1. INTRODUCTION Pedogenic iron oxides and manganese oxides and hydroxides frequently occur in soils of different climat- ic regions. Due to their high pigmenting capacity they may influence the colour of soils or particular soil hori- zons. Manganese oxides and hydroxides have a stron- ger pigmenting effect upon soils than iron oxides and tend to overwhelm their colour when both constituents are present (SCHWERTMANN & FANNING, 1976). Iron oxides may occur evenly dispersed throughout the soil, concentrated in particular horizons or in certain morphological features such as mottles and nodules Distribution of Iron and Manganese in Terra Rossa from Istria and its Genetic Implications Goran DURN, Dragutin SLOVENEC and Marta »OVI∆ (SCHWERTMANN & TAYLOR, 1989). Manganese oxides and hydroxides mainly occur as coatings on oth- er soil particles, as concretions and mottles, in cracks and veins or are mixed with iron oxides and other soil constituents in nodules (FANNING & FANNING, 1989; McK E N Z I E, 1989). The most common pedoge- netic iron oxides in soils are goethite and haematite. Lepidocrocite, maghemite and ferrihydrite are less common and occur in more specific pedogenic environ- ments. Although manganese oxides and hydroxides in soils are often reported to be amorphous (e.g. SIDHU et al., 1977) this is probably caused by difficulties in detecting small concentrations of poorly crystalline phases as with amorphous iron oxides (S C H W E R T- MANN & TAYLOR, 1989). According to McK E N Z I E (1989), the most common crystalline manganese oxides in soils are birnessite, vernadite, lithiophorite and hol- landite. Both the type of Fe-oxide in soil and its degree of crystallinity may be regarded as indicators of the pedo- genic environment because it seems to reflect the con- ditions under which they form (SCHWERTMANN & T A Y L O R, 1989). Consequently, they are sensitive to changing soil environment. This means that, for exam- ple, Fe-oxides can be remobilized by reduction and again reoxidized in another favourable environment. The behaviour of manganese in soils also depends on changing soil environment. However, oxidized forms of manganese may be reduced to the mobile Mn2 + at a higher Eh than that at which Fe3+ may be reduced to the more mobile Fe2 + (FANNING & FANNING, 1989). Namely, LOVLEY (1995) stated that at pH 7 and 25°C reduction of Mn4 + begins below a redox of 526 mV , while reduction of Fe3+ starts at Eh <-47 mV. According to LOVLEY & PHILLIPS (1988) between the zone of Mn reduction and that of Fe reduction, there is little overlap because soil bacteria show an enzymatic prefer- ence for Mn4 +, and Fe3 + reduction will not begin until Mn4+ is depleted. Identification of Fe oxides and especially Mn oxides and hydroxides in soil by X-ray diffraction (XRD) may be complicated due to their small amounts, overlapping of diagnostic lines with those of associated minerals and the poor crystallinity of some phases. For identifi- cation and even quantification of secondary Fe oxides in soils, several differential dissolution treatments can Geologia Croatica 54/1 27 - 36 11 Figs. 2 Tabs. ZAGREB 2001 Key words: Terra rossa, Iron, Manganese, Iron oxi- des, Istria, Croatia. University of Zagreb, Faculty of Mining, Geology and Petroleum Engineering, Pierottijeva 6, HR-10000 Zagreb, Croatia. e-mail: gdurn@rudar.rgn.hr Abstract Haematite and goethite are the predominant pedogenic mineral phas- es in terra rossa from Istria. Limited variation of selected Fe oxide characteristics in analysed samples indicates the specific pedogenic environment in which terra rossa is formed. The mean value of the Fed/Fet ratio, taken as an index of weathering is 0.7 and reflects quite a high degree of weathering of Fe-containing primary silicates. Rela- tively uniform Fed/ clay ratios clearly indicate a predominance of co- illuviation of clay and Fe oxides. Low values of Feo point to the low content of poorly crystalline Fe oxides in analysed terra rossa. F eo/F ed ratios in the analysed profiles vary. In the upper part of the Pomer, PoreË and Novigrad profiles Feo/F ed ratios are higher than in the lower part of those profiles which may be explained as a result of different pedogenic environments and/or additions of external mate- rials in the upper part of those profiles. The lack