid 1990’s and has continued to present day. The petrological

characteristics of the kimberlites have been studied (Tyni, 1997; O’Brien & Tyni, 1999) and several

methods have been applied to determine their age (Tyni, 1997; Peltonen et al., 1999; Peltonen &

Mänttäri, 2001; O’Brien et al., 2005). Mantle xenoliths (Peltonen, 1999; Peltonen et al., 1999;

Woodland & Peltonen, 1999; Peltonen et al., 2002) as well as lower crustal xenoliths (Hölttä et al.,

2000) transported by the kimberlite magma have also been a focus of scientific interest. In addition, a

geotherm for the mantle underlying Kaavi-Kuopio has been calculated using heat flow constraints and

mantle xenolith modes and their geophysical properties (Kukkonen & Peltonen, 1999; Kukkonen et

al., 2003). The morphological characteristics of diamonds have also been studied (Kinnunen, 2001).

I began my Ph.D. project in 2001 at the Geological Survey of Finland (GTK) with the financial support

of a 3-year grant from the Academy of Finland (project #50602, “Diamonds and the lithospheric

mantle in Finland”). The leader of the project was Professor Ilmari Haapala of the University of

Helsinki. The project was carried out in collaboration with the University of Cape Town. There were

two distinct aims in the project. The first was to produce further information on the compositional

variability and evolutionary history of the mantle beneath the southwestern portion of the Archean

Karelian Craton, and to evaluate its diamond potential. The study material for this part of the work

consisted mostly of mantle-derived xenocrysts from the Kaavi-Kuopio area kimberlites. The second

objective was to study the sand-sized kimberlitic indicator mineral dispersal and fine fraction

geochemical signatures in basal till surrounding the kimberlitic bodies. The ultimate aim of this part of

the project was to provide information that could be applied directly to diamond exploration in recently

glaciated terrains.

8

1. INTRODUCTION 1.1 Kimberlites Kimberlites are volatile-rich, potassic, ultrabasic igneous rocks that are enriched in many incompatible

constituents, such as alkalis, Sr, Zr, Hf, Nb and REE, but also in some first order transition elements

(Mg, Ni, Cr and Co) (e.g., Clement et al., 1984; Clement & Skinner, 1985; Mitchell, 1986). Kimberlites

are derived from limited partial melting of the mantle at very high pressure, corresponding to depths

greater than 150 km. Kimberlites exsolve volatiles, mostly CO2, during ascent to the Earth’s surface,

and usually occur as pipe-like bodies, known as diatremes, showing considerable brecciation.

Kimberlites are separated into three facies groups: an extrusive group (crater and epiclastic facies)

and a hypabyssal group (dykes, sills and root zone) that are interconnected by diatremes (Fig. 1).

Kimberlites tend to exist in clusters containing up to hundreds of individual pipes. Kimberlite clusters

typically cover only a few square kilometers of area. The size of the pipes ranges from some tens or

hundreds of meters across to the Camafuca-Camazambo pipe in Angola that is reported to cover an

area of 3.06 x 0.2 km (Erlich & Hausel, 2002). Mitchell and Bergman (1991) reported that the total

volume of known kimberlite in the world is on the order of 5000 km3 making it one of the most rare

rock types in the world.

The mineralogy of kimberlites is highly variable and complex. They have a distinctive inequigranular

texture due to their hybrid origin as they may contain mantle xenoliths, mantle xenocrysts, megacryst

suite of minerals (Mg-ilmenite, Ti-pyrope garnet, Cr-diopside, phlogopite, enstatite, zircon and olivine),

crustal xenoliths and euhedral to subhedral phenocrysts set in a groundmass matrix crystallized from

the kimberlite magma. Mitchell (1986) provided the following definition for kimberlite, emphasizing

petrologic characteristics:

“Kimberlites are inequigranular alkalic peridotites containing rounded and corroded

megacrysts of olivine, phlogopite, magnesian ilmenite and pyrope set in fine-grained

groundmass of second generation euhedral olivine and phlogopite together with primary and

secondary (after olivine) serpentine, perovskite, carbonate (calcite and/or dolomite) and

spinels. The spinels range in composition from titaniferous magnesian chromite to magnesian

ulvöspinel-magnetite. Accesory minerals include diopside, monticellite, rutile and nickeliferous

sulphides. Some kimberlites contain major modal amounts of monticellite.”

Since their recognition in the late 19th century in South Africa, kimberlites have been regarded as the

principal host rocks for diamonds. Diamonds are not, however, genetically related to kimberlites.

Instead, diamonds occur in kimberlites as xenocrysts that have been transported from the upper

mantle by the kimberlite magma during ascent to the Earth’s surface. Diamonds exist within the upper

mantle in unique host rocks that include eclogite and some peridotite varieties, mostly garnet- and

chromite-harzburgite. Equilibrium subcontinental geotherms require depths of 150 to 200 km

(corresponding 45 to 55 kbar and 1050 to 1200 oC) to stabilize diamond within such cratonic

environments (e.g., Kennedy & Kennedy, 1976; Pollack & Chapman, 1977).

9

Figure 1. Schematic diagram (not to scale) of an idealized kimberlite magmatic system illustrating relationships between crater, diatreme and hypabyssal facies rocks. Hypabyssal rocks include sills, dykes, root zone, and “blow”. Depth of a typical kimberlite pipe is on the order of 2-3 km (after Mitchell, 1986).

Diamondiferous kimberlites are largely confined to the stable Archean cratons characterized by a low

heat flow and a thick lithospheric mantle. A low geothermal gradient combined with great thickness of

lithosphere increase the probability that the ascending kimberlite magma will intersect diamondiferous

material in the mantle (Kennedy, 1964; Helmstaedt & Gurney, 1995; Morgan, 1995). The necessity for

the Archean bedrock is probably related to the unique, high degree of partial melting of the mantle

(komatiite extraction), leaving extremely depleted harzburgite and dunite residues (Helmstaedt &

Gurney, 1995).

Already in 1914 kimberlites were divided into two classes, basaltic and lamprophyric (Wagner, 1914).

This terminology described the macroscopic appearance of the rocks without genetic significance.

These terms are today considered archaic and have been replaced by the terms Group I and Group II

kimberlite. Groups I and II can be distinguished by their isotopic compositions (87Sr/86Sr vs. 143Nd/144Nd) (Smith, 1983). Group I kimberlites are derived from a relatively primitive mantle source,

probably asthenosphere, and Group II kimberlites are thought to originate from metasomatically

enriched parts of the lithosphere (Mitchell, 1986). Group I kimberlites are found worldwide and range

in age from Lower Proterozoic to Miocene. They can be described as olivine-rich monticellite-

serpentine-calcite kimberlites containing upper mantle xenoliths and megacrystal minerals. Group II

kimberlites consist of olivine macrocrysts in a matrix of phlogopite and diopside with spinel, perovskite

and calcite. In contrast to Group I kimberlites, they lack magnesian ülvospinel and monticellite. The

type locality of Group II kimberlites is in South Africa, and they are also referred to as orangeites

(Mitchell, 1995).

10

Figure 2. Schematic diagram (not to scale) illustrating the idealized morphological relationships oflava, crater or pyroclastic, hypabyssal and plutonic facies lamproite in different types of ventcomplexes. Also illustrated are the appearances of lamproite vents after erosion of the maar crater.Depth of a typical lamproite vent is on the order of 0.5-1.0 km (after Mitchell & Bergman, 1991).

1.2 Lamproites Besides kimberlite, lamproite, or olivine lamproite more precisely, is another important hard rock

source for diamond. Lamproite is an even more exotic rock type than kimberlite. The total volume of

known lamproite in the world is less than 100 km3 (Mitchell & Bergman, 1991). Lamproitic rocks

resemble kimberlites in being mafic, peralkaline, ultrapotassic, volatile-rich igneous rocks but there

are also significant differences between these two rock types. Lamproite clan rocks have no

mineralogical or geochemical affinities with Group I kimberlites (Mitchell, 1986): they lack, e.g. the

megacryst suite minerals, monticellite and primary serpentine. However, lamproites have some

mineralogical affinities with Group II kimberlites (Mitchell & Bergman, 1991). For instance, the

abundant phlogopite that occurs in both of these rock types is of broadly similar composition. Group II

kimberlites differ from lamproites in that they contain abundant calcite. Lamproites have varying

mineralogical composition that may include diopside, phlogopite, K-Ti-richterite, leucite, sanidine,

wadeite, priderite and/or olivine with minor apatite, perovskite, ilmenite and spinel. Sanidine, K-Ti-

richterite and leucite are absent from Group I kimberlites and extremely rare in Group II (Mitchell &

Bergman, 1991). Geochemically lamproites are similar to Group II kimberlites being enriched in K, Zr,

Nb, Sr, Ba and Rb relative to Group I kimberlite (Kirkley et al., 1991). Lamproites are probably derived

11

from a mantle source similarly enriched as the source for Group II kimberlites (Mitchell & Bergman,

1991).

