Carl Troll

I Introduction

Carl Troll (1899–1975; Figure 1) was an outstanding geographer of the 20th century who enjoyed great international reputation in the scientific community (e.g. Dickinson, 1969; Fischer et al., 1969). A glacier on Ellesmere Island, for example, was given his name (Carl Troll glacier) (Barsch, 1983). He was president of the International Geographical Union (1960–1964) and chairman of the International Geographical Congress in 1964.

OPEN IN VIEWER

Figure 1.

Carl Troll. Photo courtesy Carl Troll 18 October 1973.

Troll published 361 papers (list of publications 1922–1959, in Lautensach, 1959; 1960–1979 in Lauer, 1970: 1970–1975 in Lauer, 1976) on various fields of geography, including plant geography, geomorphology, landscape ecology, climatology, cartography, air-photo interpretation, photogrammetry and also human geography (e.g. agricultural and colonial geography). Troll held a central position in international geographical high-mountain research. In his papers on high-mountain areas in Europe, South America, East Africa, South Africa and in the Himalayas he addressed physical, biological and ecological topics as well as the interactions between humans and the mountain environment. Forty years have passed since Troll died, so one might ask which of his many publications have had a lasting effect on geographical research. I have focused on some articles that, in my mind, have essentially contributed to progress in physical geography.

II Troll’s approach and contribution to physical geography

Troll’s scientific roots were in physical geography and biology. His way of thinking was deeply influenced by Alexander von Humboldt. Like Humboldt, Troll held a very complex many-facetted view of nature and human response to environmental challenge. Inspired in particular by his extensive personal field experience in many of the earth’s high-mountains, Troll developed his idea of the three-dimensional character of climate (i.e. the change of climate by latitude, longitude and elevation) and its influence on the distribution pattern of vegetation zones, altitudinal belts, life forms and also on certain geomorphic features (see below). This concept is reflected in a series of remarkable publications (e.g. Troll, 1942, 1943b, 1948b, 1955, 1956b, 1958a, 1959, 1961, 1962, 1964, 1978a) that may be considered ‘milestones’ in Troll’s scientific work. Among them is the map of the seasonal climates of the earth (Troll, 1964). This map is based on the relationships between the seasonal and diurnal course of temperature and on the precipitation regimes (distribution of humid periods and seasons). In the accompanying article the thermal regimes are clearly explained by 15 characteristic thermoisopleths. Undoubtedly this was one of the best effective classifications of climates (Lauer, 1970) and certainly one of Troll’s most outstanding publications. Troll’s concept was broadened in particular by the numerous studies of his post-graduate student W. Lauer (1952 onwards; for further references see Lauer and Rafiqpoor, 2002) that resulted in a new map of the Earth’s climates. This map is based on the ecophysiological characteristics of the real vegetation (Lauer and Rafigpoor, 2002) and also shows, for the first time, the three-dimensional arrangement of the altitudinal belts, thus completing Troll’s original concept.

Troll, who took part in the fateful German Himalaya expedition of 1937 as leader of the group of scientists, also produced a very detailed and technically perfect vegetation map (1: 50,000) of the Nanga Parbat area in the northwestern Himalaya (Troll, 1939b). This map shows the three-dimensional distribution of vegetation (20 vegetation units) as related to the most outstanding site characteristics (e.g. slope exposure, topography and its effects on snow cover, soil moisture, etc.) and gives a more detailed view of the vegetation pattern than could be deduced from the Landsat-TH-image (149/35) about 30 years later (Nüsser and Schickhoff, 1996). Without any qualification, this map must be considered a highlight in high-mountain plant geography and cartography and should have received more international appreciation (e.g. Küchler, 1967).

It was mainly Troll’s comparative approach to high-mountain geography that made his studies so valuable and unique (e.g. Troll 1941; 1956a; 1968b). By their studies in the world’s high mountains, many of his numerous post-graduate students have contributed to spreading Troll’s idea on common traits and variation in high-mountain nature as influenced by zonal climates (diurnal, seasonal), climate character and floral development (e.g. Bader, 1960; Böhm, 1965; Ern, 1966; Fränzle, 1959; Hastenrath, 1960; Holtmeier, 1965, 1974, 2009; Schweinfurth, 1957, 1966).