of positive correla- tion between Mnd and clay and Mnd and Fed is a consequence of remobilization of manganese due to hydromorphic processes which post-dated co-illuviation of clay and Fe oxides. During the Late Tertiary and Quaternary the pedogenic environ- ment on hard carbonate rocks of the Jurassic-Cretaceous-Palaeogene carbonate plain of southern and western Istria generally remained suitable for rubification. However, (Neo)tectonic activity and the input of external material, as well as various climatic fluctuations might have significantly effected terra rossa through the processes of erosion, colluviation, yellowing and secondary hydromorphy. 28 Geologia Croatica 54/1 be used. The most useful among them are an acid ammonium oxalate treatment and dithionite-citrate bicarbonate treatment. The former treatment extracts only poorly crystalline Fe oxides, predominately ferri- hydrite (SCHWERTMANN, 1964; SCHWERTMANN et al., 1982; SCHWERTMANN & TAYLOR, 1989) , while the dithionite-citrate bicarbonate treatment (pro- posed by MEHRA & JACKSON, 1960) extracts practi- cally all secondary Fe oxides (SCHWERTMANN & TAYLOR, 1989). Acid ammonium oxalate extractable iron and manganese are usually labelled Feo and Mno while dithionite-citrate bicarbonate extractable iron and manganese are labelled Fed and Mnd. Those parameters, when compared to total contents of given elements can be very helpful in understanding some aspects of the genesis of soils which contain iron oxides and man- ganese oxides and hydroxides such as Red Mediter- ranean soils which are situated on hard limestones and dolomites and are called terra rossa. For example, the occurrence and transformations of iron and manganese in a colluvial terra rossa toposequece of Northern Italy are excellently presented by BOERO & SCHWERT- MANN (1987). The Croatian soil classification puts terra rossa in the class of Cambic soils (©KORI∆, 1986). F.A.O. leg- ends (FAO, 1974) and Soil Taxonomy (SOIL SURVEY STAFF, 1975) classify most of terra rossa as Luvisols or Alfisols. However, the problem of classification of Red Mediterranean soils is still open and controversial. Namely, ROQUERO (1993) insisted on a more accu- rate classification of Red Mediterranean soils while FEDOROFF (1997), recognised and proposed six dif- ferent types of argillic horizons in these soils, based on abundance and morphology of textural features in the B horizons. RACZ (1999) proposed a new approach for the classification of terra rossa soils in Croatia. The aim of this paper is to present data on the distri- bution of iron and manganese in terra rossa from Istria and to evaluate any genetic implications. Terra rossa is the most extensive soil type in Istria. Its diagnostic fea- ture is a red colour (5YR to 10R Munsell hues) and is the result of rubification, i.e. formation of haematite. Terra rossa covers hard limestone and dolomite in the form of a discontinuous layer ranging in thickness from a few centimetres to several metres and is considered a polygenetic palaeosol or pedo-sedimentary colluvial Fig. 1 Distribution of terra rossa in Istria (modified after ©KORI∆, 1987). Legend: 1) terra rossa, calcocambiosol, eutric cambio- sol, rigosol from terra rossa (area of terra rossa >70%); 2) terra rossa, calcocambiosol, eutric cambiosol (area of terra rossa is 50-70%); 3) eutric cam- bisol on loess, luvisol on loess, rigosol from luvic and eutric cambiosol (40:30:30); (4) loca- tion of terra rossa samples. complex (DURN, 1996; DURN et al., 1999). DURN et al. (1999) presented evidence for the polygenetic nature of terra rossa in Istria, based on a detailed mineralogical and geochemical investigation. They concluded that both loess, older than that of the Upper Pleistocene age, and flysch might have contributed to the genesis of ter- ra rossa. DURN et al. (1999) also found that the arith- metic means of two populations (Fed in 45 terra rossa samples from various locations around the world and F ed in 40 samples from Istria) represent two indepen- dent estimates of the same population (Fed in terra rossa) and concluded that this supports BOERO & SCHWERTMANN’s (1989) conclusion, that the rather limited extent of the variation of selected Fe-oxide characteristics may indicate a specific pedogenic envi- ronment in which terra rossa is formed. In this paper terra rossa refers to reddish soils (5YR to 10R dry Munsell colours) overlying hard and perme- able limestone and dolomite. 2. STUDY AREA The Istrian peninsula belongs to the NW part of the Adriatic Carbonate Platform and is composed of Upper Jurassic and Cretaceous shallow water carbonate-rocks, Palaeogene carbonate and clastic rocks, and Neogene and Quaternary sediments. The Jurassic-Cretaceous- Palaeogene carbonate plain of southern and western Istria, Cretaceous-Palaeogene carbonate-clastic zone characterised by overthrusting structures in eastern and north-eastern Istria, and the Palaeogene flysch basin of central Istria are three main regions in Istria from the geological point of view (VELI∆ et al., 1995). Istria is an example of a non-isolated karst terrain which was effected by karst processes, (Neo)tectonic activity and input of external material since the Late Tertiary (DURN et al., 1999). The most important external material was loess, the deposition of which, according to CREMASCHI (1990), has affected Istria and the Dalmatian Archipelago since the early Middle Pleis- tocene. Terra rossa is dominantly situated on the Jurassic- Cretaceous-Palaeogene carbonate plain of southern and western Istria (Fig. 1) in the form of a discontinuous surface layer up to 2.5 metres thick or the infill of cracks and sinkholes. Pedo-sedimentary colluvial com- plexes which are found in karst depressions and are up to 14 metres thick also contain terra rossa like material (DURN 1996; DURN et al., 1999). 3. MATERIALS AND METHODS Terra rossa samples from B horizons on limestone and dolomite of Jurassic and Cretaceous age were sampled. Forty samples were taken at 14 localities. Particle size analysis was determined on the <2 mm fraction after dispersion in water and treatment with ultrasound. Selected fractions were obtained using wet sieving and quantitative separation by sedimentation in cylinders after the appropriate settling time. Total iron (Fet) and manganese (Mnt) were measured in <2 mm fraction of terra rossa samples with an XRF spectrometer (Philips PW 1404) using pressed powder pellets. Iron and man- ganese extractable with Na dithionite-citrate bicarbon- ate (Fed, M nd) were obtained after the method of MEHRA & JACKSON (1960) modified after U. SCHWERTMANN (pers. comm.), and measured with AAS (Pye-Unicam SP9). Iron and manganese extracta- ble with ammonium oxalate (Feo, M no) were acquired according to SCHWERTMANN (1964) and measured with AAS (Pye-Unicam SP9). The iron oxide mineralogy of the clay fraction (<2 µm) of terra rossa was determined by X-ray powder dif- fraction (XRD) using a Philips diffractometer (graphite monochromator, CuKα radiation, proportional counter). XRD patterns of non-oriented samples were taken after the following treatments: (a) air-drying and (b) dissolu- tion in HCl (1:1). 4. RESULTS 4.1. Mineralogy XRD analysis showed that the clay fraction of all terra rossa samples from Istria contains both haematite and goethite (Fig. 2). Broadening of diagnostic haematite and goethite XRD peaks indicates a low degree of crys- tallinity and their pedogenic origin. Munsell hue of dry terra rossa samples ranges between 5YR and 10R (Table 1) which indicates that red haematite masks the yellow colour of goethite. Ferrihydrite (which also gives a red colour to soil) and Mn-oxides or hydroxides were not detected by XRD, probably due to their small amounts and poor crystallinity. 29Durn, Slovenec & »oviÊ: Distribution of Iron and Manganese in Terra Rossa... Fig. 2 Characteristic parts of XRD patterns of terra rossa sample 1 (<2 µm fraction). Legend: a) air dried; b) treated with HCl (1:1, 24 h); Kl - kaolinite; I - illitic material; Q - quartz; H - haematite; G - goethite. 30 Geologia Croatica 54/1 SCHWERTMANN & TAYLOR (1989) stated that synthetic haematite crystals produced from ferrihydrite at near-pedogenic conditions generally exhibit a hexa- gonal morphology. Estimated crystal size of haematite from the XRD and TEM data according to the same authors is 20-50 nm in width and 10-20 nm in thick- ness. According to TORRENT & SCHWERTMANN (1987) with the increased size of haematite crystals the colour changes from red to purple. In order to establish the colour of haematite in relation to the size of haema- Sam. Depth Color Clay* Silt* Sand* Fe t* Fed* Feo Fet/ Feo/ Fed/ Mn t Mnd Mno Mnd/ Mno/ (cm) Fed* Fed clay Mn t Mnd Profile Sjenokoπa 1. 