Like kimberlites, lamproites typically occur within groups or fields. Individual lamproite bodies consist

of dykes, sills, vents, flows and scoria cones with surface areas of some hectares (Fig. 2). Diatreme-

type formations characteristic of kimberlites are not common among lamproites. The difference in

morphology reflects the difference in volcanology of lamproites and kimberlites that in turn results

from the different volatile contents and composition. Lamproites are characterized by air-fall tuffs,

base surge-deposits, lava flows, and a central core of massive lamproite (Mitchell & Bergman, 1991).

The volatiles in lamproites are highly soluble silicate melts rich in H2O and F. During the ascent of this

type of magma, it is not until at 0.5-1 km depth that significant exsolution of an aqueous fluid phase

occurs. Explosive expansion of this fluid probably results in the formation of funnel-shaped lamproitic

vents (Mitchell & Bergman, 1991). In contrast, kimberlitic magmas enriched in CO2 exsolve volatiles at

much greater depths, 2-3 km. As a result, kimberlitic magmas will produce deeper eruptions that form

carrot-shaped diatremes (Mitchell, 1986).

1.3 Kimberlites and lamproites in Finland The Archean Karelian Craton is prospective for diamondiferous kimberlitic and lamproitic rocks based

on the empirical evidence necessary for diamond preservation: the geothermal gradient is low

(Kukkonen & Jõeleht, 1996) and the lithosphere is thicker than 170 km according to seismic

observations (Calcagnile, 1982). The area where thick lithospheric mantle, low heat flow and Archean

bedrock overlap is the most prospective for diamond exploration. In the eastern Fennoscandian

Shield, such area covers a major portion of Finland, Russian Karelia and the Kola Peninsula.

So far, kimberlitic rocks have been found in the eastern and northern parts of Finland, in the regions

of Kaavi-Kuopio, Kuhmo and Kuusamo (Tyni, 1997; O’Brien & Tyni 1999; Fig. 3). The first kimberlite

in Finland was discovered in Kaavi in 1964 by Malmikaivos Oy, a small Finnish mining company,

while prospecting for base metals (Tyni, 1997). The discovery of this non-diamondiferous kimberlite

was nearly forgotten for 20 years, until in the early 1980’s the company found two more kimberlite

pipes in the area. In 1986 Malmikaivos signed a joint venture with Australian Ashton Mining Ltd. with

the aim to prospect for economic kimberlites in Finland. By 1996 Malmikaivos/Ashton had discovered

24 kimberlitic bodies in the Kaavi-Kuopio, Kuusamo and Kuhmo areas, most of them diamondiferous,

varying in size from less than 1 ha to nearly 4 ha. Since then several other companies have been

involved in diamond exploration in Finland. Diamet, DeBeers and Rio Tinto were the other major

diamond companies operating in Finland before 2000, but none of them were as successful as

Malmikaivos/Ashton. European Diamonds Plc. started operating in Finland in the late 1990’s and has

explored a large part of the prospective Karelian Craton in the eastern and central parts of the

country. The company has identified a number of kimberlite indicator mineral dispersal fans in

Quaternary till and reported a large kimberlite complex in Lentiira in the Kuhmo municipality.

Presently, European Diamonds Plc. is re-evaluating the Lahtojoki kimberlite (also known as Pipe 7) in

12

Kaavi that was first discovered by Malmikaivos/Ashton in 1989. Some other Malmikaivos/Ashton

kimberlites in the Kaavi-Kuopio region are being re-examined by Nordic Diamonds Plc., which is also

exploring for new kimberlites in the area. In addition, Gondwana Investments SA has claimed ground

for diamond exploration in Kaavi. Karelian Diamonds Resources Plc. (an associate of Conroy

Diamonds and Gold Plc.) has been operating mainly in the Kuhmo area for several years and has

recovered kimberlite and diamond indicators in till samples. Farther north in Finland, exiting new

discoveries have recently (2004-2005) been reported by Tertiary Minerals Plc. and it’s associate

Sunrise Diamonds Plc. – the companies have drilled into five kimberlites south of Kuusamo. Despite

of all these activities, the Karelian Craton remains even today (2005) under-explored given its size

and potential.

.3.1 The Kaavi-Kuopio Kimberlite Province

ated at the southwestern margin of the Karelian Craton.

N

Figure 3: ing the knownkimberlite lo

Simplified geological map of Finland (after Korsman et al., 1997), showcations in the eastern and northern parts of the country.

1

The Kaavi-Kuopio Kimberlite Province is situ

Presently approximately twenty classical Group I kimberlites are known from the district. The

kimberlites contain abundant macrocrysts of olivine, Mg-ilmenite, Cr-diopside, pyrope garnet;

13

phenocrysts of olivine and microphenocrysts of monticellite, perovskite, kinoshitalite mica and spinel

occur in a calcite + serpentine matrix (Tyni, 1997; O’Brien & Tyni, 1999). Phlogopite is not abundant,

and there are no indications of Group II kimberlite affinities in these rocks. The pipes have intruded

Archean (3.5-2.6 Ga) basement gneisses and allochthonous Proterozoic (1.9-1.8 Ga) metasediments

(Kontinen et al., 1992). The intrusions range from purely hypabyssal kimberlite dykes to multiphase

pipes of diatreme facies rocks. The Kaavi-Kuopio kimberlites generally contain abundant diamond

indicator minerals and almost all of them contain at least trace amounts of microdiamonds (Tyni,

1997).

Over the years, several methods have been applied to age dating the Kaavi-Kuopio kimberlite

.3.2 Kimberlites and related rocks in the Kuusamo and Kuhmo areas

are carbonate-rich ultramafic

magmatism. The first age determinations were K-Ar measurements of two whole-rock samples from

Pipes 1 and 2, which yielded ages of 430 Ma and 560 Ma, respectively (Tyni, 1997). These ages were

regarded as rough estimates of the true age. Another approach to date the kimberlite emplacement

was attempted by Peltonen et al. (1999): garnet and clinopyroxene separates from four garnet-

peridotite xenoliths from Pipe 7 yielded two-mineral Sm-Nd isochron ages between 525±10 Ma and

607±20 Ma. The younger age is considered to be closer to the real age of the kimberlite

emplacement, because garnet and clinopyroxene have most likely coexisted at (or close to) chemical

equilibrium at the time of detachment of the xenoliths from their lithospheric mantle source. Yet

another method, U-Pb age determinations on zircon xenocrysts from Pipe 7, resulted in two

concordant age groups, 2.7 Ga and 1.8 Ga (Peltonen & Mänttäri, 2001). Clearly, these zircons have

not crystallized from the kimberlite magma, but were derived from older sources. The authors believe

that the two different ages result from partial resetting of Archean grains, possibly originating from

lower crustal mafic pegmatitites and hydrous coarse-grained veins, in the 1.8 Ga thermal event

related to the Svecofennian orogeny. Probably the most reliable ages determined for the kimberlite

magmatism are the ion microprobe U-Pb ages of 589-626 Ma from perovskite grains recovered from

several Kaavi-Kuopio kimberlites (O’Brien et al., 2005). Perovskite is an ideal mineral for dating

kimberlites, as it crystallizes from the kimberlitic melt.

1

Farther into the Karelian Craton, in the Kuusamo area, there

lamprophyre (aillikite) dykes, and in Kuhmo, there is a rock type, which appears to be intermediate

between olivine lamproite and Group II kimberlite (O’Brien & Tyni, 1999). The latter, referred as

Seitaperä kimberlite or Dyke 16, shows many similarities to the diamondiferous rocks of the

Lomonosova deposit in the Archangelsk area. It contains most of the typomorphic minerals of

lamproite and in major element composition it is equivalent to average olivine lamproite.

Serpentinized olivine macrocrysts and phenocrysts along with microphenocrysts of phlogopite, K-

richterite, diopside, apatite and perovskite are found in a serpentine and calcite matrix (O’Brien &

Tyni, 1999). However, Dyke 16 also has certain mineralogical characteristics that are more typical of

those seen in Group II kimberlites, such as Ca-Zr-silicates, specific zoning in micas, and more

importantly primary carbonate in the matrix. The diamond content of Dyke 16 is marginal based on

14

the claim report delivered by Malmikaivos-Ashton to the Ministry of Trade and Industry. Based on U-

Pb ages from perovskites, ca. 1250 Ma (O'Brien et al., 2005), and Ar-Ar ages from phlogopite

microphenocrysts, ca. 1200 Ma (Hugh O’Brien, personal communication, 2004), the Kuhmo rocks

appear to be considerably older than the kimberlites in the Kaavi-Kuopio region as well as those in the

Archangelsk area – the latter hosts 365 Ma kimberlites (Beard et al., 1998).

