Cent. Eur. J. Geosci. • 4(3) • 2012 • 376-382 DOI: 10.2478/s13533-011-0075-6

Central European Journal of Geosciences

The emergence of an ‘evolutionary geomorphology’?

Commentary

Johannes Steiger1,2∗, Dov Corenblit1,2

1 Clermont Université, Maison des Sciences de l’Homme, 4 rue Ledru, 63057 Clermont-Ferrand Cedex 1, France 2 CNRS, UMR 6042, GEOLAB – Laboratoire de géographie physique et environnementale, 63057 Clermont-Ferrand, France

Received 23 March 2012; accepted 1 July 2012

Abstract: Earth surface processes and landforms are modified through the actions of many microorganisms, plants and animals. As organism-driven landform modifications are sometimes to the advantage of the organism, some of these landform features have become adaptive functional components of ecosystems, concurrently affecting and responding to ecological and evolutionary processes. These recent eco-evolutionary insights, focused on feedback among geomorphologic, ecological and evolutionary processes, are currently leading to the emergence of what has been called an ‘evolutionary geomorphology’, with explicit consideration of feedbacks among the evolution of organisms, ecosystem structure and function and landform organization at the Earth surface. Here we provide an overview in the form of a commentary of this emerging sub-discipline in geosciences and ask whether the use of the term ‘evolutionary geomorphology’ is appropriate or rather misleading. Keywords: evolutionary geomorphology • biogeomorphology • ecosystem engineer • niche construction • eco-evolutionary dynamics © Versita sp. z o.o.

1. The concept of evolutionary geo- ary dimension, i.e. the effects on and responses of vege- morphology tation, microorganisms and multicellular animals [2]. The responses of living organisms to long-term geomorphic and ecological feedback have to be considered as the result of passive biotic evolutionary processes acting on individu- Organisms respond and contribute to the modification of als. The development of new adaptations to modify envi- their physico-chemical environment from the micro- to the ronmental constraints should not be considered as an ‘ac- global scale. As living organisms evolved over geologi- tive choice’ of building an improved physical environment. cal timescales, these biotic evolutionary changes affected Biotic evolution, in Darwin’s sense, strictly occurs as a Earth surface processes and a variety of landforms ad- consequence of natural selection at the organism level, justed to the new evolving life forms [1, 2]. In turn, certain where genotypic mutations, biotic and abiotic environmen- geomorphic modifications fed back to community structure tal selection pressures (e.g. physical disturbances and and function as well as organism evolution. stress, presence of predators, inter- and/or intra-species competition) and species adaptation (i.e. the development Thus, long-term feedback systems comprise an evolution- of biological traits beneficial for the survival and repro- duction of the organism or population) are involved [3]. Above the level of organisms, i.e. at the community and ∗E-mail: [email protected]

376 Johannes Steiger, Dov Corenblit

Figure 1. Feedback model of reciprocal interactions and adjustments between organisms and Earth surface processes and landforms. The model integrates in an eco-evolutionary framework geomorphic, ecological and evolutionary processes. Black colour represents physico- chemical elements and controls. Grey colour represents living organisms and biological controls. