The variability of soils and vegetation of hydrothermal fields in the Valley of Geysers at Kamchatka Peninsula

The catenary sequence of soils

The catena of Andosols down a slope near a hot spring in the Valley of Geysers was subdivided into four thermal zones (Fig. 1a–e), which are described below.

Zone I. Non-heated Eutrosilic Silandic Andosols (Arenic, Cutanic) under Kamchatka’s tall herb communities and fragmented Erman’s birch woods

Non-heated Andosols with temperatures of < 25 °C at the depth of 15 cm were stratified (Table S1, Figure S1). In the sandy A-horizon (Fig. 1f), the micromass of soil fabric was well-structured due to highly active micro- and mesofauna and contained numerous brownish coprolites typical for Andosols^30,31. Both modern and buried A-horizons showed a granular–angular blocky structure and contained plant remains, oval-shaped dense biogenic aggregates and small concentrations of brown collomorphic Al–Fe-humus material (Fig. 2a).

The Bt-horizon had a massive macrostructure (Table S2) and a pellicular grain microstructure consisting of microaggregates of loose consistency (containing silt-sized grains of plagioclases, weathered volcanic ash and heavy minerals³²) covered with collomorphic Fe-humus coatings (Fig. 2b). The soil horizon buried under volcanic ash contained many sand-sized opaque and optically isotropic minerals (opal/cristobalite, volcanic glass). The coagulation of finely-dispersed ferruginous and collomorphic substances resulted from ‘intra-soil metamorphism’³³ or ‘andozolization’³⁴—the main pedogenic process in non-heated Andosols. Neither illuvial clay coatings nor unaltered volcanogenic sand-silt material were present.

In the topsoil (0–10 cm), the particle-size distribution (Table S3) was characterized by the prevalence of PM_>50 with very small amounts of PM_5–50 (< 10%) and PM_<1 (0.5%), while the subsoil (Table S4) had higher contents of PM_<5 and PM_10–50 (abbreviations of particle-size fractions are explained in Methods), lower contents of PM_500–1000 and similar contents of PM_50–250 (Table S5). With depth, pH increased from 5.7 to 6.8 and redox potential (Eh) from − 35 to + 497 mV, while the value of electrical conductivity (EC) and the content of soil organic carbon (SOC) decreased by 2 and 30 times, respectively. A low EC value reflected an absence of salinization in non-heated Andosols. A high SOC content and a weakly acid reaction in the topsoil were favorable for the formation of Al-organic complexes and unfavorable for allophane formation^33,35.

In the Zone I soils, feldspars (albite-anorthite) were the main primary minerals constituents with cristobalite and rarer zeolites presented in smaller amounts. The non-heated Andosols also contained clay minerals, predominantly swelling minerals—dioctahedral smectites (probably montmorillonites). Smectites were unambiguously diagnosed by the X-ray diffraction patterns obtained from both powder and oriented air-dried and saturated with ethylene glycol specimens (Fig. 3a,e).

The upper 2 m-thick layer of non-heated Andosols had the mineralogical composition typical for this soil type^27,36 and parent materials in the Valley of Geysers³², with predominant feldspars and smectite, frequent kaolinite, zeolites and occasional quartz, anatase, pyrite and goethite (Table S6). According to SEM and XRD data, the mineralogical and elemental composition changed insignificantly with depth, except for the appearance of newly formed opal and rare fragments of etched volcanic glass at a depth of more than 15 cm (Fig. 2i).

Zone II. Slightly heated Eutrosilic Aluandic Andosols (Cutanic, Loamic, Natric) under tall herb communities at the periphery of the steam hydrotherm

Slightly heated Andosols had temperatures of 25–40 °C at a depth of 15 cm (Table 1). Their A-horizons (Fig. 1g) was better structured as compared to that of non-heated Andosols, which was due to higher contents of collomorphic Fe-organic material and clays (Table S7). Excrements of soil microbiota occurred within some microzones of the topsoil. Dense silt–clay microaggregates contained inclusions of volcanogenic rock fragments, clay minerals and ash and had a random packing, according to SEM data.