20-35 5YR4/8 41.1 58.3 0.6 4.20 2.78 0.39 0.66 0.14 0.068 1037 909 606 0.88 0.67 2. 35-100 5YR5/6 47.5 51.2 1.4 4.44 2.96 0.39 0.67 0.13 0.062 937 821 646 0.88 0.79 3. 100-145 5YR5/6 49.4 49.1 1.5 4.60 2.89 0.37 0.63 0.13 0.059 999 805 518 0.81 0.64 4. 145-200 5YR5/6 51.8 46.0 2.2 4.68 2.84 0.36 0.61 0.13 0.055 1192 558 330 0.47 0.59 5. 200-225 5YR5/6 54.2 45.3 0.5 5.12 3.22 0.44 0.63 0.14 0.059 1015 1000 839 0.99 0.84 6. 225-265 5YR5/6 53.0 46.1 0.9 4.90 3.34 0.53 0.68 0.16 0.063 984 969 694 0.98 0.72 7. 265-300 5YR5/6 57.1 42.6 0.2 5.29 3.22 0.40 0.61 0.12 0.056 883 800 555 0.91 0.69 8. 300-325 5YR4/6 61.1 38.6 0.3 5.75 3.78 0.60 0.66 0.16 0.062 1185 1100 1093 0.93 0.99 9. 325-335 5YR4/6 59.8 40.1 0.1 5.45 3.80 0.56 0.70 0.15 0.064 875 667 428 0.76 0.64 10. 335-360 5YR4/6 71.7 28.3 0.1 5.92 3.57 0.39 0.60 0.11 0.050 976 494 303 0.51 0.61 11. 360-400 5YR5/6 59.9 39.8 0.3 4.91 3.43 0.48 0.70 0.14 0.057 782 484 302 0.62 0.62 12. 400-450 2.5YR4/6 60.1 39.4 0.4 4.97 3.94 0.62 0.79 0.16 0.066 1084 1020 909 0.94 0.89 13. 450-550 2.5YR4/6 63.1 36.4 0.5 5.22 3.39 0.36 0.65 0.11 0.054 1200 391 251 0.33 0.64 14. 550-650 2.5YR4/6 64.6 34.8 0.6 5.51 3.79 0.31 0.69 0.08 0.059 1386 1173 933 0.85 0.80 15. 650-750 2.5YR5/8 66.4 32.7 0.9 5.51 3.96 0.30 0.72 0.08 0.060 1348 1127 909 0.84 0.81 16. 750-850 2.5YR4/6 62.9 36.2 0.8 5.41 4.00 0.28 0.74 0.07 0.064 1348 1255 1052 0.93 0.84 Profile Kamnik 19. 10-25 5YR4/6 32.5 65.1 2.4 3.37 1.86 0.29 0.55 0.16 0.057 930 614 532 0.66 0.87 19A 25-50 5YR4/6 35.7 61.9 2.4 3.40 2.03 0.30 0.60 0.15 0.057 852 523 513 0.61 0.98 Profile Pomer 22. 18-30 5YR5/6 45.1 52.7 2.2 3.74 2.42 0.31 0.65 0.13 0.054 1084 733 655 0.68 0.85 23. 30-62 2.5YR4/8 64.2 34.0 1.9 6.20 4.15 0.47 0.67 0.11 0.065 1084 782 626 0.72 0.80 24. 62-105 2.5YR4/8 71.1 28.0 0.9 6.03 4.11 0.37 0.68 0.09 0.058 1084 817 661 0.75 0.81 25. 105-130 10R4/8 70.2 28.7 1.1 5.60 3.91 0.38 0.70 0.10 0.056 930 662 512 0.71 0.77 Profile PoreË 53. 15-35 5YR4/6 59.4 39.5 1.1 6.10 4.01 0.46 0.66 0.12 0.054 1240 1093 741 0.88 0.68 55. 35-50 2.5YR5/6 74.1 25.2 0.8 4.70 3.19 0.36 0.68 0.11 0.054 1007 714 490 0.71 0.69 Profile Mondolaco 60. 12-30 2.5YR5/6 61.3 36.0 2.7 6.57 5.20 0.35 0.79 0.07 0.085 930 897 630 0.96 0.70 61. 30-60 2.5YR5/6 67.7 30.8 1.5 6.70 5.11 0.30 0.76 0.06 0.075 1782 1596 1072 0.90 0.67 Profile Novigrad 131. 9-20 5YR4/6 55.7 43.2 1.1 4.04 3.19 0.46 0.79 0.14 0.057 1471 956 773 0.65 0.81 132. 20-42 5YR4/6 60.3 38.6 1.2 412 3.64 0.36 0.88 0.10 0.060 1394 946 760 0.68 0.80 133. 42-63 2.5YR4/6 70.1 29.0 1.0 6.07 4.61 0.34 0.76 0.07 0.066 775 543 337 0.70 0.62 134. 63-80 2.5YR4/6 73.1 26.0 0.9 6.12 4.62 0.37 0.75 0.08 0.063 775 676 446 0.87 0.66 135. 80-112 2.5YR4/6 73.5 25.4 1.1 6.20 4.71 0.30 0.76 0.06 0.064 620 395 202 0.64 0.51 136. 112-150 2.5YR4/6 77.2 21.9 0.9 6.51 4.90 0.36 0.75 0.07 0.063 775 477 308 0.62 0.65 Single samples 42. 5-15 5YR5/6 46.8 47.9 5.3 4.06 2.67 0.52 0.66 0.20 0.057 2401 1956 1900 0.81 0.97 52. 10-30 2.5YR4/8 42.1 55.3 2.5 5.47 3.76 0.30 0.69 0.08 0.089 387 287 87 0.74 0.30 100. 10-25 2.5YR4/8 72.1 26.8 1.1 6.67 5.24 0,47 0.79 0.09 0.073 620 326 275 0.53 0.84 101. 10-25 2.5YR5/8 73.8 25.4 0.8 7.06 5.74 0,48 0.81 0.08 0.078 779 394 201 0.51 0.51 234. 5-15 5YR4/8 35.8 61.1 3.1 3.38 2.30 0.25 0.68 0.11 0.080 1007 749 656 0.74 0.88 235. 5-10 2.5YR4/6 35.1 63.9 1.0 4.13 3.53 0.32 0.85 0.09 0.101 2169 1495 1252 0.69 0.84 236. 5-10 2.5YR4/6 42.1 57.3 0.6 4.33 3.31 0.36 0.76 0.11 0.079 1162 757 549 0.65 0.73 237. 5-10 2.5YR4/6 59.1 40.4 0.5 5.45 3.99 0.36 0.73 0.09 0.068 1084 730 463 0.67 0.63 Table 1 Particle size analysis, total, dithionite and oxalate extractable iron (%) and manganese (ppm) in terra rossa samples (<2 mm). Color (dry) after MUNSELL SOIL COLOR CHARTS (1994). Clay (<2 µm), silt (2-63 µm) and sand (>63 µm); * data from DURN et al. (1999). 