1.4 Diamond exploration methods in glaciated terrains

s and lamproites, in recently glaciated

hen recovered from heavy mineral concentrates of sediment samples, the kimberlitic indicator

Another ular tool in regional diamond exploration is the use of till geochemistry. The method is

A common method to explore for diamond host rocks, kimberlite

terrains, such as the Canadian and Fennoscandian Shields, is to track kimberlitic/lamproitic indicator

mineral grains dispersed in Quaternary till and stream sediments. This prospecting method has been

applied successfully worldwide as a diamond exploration technique, especially during reconnaissance

and regional stages of exploration programs (e.g., Gurney, 1984; Atkinson, 1989). The indicator

minerals meet the following requirements (McClenaghan & Kjarsgaard, 2001): they are (1) far more

abundant than diamonds in the source rock; (2) visually and chemically distinct; (3) sand-sized (0.25-

2.0 mm); (4) sufficiently dense to be concentrated by gravity; and (5) chemically and physically

resistant to survive preglacial weathering and subsequent glacial transport. The classic kimberlitic

indicator mineral suite used in diamond exploration includes Cr-pyrope, Ti-pyrope and eclogitic

garnet, Cr-diopside, Mg-ilmenite, and chromite. Indicator mineral content and relative abundance of

different indicators may vary greatly between individual kimberlites and lamproites (e.g., Mitchell

1986; Mitchell & Bergman, 1991).

W

minerals are traced up-ice or up-stream to the source area. Ideally, the indicator trail will lead to the

kimberlitic host rock. At the extreme, indicator grains may form dispersal trains extending several tens

of kilometers in till from individual kimberlites and kimberlite fields, as in the Lac de Gras region,

Canada (e.g., Armstrong, 1999; McClenaghan et al., 2002). Exploration based on indicator minerals

requires profound understanding of the glacial history in the region as well as sensitive processing

methods, as the indicator contents in glacial sediments can be extremely low, even relatively close to

the kimberlite. Recovery of a few sand-sized indicator grains from tens of kilograms samples requires

a methodology necessitating 1 ppb sensitivity or better. GTK has developed such a system by

modifying and augmenting 3" and 4.5" Knelson Concentrators that accomplishes nearly complete

recovery of moderately heavy minerals (>0.25 mm) from till (Chernet et al., 1999).

pop

based on the distinct chemistry of kimberlitic rocks, which are enriched in incompatible elements (Sr,

Ba, LREE, Nb, Ta, Hf, Zr, P, Ti) as well as some compatible first order transition elements (Mg, Ni, Cr,

Co) (e.g., Fipke et al., 1995). Geochemical anomalies formed by these elements can be used as

pathfinders in diamond exploration in a similar way to indicator minerals. The set of pathfinder

elements are typically region-specific because geochemical anomalies are created by the contrast

between kimberlite and country rocks (McClenaghan & Kjarsgaard, 2001). Geochemical signatures in

15

till derived from individual kimberlites can be detected usually over a distance of a few hundreds of

meters to some kilometers only.

For successful application of the two diamond exploration methods, indicator minerals and

1.5 The aim of the thesis ond, kimberlites and lamproites are also valuable sources of information

nother major objective of this thesis was to study glacial dispersal of indicator minerals and

geochemistry of till, it is essential to carry out case studies around known kimberlitic bodies. Active

diamond exploration in Finland has been ongoing for over two decades (Tyni, 1997) but systematic

glacial dispersal studies around the known kimberlites, similar to those conducted on several

kimberlites and kimberlite fields in the shield terrain of Canada (listed in McClenaghan & Kjarsgaard,

2001), have not been published to date.

Not only host rocks of diam

about the underlying lithosphere in the form of mantle xenoliths and xenocrysts that the ascending

kimberlitic/lamproitic magma has brought to the Earth’s surface. The xenolith suite in the Kaavi-

Kuopio kimberlites provides evidence of a compositionally stratified lithospheric mantle adjacent to the

ancient suture zone between the Archean Karelian Craton and the Proterozoic Svecofennian mobile

belt (Peltonen et al., 1999). A shallow layer (<900 °C) of garnet-spinel harzburgites is underlain by a

layer of garnet facies peridotites (180-240 km). Mantle eclogite xenoliths are also present, some of

them being highly diamondiferous (Peltonen et al., 2002). The relatively small number of mantle

xenoliths available from the Kaavi-Kuopio kimberlites, however, gives a limited sampling of the

lithosphere relative to the mantle xenocrysts. One of the objectives of this doctoral thesis was,

therefore, to obtain additional information on the compositional variability and evolutionary history of

the Kaavi-Kuopio mantle section by studying major and trace element geochemistry of peridotitic

xenocrysts and by applying thermobarometric methods (Papers I and II). An important aspect was

also to estimate the thickness of the peridotitic diamond-prospective horizon in the lithosphere.

A

geochemical anomalies in Quaternary till surrounding kimberlitic bodies. A case study in Lapland was

carried out in order to test and develop methods for this kind of work (Paper III). Two eastern Finland

kimberlites were selected as targets of detailed heavy mineral surveys (Paper IV): the Lahtojoki pipe

(Pipe 7) in Kaavi and the Seitaperä dyke swarm (Dyke 16) in Kuhmo. The aim was to contribute to the

overall understanding of the Quaternary history in these two regions and, more importantly, to provide

key information for diamond exploration in Fennoscandia and elsewhere in glaciated terrains.

16

2. REVIEW OF ORIGINAL PUBLICATIONS Paper I Paper I focuses on the subcontinental lithospheric mantle (SCLM) underlying the 600 Ma Kaavi-

Kuopio Kimberlite Province situated at the Karelian Craton margin. Peridotitic clinopyroxene and

pyrope garnet xenocrysts from four kimberlite pipes have been studied using major and trace element

geochemistry and thermobarometric methods. The objective of the work has been to study the vertical

compositional variability and peridotitic diamond potential of the lithosphere. The xenocryst data,

when combined with the petrological constraints and geotherm determined from peridotite xenoliths,

yield a relatively complete section through the 600 Ma SCLM, which can be divided into three layers.

1. A low temperature (700-850 °C) harzburgite layer, corresponding to 80-110 km in depth. This

horizon is distinguished by Ca-rich but Ti-, Y- and Zr-depleted “CCGE” pyropes. The few xenoliths

originating from this layer are garnet-spinel harzburgites containing secondary clinopyroxene. 2. A

variably depleted lherzolitic, harzburgitic and wehrlitic horizon from 950-1150 °C, or 130 to 180 km.

This layer contains depleted subcalcic “G10” garnets, characteristic of diamond-bearing harzburgites.

3. A deep, mainly refertilized lherzolitic, horizon from 180 to 240 km, where the bulk of the peridotitic

(and eclogitic) xenoliths have been derived. The deepest layer contains a very minor amount of

depleted material and possibly represents a melt-enriched version or extension of layer 2. This

stratigraphy implies that the peridotitic diamond prospective zone is between 140 km and 180 km, the

lower temperature stability limit of diamond and the deepest level where G10 pyropes are found.

Paper II The aim of Paper II is to obtain information on the origin and evolutionary history of the three

lithospheric mantle layers described in Paper I by studying rare earth element (REE) geochemistry of

peridotitic pyrope garnet xenocrysts from Kaavi-Kuopio kimberlite pipes. The C1-chondrite-normalized

REE contents of subcalcic harzburgitic garnet xenocrysts originating from layer 2 bear evidence of an

extensive ancient melt extraction event, similar to observations from lithosphere underlying Archean

cratons elsewhere. Memory of this event has also been preserved in the REEN signatures of the rare

depleted garnets from layer 3 and possibly in the low-temperature CCGE pyropes despite their

contradictory saturation in Ca. The highly depleted mantle has been subsequently affected by fluid

metasomatism involving introduction of a fractionated fluid/melt low in HREE giving rise to the

sinusoidal REEN patterns, characteristic for subcalcic harzburgitic garnets. Most of the lherzolitic and

wehrlitic garnet xenocrysts exhibit “N-type” REEN patterns with LREEN depletion relative to MREEN

and HREEN typical of Ca-saturated mantle garnets. The enrichment of MREE and HREE probably

derives from another metasomatic event caused by silicate melts, leaving the harzburgitic and some

rare depleted lherzolitic varieties unaffected in zones or bands. The metasomatism has also caused

the crystallization of abundant megacrystal Ti-pyropes with N-type REEN patterns. Lower LREE

concentrations in lherzolitic garnets compared to harzburgitic varieties may derive from the

redistribution of REE during re-equilibration with clinopyroxene. The history of the shallowest mantle

layer (<110 km) consisting of garnet-spinel-harzburgite and characterized by the CCGE garnets is

obviously multi-stage, including several depletion and enrichment events.