The hydrothermally-induced alteration of slightly heated Andosols caused considerable changes in soil fabric (Fig. 2c,d). In the Btn-horizon, the number and distinctness of clayey layers of different colors consisting of a brown collomorphic organomineral material and a finely-dispersed material increased in the direction towards the hydrothermal field. The number of clay coatings within intrapedal pores increased with depth. Such coatings were probably formed due to alkalization and release of Na during weathering of primary minerals (with exchangeable Na content of 12.9 (+)µmol/kg in the Btn-horizon and only 0.3–0.8 (+) µmol/kg in upper horizons, Table S5). Such a process is typical for soils of the margins of thermal fields in the ultra humid environments of Kamchatka¹⁸.

The particle-size distribution was characterized by the predominance of PM_>10. Compared to the non-heated Andosols, the slightly heated Andosols of zone II had a higher content of fine fractions (PM_<10) and a lower content of PM_250-500 (Tables S8, S9). The content of PM_<5 increased with depth, with insignificant differences in the contents of coarser fractions. Properties (pH, SOC and EC) of Andosols from both zones showed similar absolute values and similar vertical distribution.

Slightly heated Andosols had higher contents of Si, K and Sr, lower contents of Ti, V, Fe, Co and Ca and similar contents of Cr, Ni, Cu, Zn, As, Pb, Al, P and Mg as compared to those of Zone I. Most elements showed uniform vertical distribution patterns, with the exception of P, which showed increased values in the topsoil (Fig. 4a).

The topsoil of zone II differed from that of zone I in the predominance of smectite, noticeable amounts (> 10%) of feldspars and quartz (Table 2), lower amounts (< 5%) of kaolinite, pyrite and cristobalite and the absence of considerable vertical differentiation (Table S8, Fig. 3b). Based on SEM data, newly formed hydrothermal pyrite and cristobalite also occurred (up to 5%) and occasionally large intrapedal pores filled with acicular mordenite (> 10%) (Fig. 2j).

Zone III. Moderately heated Eutrosilic Gleyic Aluandic Andosols (Loamic, Reductic, Protosalic, Hyperthionic) under different moss and microzonal communities

Moderately heated Andosols with temperatures of 40–80 °C at a depth of 15 cm (Figure S2) were characterized by the formation of a surface crust (brown or cream colored, humified, with a fine subangular blocky—crumbly structure; Table S2) or, at the margin of this zone, an A-horizon with a structure from fine angular blocky to coarse subangular blocky and inclusions of ferruginized plant remains (Fig. 2h). Fragments of volcanic rocks were also considerably ferruginized, which indicated an intensive inflow of fluids saturated with mobile Fe. The topsoil (0–4 cm) of zone III as compared to that of zone I had similar porosity, lower content of organic debris and collomorphic organomineral microaggregates and higher degrees of ferrugination of groundmass which isometric fragments consist of cracked sand-sized grains of plagioclases, weathered volcanic ash and heavy minerals. The SEM analysis showed that the topsoil had inherited the matrix microstructure similar to that in soils of zone II with apparent pseudomorphs over primary structural elements and a total replacement of initial rocks as a result of hydrothermal alteration as previously reported for thermally altered tuffites in Kamchatka²⁸. The non-oriented clay matrix contained silty peds (Fig. 2k). In the subsoil, volcanogenic source rocks were strongly argillized.

Below the root zone, a relatively homogenous brown Bw-horizon of an altered volcanogenic material with a subangular blocky—angular blocky structure containing less than 20% of PM_>10 was most often present. The Bw-horizon was underlain by a clayey B1-horizon with an angular blocky—prismatic or massive structure and various colors (alternating white, violet, ochric, red, brown and lilac stripes and mottles). On a micro scale, this material showed mostly dark gray hues in plain polarized light and a red, fluidal texture in reflected light (Fig. 2f), indicating the presence of intrusive ferruginous pedofeatures and a high content of anatase and alunite-jarosite. Similar brightly colored B1-horizons have been detected in endothermal soils of Iceland⁵, Kamchatka^18,24 and California³⁷. Therefore, it could be suggested that such horizons, which form under the influence of steam-heated acid sulfate water, occurred worldwide.

The Bl-horizon was underlain by a bleached white (whitish gray background matrix with cream and light blue mottles) clayey parent material (Fig. 1h). Some angular blocky and prismatic peds had red films on their surface and sides.