In his articles on high-mountain landscapes Troll regularly addressed timberline (treeline) position and physiognomy, as timberline is the most conspicuous and ecologically very important vegetation boundary in most high-mountain landscapes (e.g. Holtmeier, 1974, 2009; Troll, 1973). On a common field trip to the treeline in the Canadian Rocky Mountains (1972), Carl Troll, told the present author that he was planning a comprehensive monograph on timberline in the global view. However, Troll died in 1975 before he could realize this project.

Linking biological and geographical thinking, Troll formulated the concept of landscape ecology (Troll, 1939a). Landscape ecology is concerned with the internal structure and function of landscape units (ecosystems) and their spatial and functional relationships. A riparian landscape would be a good example. It is characterized by a stream, an episodically or periodically flooded floodplain (i.e. back waters, sand banks), the adjacent terrestrial habitats (biotopes) and fluxes of anorganic and organic matter (solid material and soluble substances) between these ‘units’. Three years before, in an extensive and detailed article on ‘termite savannas’ (Troll, 1936), Troll had already presented instructive examples that clearly reflected his ‘landscape ecological approach’.

Landscape ecology became rapidly established in geography, as the interactions of humans and nature in space and time in their entirety were the objective in traditional geography. Troll wanted to counteract the increasing specialization and splitting up of sciences that were also apparent in geography, and come back to an holistic approach again (1968a). Thirty years after introducing landscape ecology, Troll suggested replacing the term ‘landscape ecology’ by ‘geoecology’, which was unburdened by historical linguistics concerning the term ‘Landschaft’, of which the meaning varies in languages, and independent from the aspects where landscape is considered (geography, ecology, history, archeology, agricultural sciences, painting, poetry, etc.) (Troll, 1970). Thus, the distinction often made by many contemporary geographers between ‘landscape ecology’ and ‘geoecology’ (e.g. Blumenstein et al., 2000) and considering them to be separate subjects lacks any historical justification, at least. However, if geoecology is considered a particular subject with objectives going beyond those of landscape ecology – and there might be more or less plausible reasons to do so – one should choose another term, as ‘geoecology’ was introduced to geography just as a surrogate of landscape ecology (Troll, 1970). Internationally, however, the term ‘landscape ecology’ has become well accepted (e.g. Forman, 1995; Leser, 1997).

Troll also realized and emphasized the increasing possibilities of using aerial photography in physical and cultural geography, and particularly in landscape ecology (review and references in Troll, 1939a, 1942/43). Air photos were already successfully used at that time in geology, geomorphology, deposit reconnaissance, forestry and archeology, for example. Troll considered vertical air photos combined with ground truths to be an ideal instrument in landscape ecology, as the air photos provide a synopsis of the landscape pattern and allow a complex first approach to the spatial structures and functional relationships of the landscapes units.

Thus, to most geographers, Troll’s name is linked mainly with plant geography and landscape ecology (geoecology). He was, however, also an expert in periglacial geomorphology. In this context, his early studies on the Inn-Chiemsee glacier and its effects on the landscape pattern (character and distribution pattern of landforms, soils, moisture and vegetation) in the northern foreland of the European Alps must be mentioned first (Troll, 1923, 1924, 1926). Geomorphology, and periglacial geomorphology in particular, was Carl Troll’s favourite field of research, even more than plant geography and landscape ecology. His first ideas of ‘landscape ecology’ were already born during his geomorphological studies in the Alpine Foreland (Troll, personal communication), i. e. before he introduced the term into landscape research (Troll, 1936, 1939a).

Most of Troll’s numerous articles were written in the German language and thus were not very popular (e.g. Ågren and Andersson, 2012; Sanderson and Harris, 2000) in the Anglophone scientific community. This holds true also for the extensive (149 pp.) and profound article on ‘Patterned ground, solifluction and frost climates of the earth’ (Strukturböden, Solifluktion und Frostklimate der Erde) (Troll 1944a; Figure 2). This article was published in 1944 by the Geologische Vereinigung in a special issue of the Geologische Rundschau, dealing with ‘Diluvial-Geology and Climate’. It is relatively seldom referred to and is missing in many textbooks of physical geography and high-mountain geography (e.g. Bonan, 2002; Briggs and Smithtson, 1993; Goudie, 2001; Price, 1981; Strahler and Strahler, 1992). Therefore, Troll’s article on ‘patterned ground’ is addressed here in detail.