17

Paper III Paper III describes a case study from eastern Lapland applying and testing methodology used in

diamond exploration in recently glaciated terrains. The emphasis of this work is on chromite extracted

from till by studying its composition and morphology with the aim to discriminate local and regional

populations, and to identify possible kimberlitic/lamproitic indicators. The study area, 20 x 50 km in

size, was selected on the basis of the regional occurrence of a variety of mantle-derived rocks, the

recovery of a kimberlitic indicator garnet from till, and the well-established Quaternary stratigraphy in

the region. The sampling area is proven to contain at least two compositional (and morphological)

populations of till chromite. The first population is apparently regional and derived from voluminous

layered mafic intrusions situated up-ice. The second population, much fewer in number, is present in

one third of the samples and concentrated in a couple of clusters within the target area. It is

characterized by low contents of Ti, high Cr and Mg, similar in composition to chromite inclusions in

diamond. The source for these grains remains uncertain, but is probably not kimberlitic, as no high-Ti,

high-Cr chromites diagnostic for kimberlites and lamproites are present in the samples. Although the

outcome of the study is negative in terms of diamond exploration, the main goal of the work was

realized by testing the applicability of the system for heavy mineral separation from glacial deposits.

Paper IV Paper IV describes glacial dispersal studies of two eastern Finland kimberlites, Pipe 7 in Kaavi and

Dyke 16 in Kuhmo. The objective is to document and interpret indicator mineral (0.25-2.0 mm) trains

and geochemical signatures in the surrounding basal till, and thus, to provide information that can be

applied to diamond exploration. The processing methodology applied was tested and partly

developed during the Eastern Lapland case study (Paper III). The selected kimberlites are

characterized by different shape, size, age and petrology, as well as varying bedrock lithology and

Quaternary deposits. The 2-ha Pipe 7 is a typical Group I kimberlite that belongs to the 600 Ma Kaavi-

Kuopio Kimberlite Province (Papers I and II) and contains abundant indicator minerals, Mg-ilmenite,

pyrope and eclogitic garnet, and Cr-diopside. The 1.2 Ga Dyke 16 in Kuhmo, 200 km northeast of

Kaavi and further on-craton, has characteristics of both olivine lamproite and Group II kimberlite.

Virtually the only indicator mineral in the overall indicator-poor dyke swarm is chromite. The swarm

intrudes a 300 x 600 m area. The indicator grains down-ice from Pipe 7 form a symmetrical and well

defined fan in the basal till that can be followed at least 2 km. The maximum concentration of

indicators is found at 1.2 km distance from the pipe. Another kimberlitic body discovered during the

study 300 m down-ice from Pipe 7 emphasizes that there might be more undiscovered kimberlitic

sources in the area. In the Kuhmo sampling area the indicator dispersal trail from Dyke 16 is

considerably shorter (<1 km) and not as well defined as in Kaavi, mainly due to lower indicator counts

in till and a strong background population of till chromites. Results of till geochemistry reveal only

weak kimberlitic signatures in both areas, probably reflecting the small size of the bodies.

18

3. DISCUSSION 3.1 Stratification and diamond potential of the lithospheric mantle underlying Kaavi-Kuopio It has been previously shown that the structure of the SCLM is stratified adjacent to the ancient suture

zone between the Archean Karelian Craton and the Proterozoic Svecofennian mobile belt. The mantle

xenolith suite demonstrates that a shallow zone (<900 °C) of garnet-spinel harzburgite is underlain by

a zone of garnet facies peridotites (180-240 km) including lherzolitic, harzburgitic, olivine websteritic,

and wehrlitic varieties (Peltonen et al., 1999). Papers I and II provide more information on the

compositional variability and evolutionary history of the Kaavi-Kuopio mantle section based on

peridotitic mantle xenocrysts. Due to the absence of coexisting mineral assemblages, xenocrysts

cannot provide as much information as xenoliths, but they are far more abundant in kimberlites, and,

at least in theory, provide a more statistically reliable sample of the underlying mantle (Schulze,

1995). Ni thermometry (Griffin et al., 1989a; Ryan et al., 1996) on pyrope garnet xenocrysts gives a

temperature range of 650-1350 °C, indicating a sampling interval of 80-230 km projecting the

temperatures into the local geotherm determined using mantle xenoliths (Kukkonen & Peltonen, 1999;

Kukkonen et al., 2003). Three distinct layers in the lithospheric mantle (Fig. 4) can be identified based

on garnet xenocryst compositions and temperatures combined with the petrological framework

provided by the mantle xenoliths: (1) A shallow, <110 km, garnet-spinel peridotite layer distinguished

by CCGE pyropes, named by Kopylova et al. (2000) after ‘‘chromite–clinopyroxene–garnet

equilibrium’’. (2) A variably depleted lherzolitic, harzburgitic and wehrlitic horizon from 110 to 180 km

containing diamond-indicative subcalcic G10 garnets of Dawson & Stephens (1975). (3) A deep layer,

>180 km, composed largely of fertile peridotites. There are, however, some uncertainties involved in

the proposed stratigraphy. Ni thermometry gives a temperature of equilibration of a single pyrope

garnet to a typical accuracy of ±50 °C (Ryan et al., 1996), corresponding to ca. ±10 km in depth, for

the range of 650-1250 °C when extrapolated to the Kaavi-Kuopio geotherm. Bearing in mind these

factors of uncertainty, the stratigraphy implies a peridotitic diamond window between 140 km and 180

km, the lower temperature stability limit of diamond and the deepest level where G10 pyropes are

found. Significantly, this roughly 40 km diamond prospective zone is very close to the same thickness

as that sampled by the Cretaceous diamondiferous kimberlites from South Africa (e.g., Griffin & Ryan,

1995). It is still worth emphasizing that subcalcic G10 garnets form only a minor portion of the

peridotitic garnet xenocrysts derived from the 140-180 km mantle section beneath Kaavi-Kuopio.

Instead, lherzolitic pyrope (G9), which is less commonly associated with diamonds, is dominant

(Papers I and II). Nevertheless, a well-sampled component of highly diamondiferous eclogite xenoliths

(Peltonen et al., 2002) as well as a considerable amount of diamond-indicative Group I eclogitic

garnet xenocrysts in the Kaavi-Kuopio kimberlites (Fig. 5; McCandless & Gurney, 1989; Gurney &

Zweistra, 1995) add significantly to the diamond potential of this mantle section. Overall, as regards

diamond exploration in the area, this is a promising outcome.

The depleted contents of Ti, Y, Zr and HREE of the G10 garnet xenocrysts originating from layer 2

bear evidence of an ancient komatiite extraction event that started the formation of cratonic peridotites

(e.g., Shimizu & Richardson, 1987; Nixon et al., 1987; Stachel et al., 1998). The highly depleted

19

lithosphere was subsequently affected by fluid metasomatism with invasion of a fractionated fluid/melt

low in HREE giving rise to the sinusoidal C1-chondrite normalized REE patterns, characteristic of G10

garnets (e.g., Shimizu, 1975). Lherzolitic garnets originating from layers 2 and 3 are mostly Ti-rich

and form REEN patterns typical for Ca-saturated mantle garnets (e.g., Shimizu, 1975) documenting an

enrichment of MREE and HREE. This metasomatic re-enrichment event was probably caused by

silicate melts (Griffin et al., 1989b; Stachel et al., 1998) close in composition to a megacryst-forming

magma (Burgess & Harte, 2004). The melt metasomatism added a Ti-overprint to most of the

lherzolitic pyropes and caused the crystallization of abundant megacrystal garnets. Harzburgitic and

some Ti-poor lherzolitic varieties remained unaffected possibly as remnant zones, pockets or bands.

The results of Papers I and II demonstrate that some depleted material exists in layer 3 at depths

greater than 180 km. This material is interpreted to represent remnants of layer 2 that once extended

to these depths but was almost entirely destroyed by the voluminous melt metasomatic event.