Compared to the non-heated Andosols, the moderately heated Andosols had higher amounts of fine particle-size fractions and lower content of PM_250–500. The topsoil of zone III was dominated by PM_>10 and the subsoil (> 10 cm) by PM_<5 (Tables S8, S9). In the subsoil, PM_250–1000 were absent, the content of PM_10–50 was lower and the content of PM_<5 higher than in the topsoil, while the content of PM_5–10 and PM_50–250 was similar.

In comparison to non-heated Andosols, moderately heated Andosols had higher Eh and EC and lower pH and SOC values due to the biogenic oxidation of sulfides with the production of sulfuric acid, which was confirmed by the presence of colonies of iron-oxidizing bacteria (Fig. 2f). SOC, pH, EC and Eh changed insignificantly with depth.

The elemental composition of moderately heated Andosols differed from that of non-heated Andosols by a higher content of Fe, As, Pb and Al and a lower content of Mn, Ca, K, Mg and Sr, with similar content levels of V, Cr, Co, Ni, Cu, Zn, Si and P (Tables S8, S10). The topsoil of zone III had higher contents of Ca, K, Mg, Mn and Si and the subsoil was enriched in Al, As, Cu, Fe, Ni, Pb, Ti, V and Zn. The content of P, Sr, V, Cr, Co and Ca varied insignificantly with depth.

Compared to non-heated Andosols, moderately heated Andosols had a higher content of kaolinite (the predominant phase in matrix and peds) and anatase, a lower content of smectite, potassium feldspars and plagioclases and a similar content of pyrite, zeolites, quartz, cristobalite, hematite and goethite (Table S8, Fig. 4b) which coincides with data³⁸ on the disappearance of smectite in hydrothermal altered rocks with a temperature exceeding 85–95 °C at the Wairakei geothermal field, New Zealand. The topsoil was enriched in potassium feldspars and albite and impoverished in zeolite, anatase, quartz and cristobalite (SEM data showed that the last three minerals occurred within large intrapedal pores inherited from soils), with similar amounts of kaolinite, smectite and pyrite.

Judging from the XRD full patterns of kaolinite, it had a disordered structure (Fig. 3d–f). It should be mentioned that the calculation of different indices (Hinckley, AGFI) using several peaks of the XRD pattern was a very popular method for assessing the degree of kaolinite order/disorder. However, recent studies³⁹ showed that only the analysis of the XRD full pattern was suitable for this. All the studied kaolinites were characterized as low-ordered, since they had poorly defined non-basal (h,k) peaks and the modulation of peaks at large angles 2θ was almost absent. The low ordering of the studied kaolinite structure was apparently caused by the following reasons: (1) low stacking order of the layers due to certain geological conditions during kaolinite synthesis, (2) occurrence of a small amount (up to 5–10%) of mixed-layered kaolinite-smectite (K/S) minerals. This was also indicated by the presence of a specific displacement of basal (00l) peaks after oriented specimen saturation with ethylene glycol (Fig. 3e,f). When calculating the mineral composition, the phases of kaolinite and K/S did not differ.

According to the SEM study of the moderately heated Andosols, the kaolinite had a diverse particle morphology, which obviously showed the degree of its structure order/disorder. Pseudohexagonal crystals of kaolinite (Fig. 5b) were found within a non-oriented matrix (Fig. 5a). In contrast, crystals with poorly defined habitus or with irregular sheet-like curved shape (Fig. 5d) were found in microaggregates oriented along the walls of open microcracks (Fig. 3c). Thus, based on SEM and XRD, we suggested that these were particles of disordered kaolinites and kaolinite-smectite mixed-layer minerals. Such elongated aggregates were found in the dense groundmass of the layer (Fig. 2k).