OPEN IN VIEWER

Figure 2.

Carl Troll visiting patterned ground on Plateau Mountain (2.519 m, Highwood Range, Alberta) at the occasion of the second meeting of the International Geographical Union Commission on High Altitude Geoecology, Calgary, 1972. Photo by the author, August 1972.

Carl Troll himself was the editor of ‘Diluvial-Geology and Climate’. The volume consists of 13 papers. These address climatological, geological and geomorphological aspects of the Ice Age as well as of pre-glacial and post-glacial times, and in one article the late-glacial history of climate and vegetation of Finland is considered. The introduction (22 pp.) by the editor is entitled ‘The face of the Earth’. It outlines the scope of this special issue of the Geologische Rundschau and provides detailed information on the historical background of research, in particular by German researchers, up to the 1940s. In addition, main characteristics of patterned-ground forms are considered and terminology and problems of their genetic differentiation in relation to different climates are addressed.

In the following I will address Troll’s article ‘Patterned ground, solifluction and frost climates of the earth’ only. At 149 pages, it can be considered to be an instructive monograph rather than a ‘normal’ journal article. However, there are also a few shorter papers by Carl Troll dealing with patterned ground and related subjects that were published either prior to or following the article in Geologische Rundschau (Troll, 1943a, 1943b, 1944b, 1948a, 1961, 1971). In the present author’s opinion, the article (1944a) under consideration is one of the most important publications by Carl Troll. The article is well-illustrated by photos and diagrams. As in his other papers, Troll systematically used thermoisopleths to characterize seasonal and diurnal types of climate, a method that was improved upon later by Wace (1990). The 1944 article provided an essential contribution to the understanding of the nature of cold climate areas, no matter whether they are situated at high altitudes or at high latitudes, or in areas that had been previously influenced by such processes during earlier periods of colder climate (Ives, 1976).

It is still amazing that Troll could keep up with the widely distributed specific international publications even though access to international literature was difficult during World War II and the possibilities of worldwide travel were also limited. Thus, as he said somewhat ironically, Troll found ‘plenty of time’ to focus on completing this many-facetted synthesis that represents the state of research on frost-patterned ground at that time.

Although more reviews on periglacial geomorphology by other authors followed later (e.g. Cailleux and Taylor, 1954; French, 2008; Hall, 2002; Tricart, 1963; Washburn, 1950, 1956, 1979), it still is worth having a look at Troll’s profound and stimulating work. The objective was to classify frost-affected soils according to the climatic regimes (seasonal, diurnal climate, temperature, precipitation, moisture) controlling the regions where the many forms of patterned ground and associated vegetation (if present) have developed. In one way or the other this approach is similar to that of the distribution pattern of the life forms of plants in relation to the environmental conditions in the different climate zones of the earth (e.g. Troll, 1958a, 1960, 1961, 1978b).

In the introduction Troll gives a detailed review of older literature about occurrences of frost-sorted material and patterned ground reported by researchers on expeditions to high latitudes in northern Europe, Arctic islands and continental Siberia since the early 20th century. In the following, terminology and hypotheses on the multiple causes and the great variety of frost-patterned ground are discussed, including polygonal forms, sorted and non-sorted stone stripes, solifluction lobes and turf-banked terraces as well as stone garlands, stone pavements and other patterns. Problems of genetic differentiation are highlighted and the role of fossil frost-sorted forms as indicators of palaeoclimatic conditions is considered. Troll showed that frost-patterned ground may occur in all climatic zones where seasonal or daily nocturnal freezing of the upper top soil regularly occur. Patterned grounds were found to be best developed in the polar region and in tropical high mountains, while they appeared to be less frequent and less spectacular in the temperate mountains.

In the next chapter (II) Troll goes deeper into detail. The distribution of frost-structured and frost-textured soils is considered with special regard to lower limits in the world’s high mountains. This aspect played a major role during the following decades and led to many controversial debates among geomorphologists.

Chapter III deals with the climatic preconditions for the development of frost-patterned soils. The latter usually occur above the Pleistocene snow limit or close to it and thus may be considered to be a characteristic feature of the subnival (periglacial) belt. It is not the presence of permafrost that causes frost-patterned ground, but the periodical (seasonal or diurnal) occurrence of morphologically effective soil frost and the frequency of freeze-thaw events that are the main factors.

In chapter IV the main hypotheses on the development of frost-patterned soils are discussed. Convection, cryoturbation, frost heave and frost-thrust are considered in all detail as well as the role of free and tied (hampered) solifluction, translocation, regelation, swelling, sorting of soil material, the role of fine soil and the structure of soil ice and needle ice formation.

A detailed review (17 pp.) on occurrence, formation and effects of needle ice in the development of frost-patterned soils from the polar zones to the tropical high mountains, with a focus on the history of needle-ice research, is given in a special chapter (V). The role of needle ice as a geomorphic factor is considered not only at the micro scale, but also as an agent causing or enforcing slope erosion over larger areas at high elevation in tropical high-mountains, where needle-ice formation may occur every night (diurnal temperature regime).

Moreover, the author considers the formation of patterned ground by factors other than climate (Chapter VI) and outlines extra-zonal occurrences of structured soil patterns and the causes of formation (Chapter VII).

The following four chapters (VIII–XI), that comprise more than 50% of the monograph, focus on the regional differentiation of frost-patterned ground formation, including tropical high mountains with special regard to the South American High Andes, Mt Kilimanjaro, Mt Kenya, the High Arctic (Svalbard, Greenland) and sub-polar regions including Eurasian and North America tundra, as well as ocean islands (Faroes, Iceland) and the sub-Antarctic. Moreover, the boreal zone of Eurasia, the low mountains of Europe and the New England States, and the Alps and Pyrenees are focused upon. Not least, occurrences of frost-patterned soil in subtropical mountains of the Middle East (Taurus, Elburz) and in the Caucasus, Pamir, Himalayas, continental East Asia, Japanese Alps, South Africa and South America are considered.

Until the 1940s there was no other comparable similarly detailed overview of the regional distribution and types of frost-patterned soils, the many factors influencing their development and on other forms of patterned ground that are controlled by factors other than frost action. Fourteen years after the article had been published it was translated by the US Army into the English language (Troll, 1958b). Troll’s article goes far beyond a simple review of the state of the art. It is the particular systematic comparison of the regional (geographical) distribution, forms and formation of patterned ground and related phenomena that can be considered to be the most important merit of this article.

The detailed elaboration of the relative importance of the many other factors involved in the formation of patterned ground (such as substrate, texture, moisture, water-holding capacity, capillary water, needle-ice formation, etc.), and of terms and concepts, provides a well-differentiated idea of the periglacial (subnival) environment, with climate as the main controlling factor. Troll’s patterned-ground article and related publications reflect the same reasoning as his studies on the three-dimensional character of climate and its influence on the distribution pattern of vegetation zones, altitudinal belts and life forms (see also his map on the earth’s seasonal climates as discussed above).

Although going deep into details, the author always considered them in the entire context and opened new far-sighted perspectives for further studies. The careful substantial discussion of the existing theories from the beginning of frost-patterned ground research until the mid-1940s made the article a masterly contribution to the knowledge of patterned ground forming processes and the great variety of frost-patterned forms at that time.

Troll’s 1944 article and related publications (Troll, 1943, 1944a, 1944b, 1948a, 1961, 1971) have been a guide in particular for German speaking geomorphologists, probably because the articles were not written in the English language, with the exception of the translation by the US Army (Troll, 1958b).

Studies on patterned ground and related phenomena have considerably increased since then. Numerous papers cover the Alps and other high mountains in both hemispheres as well as in sub-Antarctic and Arctic areas (e.g. Furrer, 1954, 1969; Graf, 1981; Grötzbach, 1965; Hall, 2002; Hall and Boelhouwers, 2003; Heine, 1977; Höllermann, 1967, 1982; Lawler, 1988; Pissard, 1977; Rafiqpoor, 1994; Schunke, 1975; Schweinfurth, 1964; Spreitzer, 1960; Stäblein, 1977; Stingl, 1969; Washburn, 1973; further references e.g. in Washburn, 1979, and French, 2008), and many regional gaps in the knowledge of patterned ground occurrences have been filled. Moreover, long-term field experiments in selected locations have also considerably increased. Consequently, some statements made by Troll needed some (minor) revisions and became more differentiated.