31Durn, Slovenec & »oviÊ: Distribution of Iron and Manganese in Terra Rossa... tite crystals, we prepared several samples of haematite by heating (for 1 hour) fine grained goethite at follow- ing temperatures: 300°, 500°, 700°, 900°, 1000° and 1300°C. With the increased temperature of heating, the size of haematite crystals also increased, which resulted in smaller broadening of XRD lines in relation to the instrumental broadening. The profiles of the 110 line at Θ ≈1 7 .9° for haematite samples prepared at 300° and 1300°C are shown in Fig. 3. The size of haematite crys- tals in direction <1 10> was determined from broaden- ing of the 110 line after correction for instrumental broadening. The 110 line of haematite prepared at 1 3 00°C was used as a measure for instrumental broad- ening. The size of haematite crystals prepared at 300° , 5 00° and 700°C is about 21, 22 and 28 nm respective- ly, and their colour (powder) is red (10R4/8). The size of crystals prepared at 900°C is about 45 nm and they are dark red in colour (10R3/6). After heating to 1 0 00°C, crystal size increased to about 75 nm and colour changed to dusky red (10R3/4). The colour of a sample prepared at 1300°C is purple-red (no matching colour with Munsell Soil Colour Charts). We also pre- pared two homogeneous mixtures with a powder of kaolinite (33 wt.%), mica (30 wt.%), glass (20 wt.% ) , quartz (5 wt.%), goethite (6 wt.%) and haematite (6 w t.%). Mixture 1 contained haematite prepared at 300°C while mixture 2 contained haematite prepared at 1 0 00°C. Both mixtures matched the average composi- tion of terra rossa clay fraction with two exceptions. The clay fraction of natural terra rossa contains illitic material and amorphous inorganic compound instead of mica and glass. The colour of the clay fraction of natur- al terra rossa is almost identical to the colour of mixture 1 which contain haematite prepared at 300°C. This matching of colour and broadening of 110 line at Θ ≈1 7 .9° (Fig. 2) clearly indicate that the size of hae- matite crystals in analysed terra rossa samples is lower than 50 nm which is in accordance with investigation of BOERO & SCHWERTMANN (1989). 4.2. Total, Na dithionite-citrate bicarbonate and ammonium oxalate extractable iron and manganese The results of the analysis of total, Na dithionite-citrate bicarbonate and ammonium oxalate extractable iron and manganese are presented in Table 1. Mean values F et, Fed and Feo are 5.19 %, 3.68 % and 0.39 % respec- tively (Table 2). The mean value of the Fed/ F et r a t i o , which is taken as an index of weathering (e.g. ARDUINO et al., 1984; BOERO & SCHWERT- MANN, 1987; BECH et al., 1997) is 0.7 and reflects quite a high degree of weathering of Fe-containing pri- mary silicates. Relatively uniform Fed/F et ratios in the analysed profiles (Table 1) are also shown by the high positive correlation coefficient (Fig. 4) and point to a clear relationship between Fed andFet in terra rossa. The clay content in analysed terra rossa samples increases with depth in the profiles (Table 1). An exception is the Sjenokoπa profile, a colluvial terra rossa complex com- posed of superimposed features of colluviation and pedogenesis, where the clay content increases with depth within the each superimposed cycle (DURN, 1996; DURN et al., 1999). Fed/clay ratios are relatively uniform (Table 1, Fig. 5) and clearly indicate a predom- Fig. 3 The profiles of the 110 line of haematite prepared by heating of goethite at 300° and 1300°C. n x s CV (%) 0.95 C.I. Fet* 40 5.19 0.99 19.07 4.88-5.51 Fed* 40 3.68 0.89 24.18 3.40-3.96 Feo 40 0.39 0.09 23.08 0.36-0.42 Fed/Fet* 40 0.70 0.07 10.00 0.68-0.72 Feo/Fed 40 0.112 0.033 29.46 0.102-0.122 Mnt 40 1089 379 34.80 968-1210 Mnd 40 817 353 43.21 704-930 Mno 40 625 341 54.56 516-734 Mnd/Mnt 40 0.74 0.15 20.27 0.69-0.79 Mno/Mnd 40 0.73 0.14 19.18 0.685-0.775 Table 2 Basic statistics for total, dithionite and oxalate extractable iron (%) and manganese (ppm) in terra rossa samples (<2 mm). Legend: n) number of samples; x) arithmetic mean; s) standard devia- tion; CV) coefficient of variation (%); 0.95 C.I.) confidence interval. * data from DURN et al. (1999). 32 Geologia Croatica 54/1 Fig. 4 Ratio between dithionite soluble iron (Fed) and total iron ( F et) in terra rossa; r is signifi- cant at level p<0.01. Fig. 6 Ratio between oxalate extractable iron (Feo) and dithionite extractable iron (Fe d) of terra rossa samples; r is not significant at the level p<0.01. Fig. 5 Ratio between the amount of clay fraction (<2 µm) and dithionite soluble iron (Fed) in terra rossa; r is significant at level p<0.01. 