Zone IV. Hot Gleyic Aluandic Andosols (Clayic, Reductic, Salic, Hyperthionic) near the center of the steam hydrotherm

Shallow and hot Andosols with temperatures of > 80 °C (Figure S2) exhibited a mottled (light gray and reddish) biocrust of thermophilic algae on the surface of a bright-colored clayey Bilz-horizon (Table S2), which was underlain by greenish, bluish and whitish clays (Crz-horizon) formed as a result of a deep hydrothermal alteration and bleaching of volcanic rocks. On a micro-scale, the zone IV topsoil consisted of laminated crusts of silty material impregnated and bonded by collomorphic Fe–Al-organomineral substances with a fluidal (after⁴⁰) texture. Below, the argillized material was very heterogeneous in color (e.g., with whitish mottles corresponding to locally reducing conditions) and contained inclusions of various minerals (Figs. 1i, 2g,h). Such argillized material was typical for the shallow hydrothermal sediments of Kamchatka^28,41,42,43 and the Iturup Island²⁷. SEM analyses showed that hot Andosols consisted of kaolinite aggregates, pseudomorphs of which had replaced rock-forming minerals. The soil microfabric was typical for Kamchatka’s hydrothermal clays (Fig. 2l), i.e., inherited from the parent materials²⁵, with a high porosity resulting from an intensive rise of vapors from endogenous hot fluids. Such pores were colonized by thermophilic bacteria.

The particle-size distribution of hot Andosols was characterized by the predominance of PM_<5 and PM_10–50 (Table S8). In comparison with non-heated Andosols, hot Andosols had a lower content of PM_>10, lower pH and SOC values and higher values of Eh and EC due to an intensive inflow of endogenous fluids and the biogenic oxidation of sulfides (Table S7).

Hot Andosols as compared to non-heated Andosols were also enriched in Ti, Cr, Fe, Co, Cu, As and Pb, depleted in Mn, Ca, Si, K and Mg and had the similar contents of V, Ni, Zn, P and Sr (Table S7). Leaching of Ca, K, Na and Mg from volcanic rocks was known to be induced by steam-heated acid-sulfate waters⁴⁴, while enrichment in Fe, Cu and Co could be caused by a selective biogenic accumulation of metals (including Ca, Mn and Sr⁴⁵) within biocrusts. The impact of steam-heated acid-sulfate waters caused an enrichment of Ti within deep layers of subsurface sediments in Kamchatka⁴⁶ and the formation of mineral coatings on stones in a hot river on the Iturup Island²⁷.

The volcanogenic source rocks of the hot Andosols were deeply altered and mostly consisted of poorly-ordered kaolinite and kaolinite-smectite mixed layered minerals (Fig. 3e,f) and small amounts of boehmite, anatase and quartz (Table S8, Fig. 2l). Most samples of hot Andosols also had considerable contents of potassium feldspars, boehmite, pyrite, cristobalite and hematite. Plagioclases, goethite and zeolites occurred rarely. The smectite content was low, this being the main mineral of soils in non-heated and slightly heated Andosols. Boehmite, jarosite, pyrolusite, marcasite and calcite occurred only in hot Andosols (Table 2). In comparison with non-heated Andosols, hot Andosols were enriched in kaolinite, anatase, hematite and boehmite and impoverished in smectite, potassium feldspars, plagioclases and zeolites (Table S11). A high content of kaolinite and a low content of albite and quartz resulted from the duration of exposure to and/or high temperature of a steam hydrotherm^47,48.

The microstructure of hot Andosols (Fig. 2l) differed from that of less heated Andosols by the presence of numerous round opal grains, which often formed collomorphic microaggregates, pore infillings and bridges between clay micropeds. At the surface of opal pedofeatures, there were occasional crystals of pyrite, anatase and cristobalite.

Botanical transect description

Within the studied catena, from the periphery of the thermal field with communities (association of Betuletum ermanii calamagrostidosum subassociations typicum or dryopteridosum) on non-heated Andosols to its bare ground center, plant communities replaced each other in the following order: slightly modified tall herb meadow → modified meadow with dominance of Thalictrum minus or Pteridium aquilinum → communities with dominance of Artemisia opulenta and monodominant communities of Calamagrostis langsdorfii → microzonal communities⁸ with dominance of Agrostis geminata, Lycopus uniflorus or Potentilla stolonifera → microzonal communities with dominance of the Marchantiophyta (Gymnocolea inflate and Solenostoma vulcanicola)^49,50 and Fimbristylis ochotensis → Marchantiophyta communities → biocrust.