The origin of layer 1 is enigmatic (Papers I and II). It is distinguished by CCGE garnets that are ultra-

depleted with respect to Ti, Y, Zr and REE but fertile in their main element composition, as measured

by saturation in Ca. Kopylova et al. (2000) suggested that the Ca-Cr trend in the CCGE’s derives from

re-equilibration with clinopyroxene and chromite following invasion of Ca-rich fluid into the system

leading to the crystallization of secondary clinopyroxene and the transformation of harzburgitic

garnets to lherzolitic. However, the whole-rock analyses of the garnet-spinel facies peridotite xenoliths

(Peltonen et al., 1999) show that they actually contain less CaO than the other xenolith varieties.

Peltonen et al. (1999) suggested that these xenoliths could represent remnants of the reworked

Archean lithosphere, metasomatized by a kimberlite-derived melt or fluid. For the Slave Craton

examples Carbno & Canil (2002) concluded that this kind of enrichment event could have been

caused by interaction of carbonate melt with harzburgite, leading clinopyroxene to form from

orthopyroxene (Yaxley et al., 1998). The problem with these hypotheses arises from the ultra-

depleted contents of trace elements in the CCGE garnets. It is unlikely that these signatures would

have been preserved in this kind of metasomatic events. However, it is possible that layer 1 has been

somehow affected by the Svecofennian orogeny, as this layer does not exist further inside the

Karelian Craton based on mantle xenocryst studies from the Kuhmo area kimberlites (O’Brien et al.,

2003).

The stabilization ages of the Kaavi-Kuopio mantle layers (1-3) are not resolved in this study. The

middle (2) and the upper (1) layers are considered to be Archean (Paper II), with a similar age to that

of the overlying crust (Pearson, 1999). Peltonen et al. (1999) suggested that the lowermost section of

the mantle (layer 3) has been tectonically emplaced during the collision event at 1.88 Ga, but the

results of Papers I and II indicate that it, in fact, represents metasomatically re-enriched Archean

mantle. New, previously unavailable, data on Re-Os isotopes in mantle xenoliths (Peltonen &

Brügmann, 2005) confirm that layer 2 is indeed Archean with a ~3.3 Ga age determined for melt

extraction. This age corresponds with the oldest formation ages of the overlying crust (Mutanen &

Huhma, 2003), implying that layer 2 represents unmodified SCLM stabilized during the Paleoarchean.

20

Svec

ofen

nian

mob

ile b

elt

Kar

elia

n cr

aton

1.9

Ga

1.9

Ga

3.5

- 2.6

Ga

Asth

enos

pher

e

kim

berli

tes

eclo

gite

s

DI

Gra

ph.

Hig

h V

p m

afic

zon

e

250

200

150

10050km

Cru

st

Shal

low

SC

LM(fi

ne-g

rain

ed,

met

as. h

rz)

Mid

dle

SCLM

(coa

rse,

var

iabl

y de

plet

ed lh

z / h

rz)

Dee

p SC

LM(c

oars

e, re

fert

ilize

d lh

z)

kim

berli

tes

carb

onat

ites

3.5

- 2.6

Ga

3.5

Ga

- pre

sent

?

1.9

Ga

-pr

esen

t?

50 75 100

125

150

175

200

225

2500.

000.

250.

500.

751.

001.

25

TiO

2 wt%

km

CC

GE

GT

HR

Z G

TLH

Z G

TW

HR

GT

MEG

AC

R. G

T

n =

497

GA

RN

ET X

ENO

CRY

STS

0.010.1110100

LaC

eN

dS

mEu

Gd

Dy

ErYb

Lu

Garnet/C1-chondrite

n =

4

GA

RN

ET

XEN

OC

RYS

TS

0.010.1110100

LaC

eN

dS

mEu

Gd

Dy

ErYb

Lu

Garnet/C1-chondrite

n =

6

GA

RN

ET

XEN

OC

RY

STS

0.010.1110100

LaC

eN

dS

mE

uG

dD

yE

rY

b

Garnet/C1-chondrite

GA

RN

ET X

ENO

CRY

STS

Figu

re 4

. On

the

left,

a s

chem

atic

cro

ss-s

ectio

n th

roug

h th

e rif

ted

edge

of t

he K

arel

ian

crat

on s

how

ing

the

thre

e di

stin

ct p

etro

logi

c la

yers

in th

e m

antle

def

ined

fro

m x

enoc

ryst

and

xen

olith

com

posi

tions

. Sta

rs m

ark

the

dept

h of

orig

in o

f the

man

tle x

enol

iths

base

d on

BK

N th

erm

obar

omet

ry (P

elto

nen

et a

l., 1

999)

. In

the

mid

dle,

TiO

2 of p

yrop

e xe

nocr

ysts

vs.

dep

th c

alcu

late

d by

ext

rapo

latin

g T N

i (R

yan

et a

l., 1

996)

to th

e lo

cal g

eoth

erm

(Kuk

kone

n et

al.,

200

3). O

n th

e rig

ht, t

he

mai

n ty

pes

of C

1-ch

ondr

ite n

orm

aliz

ed (M

cDon

ough

& S

un, 1

995)

RE

E p

rofil

es o

f pyr

opes

orig

inat

ing

from

the

dist

inct

man

tle la

yers

. (A

bbre

viat

ions

: SC

LM =

su

b-co

ntin

enta

l lith

osph

eric

man

tle; C

CG

E G

T =

Ca-

rich

but T

i-, Y

-, Zr

- and

RE

E-d

eple

ted

garn

et; H

RZ

GT

= ha

rzbu

rgiti

c ga

rnet

; LH

Z G

T =

lher

zolit

ic g

arne

t; W

HR

GT

= w

ehrli

tic g

arne

t; M

EG

AC

R. G

T =

meg

acry

stal

gar

net)

Lu

n =

13

0.010.1110100

LaC

eN

dSm

Eu

Gd

Dy

ErYb

Garnet/C1-chondrite

GA

RN

ET

XEN

OC

RYS

TS

Lu

n =

20.

010.1110100

LaC

eN

dSm

EuG

dD

yE

rY

bLu

Garnet/C1-chondrite

n =

13

GA

RN

ET X

EN

OC

RYS

TS

CC

GE

LHZ,

ME

GA

CR

YSTS

HR

Z, L

OW

-Ti L

HZ

LHZ

LOW

-Ti L

HZ

Shal

low

SC

LM

Mid

dle

SCLM

Dee

p SC

LM

21

Despite the fact that xenoliths originating from layer 3 yield only Proterozoic Re-Os ages, Peltonen &

Brügmann (2005) interpreted it to represent Archean SCLM overprinted by the melt metasomatism as

described in Paper II, with the original isotopic signature being disturbed. The authors relate the origin

of layer 3 to continental break-up at ~2 Ga. A maximum age determination for layer 1 peridotites is

~2.6 Ga, suggesting that Archean domains form at least part of the shallowest mantle horizon. Layer

1, however, is interpreted by Peltonen & Brügmann (2005) to represent mostly Proterozoic arc

complex lithosphere thrust beneath the Karelian Craton margin crust during the ~1.9 Ga continental

collision.

0.00

0.20

0.40

0.60

0.80

1.00

1.20

0.00 0.05 0.10 0.15 0.20

Na2O wt%

TiO

2 wt%

GARNET XENOCRYST

GROUP I ECLOGITE

GROUP II ECLOGITE

ECLOGITEMEGACRYST

n = 393

Figure 5: TiO2 vs. Na2O classification by McCandless & Gurney (1989) for eclogitic and megacrystalKaavi-Kuopio garnet xenocrysts. Note the abundance of diamond-indicative Group I eclogitic grains.

3.2 Implications for diamond exploration based on glacial dispersal studies The results of indicator mineral dispersal (Paper IV) from the Lahtojoki (Pipe 7) and Seitaperä

kimberlites (Dyke 16) demonstrate that the till sampling frequency during detailed stages of diamond

exploration should be 2-3 samples per a square kilometer at minimum. Based on these case studies

kimberlite dispersal fans can be roughly divided into three zones (Fig. 6): (1) The proximal zone

characterized by abundant kimberlite fragments and a high concentration of kimberlitic material at the

base of the till bed. This zone extends approximately 500 m down-ice from Pipe 7 but it is apparently

absent in the Dyke 16 sampling area. (2) The intermediate zone starts where clearly elevated

numbers of indicator grains have reached the till surface. In Seitaperä this happens immediately at

the Dyke 16 contact but in Lahtojoki at a 500 m distance from Pipe 7. This is probably due to the

22

effect of grain liberation from coarser kimberlite fragments, which consequently become less

abundant in the till. The distance of this zone from the kimberlite is greatly affected by the kimberlite

rock-type. (3) The distal zone shows decreasing numbers of indicator grains, with the highest

concentrations in the upper parts of the basal till, and a virtual absence of kimberlite fragments. Zone

3 dilutes down-ice to the regional background level of kimberlitic indicators. In the Seitaperä study

area the dilution happens very near the body. Already at 1 km distance from Dyke 16 the counts are

extremely low. In the Lahtojoki study area the fan extends at least 2 km but its exact length could not

be determined because of a lake.