33Durn, Slovenec & »oviÊ: Distribution of Iron and Manganese in Terra Rossa... inance of co-illuviation of clay and Fe oxides, i.e. con- nection of Fe oxides with the clay fraction. Low values of Feo (Tables 1 and 2) point to the low content of poor- ly crystalline Fe oxides in analysed terra rossa. Feo/F ed ratios in the analysed profiles vary (Table 1, Fig. 6). In the upper part of the Pomer, PoreË and Novigrad pro- files (5YR Munsell hues) Feo/Fed ratios are higher than in the lower part of those profiles (2.5YR and 10R Munsell hues). Significant positive correlation (r = 0.7) between Feo and Fed in samples with 5YR Munsell hues was observed (Fig. 7). Mean values of Mnt, Mnd and Mno are 1089 mg/kg, 817 mg/kg and 625 mg/kg respectively (Table 2). The mean value of the Mnd/M nt ratio is 0.74 and indicates that most of the manganese in terra rossa is in the form of Mn oxides and hydroxides. A high positive correla- tion between Mnd and Mnt (Fig. 8) points to a clear rela- tionship between Mnd a n dM nt in terra rossa. Values of Mno are lower compared to those of Mnd (Tables 1 and 2), and a high positive correlation between them (Fig. 9) is in accordance with the results of BLUME & SCHWERTMANN (1969) who concluded that the presence of Mn oxides and hydroxides in soil should be recognised on Mnd values rather than on Mno v a l u e s which does not provide any new qualitative informa- tion. Consequently, in further texts only Mnd v a l u e s will be evaluated. The lack of positive correlation between Mnd and clay (Fig. 10) and Mnd and Fed ( F i g . 11) show that Mn oxides and hydroxides are not con- nected with the clay fraction and Fe oxides. Namely, they either do not participate in the co-illuviation of clay and Fe oxides, or are remobilized later due to suc- cessive processes. 5. DISCUSSION AND CONCLUSION The association of haematite and goethite as predomi- nantly pedogenic mineral phases, and the rather limited variation of selected Fe oxide characteristics in terra Fig. 7 Ratio between oxalate ex- tractable iron (Feo) and dithi- onite extractable iron (Fed) of 5YR terra rossa samples; r is significant at the level p<0.01. Fig. 8 Ratio between dithionite extractable manganese (Mnd) and total manganese (Mnt) of terra rossa samples; r is signif- icant at the level p<0.01. 34 Geologia Croatica 54/1 rossa from Istria, indicate specific pedogenic environ- ment in which terra rossa is formed. Our results match those of BOERO & SCHWERTMANN (1989) who concluded that this pedogenic environment is character - ized by an association of a Mediterranean climate, high internal drainage and neutral pH conditions. In the 30 samples examined (fraction < 2 µm) they also estimated that the haematite /(haematite+goethite) ratios averaged 0 . 62±0.055 i.e. the content of haematite is higher than the content of goethite in terra rossa. SINGER et al. (1998) interpreted the difference in iron oxide mineral- ogy between terra rossa (dominated by haematite) and rendzina (dominated by goethite) developed on lime- stones in Israel as a consequence of soil climate, specif- ically the moisture regime of two soils. Poorly crystalline Fe oxides were not detected with XRD in analysed terra rossa. However, low values of F eo (Tables 1 and 2) point to their very low (probably ferrihydrite) content. This was expected because most of the Fe-oxides in fersialitic soils are crystallised (BECH et al., 1997). Relatively uniform Fed/ clay ratios (Table 1, Fig. 5) clearly indicate a connection of Fe oxides with the clay fraction, as iron oxides are translocated downwards together with clay illuviation. The main pedogenic process which characterises terra rossa in Istria is the accumulation of clay and is manifested through: (1) clay illuviation in the form of coatings and/or (2) microaggregates incorporated in the groundmass due to argilloturbation and soil stress (DU R N, 1996). Accord- ing to FE D O R O F F (1997) rubification occurs in upper horizons, then rubified soil material is translocated with clays to depth. In general, the Feo/Fed ratio decreases with depth in the Kamnik, Pomer, PoreË, Mondolaco and Novigrad profiles. A decrease of this ratio at depth in profiles was also observed by BECH et al. (1997) in some red Mediterranean soils from northern Spain. Significant positive correlation (r = 0.7) between Feo and Fed i n samples with 5YR Munsell hues, and the higher content of Feo in samples with 5YR Munsell hues (0.137 % ) compared to samples with 2.5YR and 10R Munsell Fig. 9 Ratio between oxalate ex- tractable manganese (Mno) and dithionite extractable man- ganese (Mnd) of terra rossa samples; r is significant at the level p<0.01. Fig. 10 Ratio between the amount of clay fraction (<2 µm) and dithionite extractable manga- nese (Mnd) of terra rossa sam- ples; r is not significant at the level p<0.01. 