The cover of vascular plants dropped at around 60 °C, with cover values not exceeding 20% above this temperature. Mosses decreased in cover at 65 °C and disappeared completely at around 90 °C. The total number of species varied from 9 to 24 per 1 m² on the non-heated soils. With increase in temperature it decreased and did not reach more than 6 species per 1 m² at a topsoil temperature of more than 43 °C. Thus, together with previous studies^51,52,53,54, our research showed that in the Valley of the Geysers, non-vascular plants dominated at the moderately heated and hot soils.

The clear majority (263 of 292, see Methods) of species belonged to Magnoliophyta. Both facultative (Net Relatedness Index, NRI = 1.7 ± 0.5, p = 0.036) and obligate (NRI = 1.3 ± 0.8, p = 0.084) thermal species seemed to be clustered on the phylogenetic tree of Magnoliophyta from the Valley of Geysers (Table S12). However, they did not form a monophyletic clade and were organized in small clusters⁵⁵ of closely related thermal species that were distributed throughout the phylogenetic tree (Fig. 4c).

Only 9 of 60 families of vascular plants contained obligate thermophyte species (Fig. 4c, Figure S3). Rosaceae (4 of 18 species) were overrepresented with a false discovery rate (FDR) level of 15%. The other two families represented by more than one species were Cyperaceae and Poaceae that had been typical for Cenozoic floras of geothermally influenced wetland hydrophytic and dryland mesophytic communities⁵⁶. As far as orders were concerned, only Rosales were overrepresented in the obligate thermal flora (Table S13).

Among 24 families that contained facultative thermophytes, no families were significantly enriched or depleted in such species. At the level of orders, Poales were underrepresented and Asparageles were overrepresented among facultative thermophytes (Table S13).

Among 6 families containing only thermal species (both facultative and obligate thermophytes) in the Valley of the Geysers, 5 families (Dennstaedtiaceae, Geraniaceae, Urticaceae, Rubiaceae, and Iridaceae) were represented by only one species, and Amaryllidaceae were represented by two species Allium ochotense and A. strictum. Although this was not surprising having a probability of 0.03 to occur by chance, the fact that both these species were thermal may relate to the fact that the main center of Allium genera diversity located in southwest and central Asia⁵⁷ i.e., Allium sp. might serve as an indicator of thermal habitats in Kamchatka.

At the non-heated habitats near the catena, among families of the Valley of Geysers with only non-thermal species, Thelypteridaceae, Equisetaceae and Betulaceae were overrepresented (Table S13). Overrepresented orders were Equisetales, Polypodiales, Fagales and Liliales, while the order Poales was underrepresented. At the class level, the nearest non-thermal flora at non-heated Andosols (group B, see Methods) was enriched with Polypodiopsida species, which could result from the higher humidity of this territory¹⁸.

Alteration of Andosols under the impact of steam-heated acid-sulfate waters

In the Valley of Geysers, the inflow of hydrothermal fluids into Andosols seemed to cause a decrease in TOC content, salinization (Figure S4), alkalinization (only at the hydrotherm margin), accumulation of Al, As, Co, Cr, Cu, Fe, Pb, Ti and V, removal of Ca, K, Mg, Mn and Si resulted from acid hydrolysis and a deep mineralogical alteration of the parent materials. The biogenic oxidation of sulfides^{15,22,23,43,58} provided forming of potential acid sulfate soils. A decrease in the TOC content resulted from both a small amount of litter—the projective cover of vegetation decreases from the periphery to the center of the hydrotherm—and sulfuric acid decomposition of organic matter. These processes were typical for endothermal soils affected by the steam-heated acid-sulfate waters in Kamchatka¹⁸, New Zealand^{9,59,60,61,62}, Iceland²², California³⁷, Japan (Hokkaido⁶³ and the Kusatsu-Shirane volcano region⁶⁴).

The hydrothermal alteration in endothermal soils was characterized by the transformation of smectite (the predominant clay mineral) to kaolinite, the assemblage of non-clay minerals (albite, anatase, anorthite, clinoptilolite, cristobalite, heulandite, microcline, quartz), a decrease in the proportion of plagioclases (with a relative increase in microcline) and the appearance of noticeable amounts of jarosite and boehmite (Fig. 4c). New mineral phases and an increased proportion of clay in the particle-size distribution were previously reported in endothermal soils of Kamchatka^18,24, Japan⁶⁴, California³⁷, and Iceland⁵.

The A-horizon of the non-heated Andosols had a weakly developed macrostructure and a good microstructure. Due to the more intense alteration of rocks and minerals and a heavier texture, in slightly heated Andosols, the structure was well-developed both at a macro and micro scale. In zones III and IV, the A-horizon degraded to a biocrust. A similar trend was observed in the B-horizon, i.e., the structure was most developed at the hydrotherm’s periphery because of the favorable combination of factors such as a loamic texture and the release of bases from weathered primary minerals.

In the topsoil (0–10 cm), the SiO₂ content in zones I–III was typical for non-heated Andosols of Kamchatka⁶⁵, but higher than those in zone IV, where Al, Fe and Ti were introduced by hydrothermal fluids and contributed to the formation of disordered kaolinite, anatase, hematite, goethite and boehmite.

Soils with temperatures below and above 50 °C differed in their mineralogical composition. Soils of zones III and IV had lower contents of plagioclases, mordenite, potassium feldspars and smectite and higher contents of anatase, boehmite, kaolinite and hematite (Table 2). The studied endothermal soils in relation to non-heated Andosols had higher contents of clay minerals and Fe- and S-containing minerals and lower contents of other non-clay minerals: albite, anorthite, clinoptilolite, microcline (Fig. 4c).

Clustering of thermal species on the phylogenic tree

In the Valley of Geysers, the Poaceae (30 species) and Cyperaceae (19) were the most represented families after the Asteraceae (31). But only one Poacea species and no Cyperacea species were among 33 facultative thermophytes, which makes Poales the most underrepresented order among facultative thermophytes (Table S13). Given that this order was the most prevalent among obligatory thermophytes (Table S14), we suggest that there are different constraints of thermophiticity and thermotolerance⁶⁶. And for Poales, specialization was the more preferred way of adaptation to the extreme environment (high acidity, salinity, temperature and clay texture) rather than generalization. According to⁶⁷, Poales and Asterales were over‐represented by alien plants on Mediterranean islands due to their late flowering, large seed size and anemochory. The prevalence of the Poales among obligate thermophytes could resulted from their ecological competitiveness⁶⁸. For example, the Poaceae and Cyperaceae are well adapted to drought⁶⁹. Additionally, the Poaceae are well adapted to saline habitats⁷⁰. Data on the fossil floras of ancient hot carbonated springs support the fact that vegetation pre-adapted to chemically and physically stressed environments colonizes heated habitats⁵⁶ i.e., plants occupy free niches.

In the geobotany analysis described above, all the species that occurred in the Valley of Geysers (in areas with either temperature, but had not been found at the catena and at the surrounding territory) were merged in the group of distant territory (both non-heated and heated). It was justified by several reasons. Firstly, the temperature, humidity, salinity and altitude differed throughout the Valley of Geysers^1,18. And our analysis aimed to compare vegetation of the specific thermal site affected by steam-heated acid-sulfate waters with its surrounding and distant territories represented by all other parts of the Valley of Geysers. Secondly, it appeared to be impossible to clarify which species had been identified as “thermophytes” without a detailed soil temperature survey in⁷¹ later provided in^1,8. When we specified species with the status “thermal” in either distant territories (the Geysernaya River valley’s plant species list compiled by L.I. Rassokhina: see Table S14) or in the catena studied as “thermal” and species with “non-thermal” status in either study as “non-thermal”, the results were overlapping, but not the same (Table S15). The Cyperaceae was the overrepresented family among obligate thermophytes, but underrepresented among facultative thermophytes. The Ericaceae and Saxifragaceae tended to be non-thermal (simultaneously overrepresented among non-thermal species and underrepresented among thermal species), while the Caryophyllaceae and Juncaceae were the thermal species (vice versa). 10 of the 12 Ericaceae and all 4 of the 4 Saxifragaceae species were non-thermal; 9 of the 10 Caryophyllaceae and 8 of the 9 Juncaceae species were thermal. We suggested that the differences between the results obtained by the two calculation methods were caused by the diversity of thermal habitats in terms of Eh, pH, cation–anion composition and salinity of waters, heating and moistening of soils^1,18,24,48.