The recommended sample size is 60 kg of till at the minimum, which will result in approximately 45-50

kg of <1 mm material for the Knelson concentrator, providing that bigger rock fragments (a few

centimeters in diameter or more) are picked out already during sampling. Till samples even that size

taken at a few hundred-meter distance from the kimberlite may contain only a few 0.25-0.5 mm

fraction indicators, as seen in the Seitaperä case study (Paper IV). The denser the sampling grid and

the larger the sample size, the better the chances are to locate an indicator anomaly. The

interpretation of indicator mineral distribution results should be carefully studied in the context of field

observations. The recommended sampling media is basal till as it is a direct result of the last

glaciation and oriented according to the ice flow (e.g., Levson, 2001). In Finland basal till is usually

abundant and easily available, even though in most cases an excavator, and sometimes even a drill

rig, is required for efficient sampling. Ablation till is more easily reached by shovel but it lacks

orientation and, thus, it does not provide exact information about the source area. It is advisable to

take samples from different layers of the basal till bed (1-2 m spacing), because the indicator content

may vary considerably at different till horizons depending, e.g., on the transport distance (Paper IV).

The recovery rate of indicators should be controlled by regularly running spiked test-samples through

the processing line (Paper III). Usually a multi-stage till sampling program is needed to identify

possible kimberlitic targets for further studies, such as geophysical measurements or drilling. If there

are several possible targets, the most promising ones can be selected based on the size and content

of their indicator mineral fans. In the Lahtojoki case (Paper IV) the indicator maximum was discovered

surprisingly far away from Pipe 7 and the results also revealed the existence of another kimberlitic

body. Significantly, the drilled till samples taken near the Pipe 7 area contained abundant coarser (2-8

mm) kimberlite fragments even though their indicator counts were overall lower than in the maximum

area located further down-ice. This emphasizes the importance of studying coarser fractions of till

from interesting locations in the exploration area, on top of the sand-sized indicators.

The most time-consuming stage of diamond exploration is usually the microscopic observing (i.e.,

hand picking) of the kimberlitic indicator mineral grains from heavy mineral concentrates, traditionally

0.25-1.0 mm in grain size. In general, the coarser split (0.5-1.0 mm) can be observed rather

effectively, and in some indicator-rich areas, such as Lahtojoki (Paper IV), it already gives a good

estimate of the indicator distribution. However, in areas similar to Seitaperä (Paper IV) observation of

the finer split (0.25-0.5 mm) is necessary and requires much more effort. There are some applications

23

24

3

1

2

2

3

?

?

?

Figu

re 6

. Ind

icat

or m

iner

al d

ispe

rsal

fans

dow

n-ic

e fro

m P

ipe

7 in

Lah

tojo

ki (K

aavi

) and

Dyk

e 16

in S

eita

perä

(Kuh

mo)

div

ided

into

diff

eren

t zon

es a

ccor

ding

to

the

cont

ent o

f kim

berli

tic m

ater

ial i

n til

l (di

scus

sed

in te

xt).

The

bord

ers

of th

e ki

mbe

rlite

s ar

e m

arke

d an

d th

e ne

wly

dis

cove

red

sate

llite

body

of P

ipe

7 is

als

o in

dica

ted.

The

arr

ows

poin

t to

the

ice

flow

dire

ctio

ns in

the

regi

ons.

The

dis

pers

al fa

n de

rived

from

Pip

e 7

has

mai

nly

resu

lted

from

the

youn

ger i

ce fl

ow e

vent

in

the

regi

on. T

he n

orm

aliz

ed to

tal i

ndic

ator

min

eral

con

tent

s in

the

0.25

-0.5

mm

frac

tion

of 1

5 kg

(Lah

tojo

ki) a

nd 4

5 kg

(Sei

tape

rä) t

ill s

ampl

es fr

om e

xcav

ated

site

s ar

e m

arke

d in

dou

ble

boxe

s. T

he v

alue

in th

e lo

wer

box

repr

esen

ts th

e in

dica

tor c

onte

nt a

t the

bas

e of

the

till b

ed a

nd th

e up

per v

alue

that

of t

he u

pper

bas

al ti

ll.

The

drill

ing

site

s in

the

Laht

ojok

i sam

plin

g ar

ea a

re m

arke

d w

ith d

ots,

the

shad

es o

f whi

ch in

dica

te th

e in

dica

tor a

bund

ance

in th

e de

epes

t 2-4

m la

yer o

f till

.

that can speed up the picking process or even replace it. For instance, a scanning electron

microscope (SEM) attached to an energy dispersive spectrometer (EDS) is a promising tool in

studying the mineral distribution in sediment samples (Vareikiene & Lehtonen, 2004). By using some

automatic applications of the SEM-EDS system, such as that described by Gu (2003), identification of

kimberlitic indicator minerals in certain types of heavy mineral concentrates can get even faster than

by traditional microscopic observing. Some minerals, such as chromite, cannot be usually identified

as a kimberlitic indicator without analysis by electron microprobe. If chromite grains form a major

portion of indicators recovered from an exploration area, at least a subset of them should be analyzed

from carefully selected samples. The key elements to be analyzed are Cr, Mg, and Ti (Papers III and

IV).

Till geochemistry failed to reveal the kimberlitic targets in the Lahtojoki and Seitaperä case studies

(Paper IV). The negative outcome can probably be explained by the small size of the selected

kimberlitic bodies and/or a wrong till size fraction (<0.063 mm) having been subjected to chemical

analysis (e.g., Lintinen, 1995). Experiences in till geochemistry from the Lake Timiskaming field,

Canada, for example, demonstrate that coarser fractions of till can be more effective sample media to

detect a kimberlitic response, as these reflect better the incorporation of kimberlite debris in the form

of kimberlitic olivine (McClenaghan et al., 2002). This example among several others indicates that till

geochemistry should not be abandoned as a regional diamond exploration method but should be

used with careful consideration.

25

4. MAIN CONCLUSIONS

The main conclusions of this thesis are:

(1) The lithospheric mantle underlying the southwestern edge of the Karelian Craton consists of

three compositional layers, each of them having a diverse evolutionary and metamorphic

history.

(2) The multi-stage origin of the uppermost garnet spinel harzburgite layer (<110 km)

distinguished by Ca-rich but Ti-, Y-, Zr- and REE-depleted “CCGE” pyropes is probably

related to the tectonic events during the Svecofennian orogeny.

(3) Subcalcic harzburgitic pyrope xenocrysts (“G10’s”) originating from the variably depleted

middle layer (110-180 km) bear evidence of an ancient depletion event, similar to that

observed from lithosphere underlying Archean cratons elsewhere.

(4) Lherzolitic pyrope varieties from the middle and the deepest (180-240 km) mantle layers

document metasomatic enrichment of Ti, MREE and HREE, probably caused by silicate melts

close in composition to a megacryst-forming magma.

(5) The melt metasomatism produced abundant megacrystal garnets and destroyed almost all

depleted material in layer 3, i.e., layer 3 is considered to represent a melt-enriched version of

layer 2.

(6) A peridotitic diamond window lies in the middle mantle layer between 140 and 180 km.

(7) Abundant Group I eclogitic garnet xenocrysts and, moreover, a few previously found highly

diamondiferous eclogite xenoliths, emphasize the potential of an eclogitic diamond source in

the Kaavi-Kuopio mantle section.

(8) The Lahtojoki and Seitaperä case studies demonstrate, that the extent, morphology, indicator

content, and internal structure of the kimberlite indicator dispersal fans formed in Quaternary

till can be different due to several reasons, such as the intrusion size, mineralogy, resistance,

and glacial history in the region.

(9) The abundance, distribution and type of the kimberlitic material present in the dispersal fan

provide information about the distance to the source and about the source itself.

(10) Local and more regional populations of till chromites can be identified and related to their

bedrock sources based on the composition and morphology of the grains.

(11) The detrital heavy mineral suite in basal till comprises both regional and local components.

Mineral compositions and relative amounts of heavy mineral fractions may vary significantly

between different localities.

(12) Diamond prospecting based on indicator minerals requires a good knowledge on the glacial

history in the region, carefully planned sampling programs to meet the local characteristic,

and efficient processing and analytical methods. Providing these factors are in order, indicator

mineral method can be a powerful tool in diamond exploration in the Fennoscandian Shield.