35Durn, Slovenec & »oviÊ: Distribution of Iron and Manganese in Terra Rossa... hues (0.089 %) may indicate a difference between those two groups of samples in respect to their pedogenic environment and/or parent material and, consequently, their content of poorly crystalline Fe oxides. The differ- ence in pedogenic environment can be explained with the higher content of organic material in the upper parts of analysed profiles. BECH et al. (1997) concluded that there is a particular tendency for Feo to correlate with organic matter content (r = 0.68). Namely, organic mat- ter in soil inhibits the crystallisation of Fe oxides (BLUME & SCHWERTMANN, 1969). During recent or subrecent pedogenesis a yellowing process might have occurred by preferential dissolution of haematite due to the higher thermodynamic stability of goethite. According to BOERO & SCHWERTMANN ( 1 9 8 7 ) , during that process the goethite/haematite ratio increas- es, Fed does not change significantly while Feo m a y increase if inhibitors which prevent its transformation to more stable Fe oxide phases are present. The pres- ence of external materials in the upper parts of terra rossa profiles in Istria was postulated by ©INKOVEC (1974) and ©K O R I ∆ (1979), and confirmed by DURN (1996) and DURN et al. (1999), who found low-charge vermiculite or high-charge smectite in the <0.2 µm fraction of terra rossa samples which are situated bel- low or near Upper Pleistocene loess in Savudrija. DURN et al. (1999) concluded that the presence of this mineral in those samples may indicate its aeolian origin because it was detected as one of the main mineral phases in fine clay of the Upper Pleistocene loess in Savudrija which post-dated terra rossa formation. They also discovered that both loess older than that of the Upper Pleistocene age and flysch might have con- tributed to the genesis of terra rossa. We interpret the lack of a positive correlation between Mnd and clay (Fig. 10) and Mnd and Fed ( F i g . 11) as a consequence of hydromorphic processes which post-dated co-illuviation of clay and Fe oxides and resulted in the remobilization of manganese. During field work we observed the tendency of manganese oxides and hydroxides to concentrate as concretions and coatings on different soil features, mainly in the lower part of investigated profiles. Istria is an example of a non-isolated karst terrain which was affected by karst processes, Neotectonic activity and external material input since the Late Ter- tiary. Although climatic and biotic factors have changed during the Late Tertiary and Quaternary the pedogenic environment on the hard carbonate rocks of the Jurassic-Cretaceous-Palaeogene carbonate plain of southern and western Istria generally remained suitable for rubification (DURN, 1996; DURN et al., 1999). However, red Mediterranean soils can be affected by eluviation, yellowing and secondary hydromorphy because they have been exposed to various climatic fluctuations (FEDOROFF, 1997). 6. REFERENCES ARDUINO, E., BERBERIS, E., CANARO, F. & FORNO, M.G. (1984): Estimating relative ages from iron-oxi- de/total iron ratios of soils in the western Po valley (Italy).- Geoderma, 33, 39-52. BECH, J., RUSTULLET, J., GARIGÓ, J., TOBÍAS, F.J. & MARTÍNEZ, R. (1997): The iron content of some red Mediterranean soils from Northeast Spain and its pedo- genic significance.- Catena, 28, 211-229. BLUME, H.P. & SCHWERTMANN, V. (1969): Genetic eva- luation of profile distribution of Al, Fe and Mn oxides.- Soil Sci. Soc. Am. Proc., 33, 438-444. BOERO, V. & SCHWERTMANN, U. (1987): Occurrence and transformations of Iron and Manganese in a colluvial terra rossa toposequence of Northern Italy.- Catena, 14, 519-531. BOERO, V. & SCHWERTMANN, U. (1989): Iron oxide mineralogy of terra rossa and its genetic Implications.