Relationships between the composition of plant communities and soil properties

The depth of plant root occurrence decreased from non-heated Andosols to heated Andosols of zones II and III (Table S2). Plant roots were not found in hot Andosols. In non-heated Andosols, slightly and moderately heated Andosols, the diameter of plant roots decreased, too. We did not notice any features (e.g., horst- or cushion-forming) of underground plant parts or roots that were preserved alive where there was direct contact with heated ground.

In the catena studied, the most drastic changes in soil morphology at the meso-, micro and nano-scales, elemental composition and particle-size distribution occurred when the temperature at a depth of 50 cm rose above 60 °C, with the most significant mineralogical changes occurring with the temperature transition at 50 °C. The species diversity of the facultative thermophytes changed sharply when the soil temperature at a depth of 50 cm rose above 98 °C, although that of obligate thermophytes changed only slightly (Figure S5). The pairwise comparison of species compositions from the thermal zones showed that zones II–IV had more similar species than expected by chance and zone I differed significantly from zone II due to the loss of non-thermal species in zones II, III and IV and their high diversity in non-heated areas. At the catena studied and its surrounding territory (both non-thermal and thermal landscapes), the number of obligatory thermophytes gradually decreased with an increase in temperature, while the number of facultative thermophytes remained the same in zones II and III and sharply decreased in zone IV that could have resulted from the absence of any visible morphological adaptations, for example (Figure S5).

The absolute majority of the thermophilic vascular plants were mesophytes and not psychrophytes in relation to temperature and humidity, respectively (Table S16). In comparison to non-thermal flora, there were no overrepresented or underrepresented biomorphs (Fisher’s exact test, p value > 0.05).

Based on the relationships between the cover values of plants and edaphic conditions at the catena we distinguished 11 ecological groups of species (Table S17–S19). The main edaphic factors limiting plant growth included EC, the content of sulfates in water extract, particle-size distribution and pH. Even for the thermal zones, such factors as the soil temperature, SOC content, concentration of toxic compounds and Eh were less important.

In thermal landscapes, biodiversity can increase or decrease depending on the maximal temperature at the certain hotspot. When a temperature of heated neutral solutions is low (less than 30 °C) e.g., in the Bolshezemelskaya tundra (Russian Subarctic)⁷², due to higher frequency of soil freezing–thawing, environmental harshness may result in decrease in the floristic and microbiota richness. Such a partition was observed at the Hengill valley in Iceland⁷³. In the South Sandwich Islands (maritime Antarctic), bryoflora richness was maximal at the thermal ecosystems where the topsoil surface temperature reached 47 °C^74,75. In the forest landscapes of Kamchatka⁷⁶, Hokkaido^48,77, California³⁷ and New Zealand^7,9,61,62,78, the inflow of heated thermal waters resulted in disappearance of woody plants. Vegetation was sparse at the endothermal soils with surface temperatures of more than 50 °C. In thermal ecosystems, the highest vascular plant⁷⁹ and moss⁸⁰ species richness was typical for areas where the topsoil temperature is 30–40 and 30–50 °C, respectively due to the upper thermal limit for cell activity lied between 45 and 55 °C.

The subsoil temperature seems to be the main factor controlling the vegetation of geothermal areas, while soil chemical factors are thought to have less influence^{1,8,81,82,83,84}. Because soil temperature is the simplest and most reliable indicator to assess the plant partition at thermal ecosystems during the field study, other environmental factors (e.g., pH, soil moisture, a content of phytotoxic substances) were studied insufficiently. But soil temperature affects the vegetation together with other physical and chemical soil properties that also depend on the intensity of the inflow of heated fluids and waters. A relationship between the vegetation and the environment has only been analyzed based on the data collected at transects from a center of a hydrotherm to its periphery. However, such approach does not allow separating the influence of soil temperature and chemical properties on vegetation as these factors are dependent variables. To understand the role of soil temperature and chemical properties on plant growth separately, more detailed sampling should be carried out into one thermal zone.

Source: Ecology - nature.com