26

ACKNOWLEDGEMENTS This doctoral thesis was funded by the Academy of Finland (project #50602, “Diamonds and the

lithospheric mantle in Finland”) and the Geological Survey of Finland. I am grateful to the leader of the

project, Prof. Ilmari Haapala of the University of Helsinki, and to my supervisors at GTK, Drs. Hugh

O’Brien and Jukka Marmo, who from the start outlined the project and did their bit to follow it through.

I express my thanks to the collaborative institution, the University of Cape Town, especially to Prof.

John Gurney and Dr. Andreas Späth. I acknowledge with gratitude the advice I got form my

colleagues and co-workers at GTK, Prof. Matti Saarnisto, Dr. Petri Peltonen, Mr. Matti Tyni and Dr.

Kari Kinnunen. The effort of Mr. Bo Johanson and Mr. Lassi Pakkanen of the GTK E-beam Laboratory

is also greatly appreciated. I value the reviews of this thesis by Dr. Arto Luttinen of the University of

Helsinki and Dr. Sven Monrad Jensen of the Geological Survey of Denmark and Greenland. I am

grateful to Prof. Tapani Rämö of the University of Helsinki for guidance during the final steps of the

dissertation. Co-operation of A&G Mining, Conroy Diamonds and Gold plc., and, European Diamonds

plc. is acknowledged with gratitude. I thank my closest co-workers at the GTK Research Laboratory

for their support, Mrs. Sirkku Mäenluoma, Dr. Tegist Chernet, Mr. Ahti Nissinen, Ms. Sari Summanen

and Mrs. Leena Järvinen, to name a few. Finally, I want to express my gratitude to my family – my life-

companion Jarkko, my parents and my siblings − and to my friends for encouragement.

REFERECES Armstrong, J., 1999. KIDD database. Indian and Northern Affairs Canada, Yellowknife. Economic

Geology Series Open file 1999-03. Atkinson, W.J., 1989. Diamond exploration philosophy, practice, and promises: a review. In: J. Ross

(Ed.), Kimberlites and Related Rocks, Geological Society of Australia, Spec. Pub. 14, pp. 1075-1107.

Beard, A.D., Downes, H., Hegner, E., Sablukov, S.M., Vetrin, V.R., Balogh, K., 1998. Mineralogy

and geochemistry of Devonian ultramafic minor intrusions of the southern Kola Peninsula,

Russia:implications for the petrogenesis of kimberlites and melilitites. Contrib. Min. Petrol. 130, 288-

303.

Burgess, S.R., Harte, B., 2004. Tracing Lithosphere Evolution through the Analysis of

Heterogeneous G9–G10 Garnets in Peridotite Xenoliths, II: REE Chemistry. J. Petrol. 45, 609-633.

Calcagnile, G., 1982. The lithosphere-astenosphere system in Fennoscandia. Tectonoph. 90, 19-35.

Carbno, G.B., Canil, D., 2002. Mantle structure beneath the SW Slave Craton, Canada: Constraints

from Garnet Geochemistry in the Drybones Bay kimberlite. J. Petrol. 43 (1), 129-142.

Chernet, T., Marmo, J., Nissinen, A., 1999. Technical Note. Significantly improved recovery of

slightly heavy minerals from Quaternary samples using GTK modified 3"Knelson preconcentrator.

Min. Eng. 12, 1521-1526.

Clement, C.R., Skinner, E.M.W., 1985. A textural genetic classification of kimberlites. Trans. Geol.

Soc. S. Afr. 88: 403-409.

Clement, C.R., Skinner, E.M.W., Scott-Smith, B.H., 1984. Kimberlite re-defined. J. Geol. 32, 223-

228.

27

Dawson, J.B., Stephens, W.E., 1975. Statistical analysis of garnets from kimberlites and associated

xenoliths. J. Geol. 83, 589-607.

Erlich, E.I., Hausel, W.D., 2002. Diamond deposits: origin, exploration, and history of discovery.

Society for Mining, Metallurgy, and Exploration, Inc., Littleton, CO, USA, 374 pp.

Fipke, C.E., Gurney, J.J., Moore, R.O., 1995. Diamond exploration techniques emphasising

indicator mineral geochemistry and Canadian examples. Geol. Surv. Canada, Bull. 423, 86 pp.

Griffin W.L., Ryan C.G., 1995. Trace elements in indicator minerals: area selection and target

evaluation in diamond exploration. In: W.L. Griffin (edit.), Diamond Exploration: Into the 21st Century.

J. Geochem. Explor. 53: 311-337.

Griffin, W.L., Cousens, D.R., Ryan, C.G., Sie, S.H., Suter, G.F., 1989a. Ni in chrome pyrope

garnets: A new geothermometer. Contrib. Mineral. Petrol. 103, 199-202.

Griffin, W.L., Smith, D., Boyd, F.R., Cousens, D.R., Ryan, C.G., Sie, S.H., Suter, G.F., 1989b. Trace element zoning in garnets from sheared mantle xenoliths. Geochim. Cosmochim. Acta, 53,

561-567.

Gu, Y., 2003. Automated scanning electron microscope based mineral liberation analysis. Journal of

Minerals & Materials Characterization & Engineering 1: 33-41.

Gurney, J.J., 1984. A correlation between garnets and diamonds. In: Glover J.E. and Harris P.G.

(eds.), Kimberlite Occurrence and Origin: A basis for conceptual models in exploration. Geol. Dept.

and Univ. Ext., Univ. of WA, Publ. 8, pp. 143-166.

Gurney, J.J., Zweistra, P., 1995. The interpretation of the major element compositions of mantle

minerals in diamond exploration. In: W.L. Griffin (edit.), Diamond Exploration: Into the 21st Century. J.

Geochem. Explor. 53, 293-309. Helmstaedt, H.H., Gurney, J.J., 1995. Geotectonic controls of primary diamond deposits:

implications for area selection. In: W.L. Griffin (edit.), Diamond Exploration: Into the 21st Century. J.

Geochem. Explor. 53, 125-144. Hölttä, P., Huhma, H., Mänttäri, I., Peltonen, P., Juhanoja, J., 2000. Petrology and geochemistry of

mafic granulite xenoliths from the Lahtojoki kimberlite pipe, eastern Finland. In: Downes, H., Demaiffe,

D. & Kramm, U. (Eds.), Alkaline magmatism and xenoliths from the Baltic Shield SVEKALAPKO

Project, EUROPROBE. Lithos 51 (1-2), 109-133.

Kennedy, C.S., Kennedy, G.C., 1976. The equilibrium boundary between graphite and diamond. J.

Gephys. Res. 81, 2467-2470.

Kennedy, W.G., 1964. The structural differentation of Africa in the Pan-African episode. Res. Inst.

Afr. Geol., Univ. Leeds, 8th Annu. Rep., pp. 48-49.

Kinnunen, K.A., 2001. Photographic interpretation of morphology and surface textures of diamond

crystals from Kaavi kimberlite province in Finland. In: Autio, S. (Ed.), Geological Survey of Finland,

Current Research 1999-2000. Geol. Surv. Finland Special Paper 31, 41-46.

Kirkley, M.B., Gurney, J.J., Levinson, A.A., 1991. Age, origin, and emplacement of diamonds.

Gems and Gemology 27 (1), 2-25.

Kong, J.M., Boucher, D.R., Scott-Smith, B.H., 1999. Exploration and geology of the Attawapiskat

kimberlites, James Bay Lowland, northern Ontario, Canada. In: Gurney, J.J., Gurney, J.L., Pascoe,

28

M.D., Richardson, S.H. (Eds.), Proceedings of the 7th International Kimberlite Conference, Cape

Town. Red Rood Design cc, Cape Town, South Africa, pp. 446-448. Kontinen, A., Paavola, J., Lukkarinen, H., 1992. K-Ar ages of hornblende and biotite from Late

Archean rocks of eastern Finland; interpretation and discussion of tectonic implications. Geol. Surv.

Finland Bull. 365, 31 pp.

Kopylova, M.G., Russell, J.K., Stanley, C., Cookenboo, H., 2000. Garnet from Cr- and Ca-

saturated mantle: implications for diamond exploration. J. Geochem. Explor. 68, 183-199.

Korsman, K., Koistinen, T., Kohonen, J., Wennerström, M., Ekdahl, E., Idman, H., Pekkala, Y., 1997. Bedrock map of Finland 1:1000000. Geol. Surv. Finland, Espoo, Finland.

Kukkonen, I.T., Jõeleht, A., 1996. Geothermal modelling of the litosphere in the central Baltic Shield

and its southern slope. Tectonoph. 255 (1-2): 25-45.