- Geoderma, 44, 319-327. CREMASCHI, M. (1990): The loess in Northern and Central Italy: a loess basin between the Alps and the Mediter- Fig. 11 Ratio between dithionite extractable manganese (Mnd) and dithionite extractable iron ( F ed) of terra rossa samples; r is not significant at the level p<0.01. 36 Geologia Croatica 54/1 ranean regions.- In: CREMASCHI, M. (ed.): The Loess in Northern and Central Italy. Centro di Studio per la Strati- grafia e Petrografia delle Alpi Centrali, Editrice Guten- berg, Milano, 15-19. DURN, G. (1996): Origin, composition and genesis of terra rossa in Istria.- Unpublished PhD Thesis (in Croatian with English summary), University of Zagreb, 204 p. DURN, G., OTTNER, F. & SLOVENEC, D. (1999): Miner- alogical and geochemical indicators of the polygenetic nature of terra rossa in Istria, Croatia.- Geoderma, 91, 125-150. FANNING, D.S. & FANNING, M.C.B. (1989): Soil, mor- phology, genesis, and classification.- John Wiley & Sons, New York, 395 p. FAO (1974): Soil map of the World, 1:5 M, Volume 1, Leg- end.- Unesco, Paris. FEDOROFF, N. (1997): Clay illuviation in Red Medi- terranean soils.- Catena, 26, 171-189. LOVLEY, D.R. & PHILLIPS, E.J.P. (1988): Manganese inhi- bition of microbial iron reduction in anaerobic sediments.- Geomicrobiology Jour., 6, 145-155. LOVLEY, D.R. (1995): Microbial reduction of iron, man- ganese, and other metals.- Advances in Agronomy, 54, 175-231. McKENZIE, R.M. (1989): Manganese oxides and hydrox- ides.- In: DIXON, J.B. & WEED, S.B. (eds.): Minerals in Soil Environments.- Soil Sci. Soc. Am., Madison, 439- 465. MEHRA, O.P. & JACKSON, M.L. (1960): Iron oxides removal from soils and clays by a dithionite-citrate-bicar- bonate system buffered with sodium bicarbonate.- 7th Nat. Conf. Clays Clay Miner., 7, 317-327. MUNSELL SOIL COLOR CHARTS (1994): Macbeth Divi- sion of Kollmorgem Instruments Corporation, New Wind- sor. RACZ, Z. (1999): Current global and local issues and prob- lems related to pedological research.- Agriculturae Con- pectus Scientificus, 64/3, 231-241. ROQUERO, C. (1993): Some problems on genesis and classi- fication of red soils in Spain.- Second International Meet- ing on “Red Mediterranean Soils”. Short papers and abstracts, Adana, Turkey, 8-10. SCHWERTMANN, U. (1964): Differenzierung der Eisen- oxide des Bodens durch photochemische Extraktion mit saurer Ammoniumoxalat-lösung.- Z. Pflanzenernähr. Bodenkund., 105, 194-202. SCHWERTMANN, U. & FANNING, D.S. (1976): Iron-man- ganese concretions in hydrosequences of soils in Bavaria.- Soil Sci. Soc. Am. J., 40, 731-738. SCHWERTMANN, U. & TAYLOR, R.M. (1989): Iron oxides.- In: DIXON, J.B. & WEED, S.B. (eds.): Minerals in Soil environments. 2nd ed., Soil Sci. Soc. Am. Book Series, 379-438. SCHWERTMANN, U., MURAD, E. & SCHULZE, D.G. (1982): Is there Holocene reddening (hematite formation) in soils of oxeric temperate areas?- Geoderma, 27, 209- 223. SIDHU, P.S., SEHGAL, J.L., SINHA, M.K. & RAND- HAWA, N.S. (1977): Composition and mineralogy of iron-manganese concretions from some soils of the Indo- Gangetic plain in northwest India.- Geoderma, 18, 241- 249. SINGER, A., SCHWERTMANN, U. & FRIEDL, J. (1998): Iron oxide mineralogy of terre rosse and rendzinas in rela- tion to their moisture and temperature regimes.- Eur. Jour. of Soil Sci., 49, 385-395. SOIL SURVEY STAFF (1975): Soil taxonomy: a basic sys- tem of soil classification for making and interpreting soil surveys. USDA Handbook No. 436, U.S. Gov. Printing Office, Washington. ©INKOVEC, B. (1974): The origin of terra rossa in Istria (in Croatian).- Geoloπki vjesnik, 27, 227-237. ©KORI∆, A. (1979): Two-layer soil profile on the area of ter- ra rossa in Istria (in Croatian).- Zemljiπte i biljka, 28, 111- 131, Zagreb. ©KORI∆, A. (1986): Origin, development and systematics of soil (in Croatian).- Spec. pub. of agricultural scientific meeting, Zagreb, 172 p. ©KORI∆, A. (1987): Pedosphere of Istria (in Croatian).- Pro- ject Council of Pedological Map of Croatia, Spec. ed., book 2, Zagreb, 192 p. TORRENT, J. & SCHWERTMANN, U. (1987): Influence of hematite on the colour of red beds.- J. Sedim. Petrol., 57, 121-125. VELI∆, I., TI©LJAR, J., MATI»EC, D. & VLAHOVI∆, I. (1995): A review of the geology of Istria.- In: VLA- HOVI∆, I. & VELI∆, I. (eds.): 1st Croatian geological congress, Excursion guide-book, 21-30, Zagreb. Manuscript received July 27, 2000. Revised manuscript accepted April 23, 2001.