Here, only a few examples are given. Thus, the lower limits of patterned ground and solifluction do not necessarily run parallel to the upper timberline and snowline, but show just the same trend (cf. Troll, 1971). Moreover, the lower limit of patterned ground turned out to be not very appropriate as an indicator of climatic-geomorphic conditions (Höllermann, 1977), because the occurrences of patterned grounds within their potential distribution area also depend to a varying degree on non-climatic factors or on only indirectly climate-controlled factors. Their effects may overlap more or less (see also Graf, 1981).

The hypothesis that, on permafrost, patterned grounds and free solifluction may occur everywhere did not hold true for many permafrost regions. In addition, Troll’s statements on the geographical distribution of polygons by different size (large polygons, miniature polygons) have needed to be more differentiated (see also Troll, 1947) as small, patterned forms have also been found in many sub-tropical high mountains (e.g. Middle East), in highly maritime sub-antarctic islands (Kerguelen archipelago, Southern Shetland Islands, South Orkney Islands, Crozet Islands, Marion and Prince Edward Islands), in New Zealand (South Island) and Stewart Island as well as in close proximity to the coastal areas of Iceland and on the Faroe Islands. Troll himself already mentioned some of these observations in a paper on solifluction that followed the ‘Patterned ground’ article (Troll, 1947; see also Troll, 1971).

Some minor correction had to be made concerning needle-ice formation and effects. Thus, the orientation of so-called ‘windgestreifte Auffrierböden’ (i.e. wind-oriented non-sorted stripes resultant from needle-ice formation) does not necessarily coincide with the prevailing wind direction (for review see e.g. Washburn, 1973; Höllermann, 1982). Moreover, formation of multilayered needle-ice, that Troll (1944a) and some other authors (e.g. Furrer, 1954; Soons, 1967) supposed to be caused by the number of freeze-thaw events, appears to be more complex (Soons and Greenland, 1970). Altogether, however, most of the causal explanations are still valid or did not principally change.

Nowadays, people searching for information on periglacial processes and their geomorphic effects (and landforms) usually reach for the reviews by Price (1972) and the textbooks of Washburn (1979) and French (2008) rather than for Troll’s article. These reviews, representing the current state of the knowledge, are almost universally accepted. Anyway, many hypotheses about patterned ground and periglacial processes are still tentative, and challenging open questions are left for continued research, particularly in tropical high mountains and high southern latitudes. However, many extant studies on periglacial processes, landforms and patterned grounds of the sub-Antarctic region (for review see Hall, 2002; Hall and Boelhouwers, 2003) are little known by many northern hemisphere geomorphologists (Hall, 2002). In view of the enormous increase of publications, available data und observations on patterned ground and related phenomena, it becomes obvious that formation of patterned ground and the geomorphic role of solifluction can be completely understood only from a synoptic landscape(-ecological) approach including climate, substrate, soils and vegetation, as was already shown by Troll’s study 70 years ago (see also Troll, 1971). Still, as ever, more and comparative studies, in particular analyzing the great regional and local varieties, and linking the many factors involved in patterned ground formation and distribution (climatic, non-climatic: e.g. substrate, texture, water-holding capacity, moisture balance), are needed.

III Conclusions

Undoubtedly, Troll has left permanent footprints in physical geography, in particular through his publications on the relationships between vegetation and climates at the global (see the map of seasonal climates), the regional and the local scales in a three-dimensional view (e.g. vegetation map of Nanga Parbat). His article on patterned ground (Troll, 1944a) and a few related publications (Troll, 1943b, 1944a, 1944b, 1948a, 1961, 1971), however, have been at least as important and inspired younger generations of physical geographers.

It is only natural that 70 years after publication the interest in the 1944 article has been fading away. Moreover, modelling factors and processes have become more important during the last decades in comparison to extensive and time-consuming field research. Anyway, this excellent, well-differentiating overall synopsis documenting the state of the art before the post-war period is worth remembering as a valuable contribution to the knowledge of geomorphic processes and formation of patterned ground structures in periglacial/subnival environments. Even nowadays it should be recommended reading. In view of the extensive body of new information on the occurrence of patterned-ground types, and of the modern remote sensing techniques (sattelite imagery) in particular, it would be a real challenge to monitor systematically their regional distribution on both hemispheres and to produce a detailed monograph in a similar way as Carl Troll did in his extensive article (Chapters VIII–XI).