Kukkonen, I.T., Peltonen, P., 1999. Xenolith controlled geotherm for the central Fennoscandian

Shield: Implications for lithosphere-asthenosphere relations. Tectonoph. 304 (4), 301-315.

Kukkonen, I.T., Kinnunen, K.A., Peltonen, P., 2003. Mantle xenoliths and thick lithosphere in the

Fennoscandian Shield. Physics and Chemistry of the Earth 28 (9-11), 349-360.

Levson, V.M., 2001. Regional till geochemical surveys in Canadian Cordillera: sample media,

methods and anomaly evaluation. In: M.B. McClenaghan, P.T. Bobrowsky, G.E.M. Hall and S.J. Cook

(Eds.), Drift Exploration in Glaciated Terrain. Geol. Soc., London, Spec. Publ. 185, pp. 45-68.

Lintinen, P., 1995. Origin and physical characteristics of till fines in Finland. Geol. Surv. Finland Bull.

379, 83 pp. McCandless, T.E., Gurney, J.J., 1989. Sodium in garnet and potassium in clinopyroxene: criteria for

classifying mantle eclogites. In: J.R. Ross et al. (Eds.), Kimberlites and Related Rocks Vol. 2: Their

Crustal/Mantle Setting, Diamonds and Diamond Exploration. Geol. Soc. Aust. Spec. Publ. 14,

Blackwell, pp. 827-832.

McClenaghan, M.B., Kjarsgaard, B.A., 2001. Indicator mineral and geochemical methods for

diamond exploration in glaciated terrain in Canada. In: M.B. McClenaghan, P.T. Bobrowsky, G.E.M.

Hall and S.J. Cook (Eds.), Drift Exploration in Glaciated Terrain. Geol. Soc., London, Spec. Publ. 185,

pp. 83-123.

McClenaghan, M.B., Ward, B.C., Kjarsgaard, I.M., Kjarsgaard, B.A., Ker, D.E., Dredge, L.A., 2002. Indicator mineral and till geochemical dispersal patterns associated with the Ranch Lake

kimberlite, Lac de Gras Region, NWT, Canada. Geochemistry: Exploration, Environment, Analysis 2,

299-320.

McDonough, W.F., Sun, S.S., 1995. The composition of the earth. Chem. Geology 120, 223-253.

Mitchell, R.H., 1986. Kimberlites, Mineralogy, Geochemistry and Petrology. Plenum Press, New York,

442 pp.

Mitchell, R.H., 1995. Kimberlites, orangeites, and related rocks. Plenum, New York, 410 pp.

Mitchell, R.H., Bergman, S.C., 1991. Petrology of lamproites. Plenum Press, New York, 447 pp.

Morgan, P., 1995. Diamond exploration from the bottom up: regional geophysical signatures of

litosphere conditions favorable for diamond exploration. In: W.L. Griffin (edit.), Diamond Exploration:

Into the 21st Century. J. Geochem. Explor. 53, 145-165.

29

Mutanen, T., Huhma, H., 2003. The 3.5 Ga Siurua trondhjemite gneiss in the Archean Pudasjärvi

Granulite Belt, northern Finland. Bulletin of the Geol. Soc. Finland 75, 51-68.

Nixon, P.H., van Calsteren, P.W.C., Boyd, F.R., Hawkesworth, C.J., 1987. Harzburgites with

diamond facies from southern African kimberlites. In: P.H. Nixon (Ed.), Mantle xenoliths, John Wiley &

Sons Ltd, pp. 523-533.

O’Brien, H.E., Tyni, M., 1999. Mineralogy and Geochemistry of Kimberlites and Related Rocks from

Finland. In: J.J. Gurney, J.L. Gurney, M.D. Pascoe and S.H. Richardson (Eds.), Proceedings of the

7th International Kimberlite Conference, Cape Town. Red Rood Design cc, Cape Town, South Africa,

pp. 625-636.

O'Brien, H.E., Lehtonen, M.L., Spencer, R.G., Birnie, A.C., 2003. Lithospheric Mantle in Eastern

Finland: 250 km 3D transect. In: Extended abstracts. 8th International Kimberlite Conference,

Victoria, BC, Canada, June 22-27th 2003. 5pp.

O’Brien, H.E., Peltonen, P, Vartiainen, H., 2005. Kimberlites, carbonatites and alkaline rocks. In: M.

Lehtinen et al. (Eds.), Precambrian Geology of Finland – Key to the Evolution of the Fennoscandian

Shield. Elsevier Science B.V., Amsterdam, xxx-yyy.

Pearson, D.G., 1999. The age of continental roots. Lithos 48, 171-194.

Peltonen, P., 1999. Silicification of garnet peridotite xenoliths from the Lahtojoki kimberlite pipe,

Finland. In: Gurney, J.J., Gurney, J.L., Pascoe, M.D., Richardson, S.H. (Eds.), Proceedings of the 7th

International Kimberlite Conference, Cape Town. Red Rood Design cc, Cape Town, South Africa, pp.

659-663.

Peltonen, P., Brügmann, G., 2005 (submitted). Origin of layered continental mantle (Karelian

craton, Finland): Geochemical and Re-Os isotope constraints. Lithos.

Peltonen, P., Mänttäri, I., 2001. An ion microprobe U-Th-Pb study of zircon xenocrysts from the

Lahtojoki kimberlite pipe, eastern Finland. Bull. Geol. Soc. Finland 73, pp. 47-58.

Peltonen, P., Kinnunen, K. A., Huhma, H., 2002. Petrology of two diamondiferous eclogite xenoliths

from the Lahtojoki kimberlite pipe, eastern Finland. Lithos 63 (3-4), 151-164.

Peltonen, P., Huhma, H., Tyni, M., Shimizu, N., 1999. Garnet peridotite xenoliths from kimberlites of

Finland: Nature of the continental mantle at an Archaean craton – Proterozoic mobile belt transition.

In: Gurney, J.J., Gurney, J.L., Pascoe, M.D., Richardson, S.H. (Eds.), Proceedings of the 7th

International Kimberlite Conference, Cape Town. Red Rood Design cc, Cape Town, South Africa, pp.

664-675.

Pollack, H.N., Chapman, D.S., 1977. On the regional variations of heat flow, geotherms and

lithosphere thickness. Tectonophysics 38, 279-279.

Ryan, C.G., Griffin, W.L., Pearson, N.J., 1996. Garnet geotherms: Pressure-temperature data from

Cr-pyrope garnet xenocrysts in volcanic rocks. J. Geophys. Res. 101, 5611-5625.

Schulze, D.J., 2003. A classification scheme fro mantle-derived garnets in kimberlite: a tool for

investigating the mantle and exploring for diamonds. Lithos 71, 195-213.

Shimizu, N., 1975. Rare earth elements in garnets and clinopyroxenes from garnet lherzolite nodules

in kimberlites. Earth Planet. Sci. Lett. 25, 26-32.

30

Shimizu, N., Richardson, S.H., 1987. Trace element abundance patterns of garnet inclusions in

peridotite-suite diamonds. Geochim. Cosmochim. Acta 51, 755-758.

Smith, C.B., 1983. Pb, Sr, and Nd isotopic evidence for sources of African Cretaceous kimberlites.

Nature 303, 51-54.

Stachel, T., Viljoen, K.S., Brey, G. & Harris, J.W., 1998. Metasomatic processes in lherzolitic and

harzburgitic domains of diamondiferous lithospheric mantle: REE in garnets from xenoliths and

inclusions in diamonds. Earth Planet. Sci. Lett. 159, 1-12.

Tyni, M., 1997. Diamond prospecting in Finland – a review. In: Papunen H. (edit.), Mineral Deposits:

Research and Exploration , Where do They Meet? Proceedings of the 4th SGA Meeting, pp.789-791.

Vareikienė, O., Lehtonen, M., 2004. Heavy minerals in the study of soil: techniques, their limitations

and advantages. Academia Scientiarum Lithuaniae. Geologija 46, pp. 1-7.

Wagner, P.A., 1914. The Diamond Fields of South Africa. 3rd Ed., 1971. Johannesburg: Transvaal

Leader. Woodland, A. B., Peltonen, P., 1999. Ferric iron contents of garnet and clinopyroxene and estimated

oxygen fugacities of peridotite xenoliths from the Eastern Finland Kimberlite Province. In: Gurney,

J.J., Gurney, J.L., Pascoe, M.D., Richardson, S.H. (Eds.), Proceedings of the 7th International

Kimberlite Conference, Cape Town. Red Rood Design cc, Cape Town, South Africa, pp. 904-911.

Yaxley, G.M., Green, D.H., Kamenetsky, V., 1998. Carbonatite metasomatism in the southeastern

Australian lithosphere. J. Petrol. 39, 1917-1930.

31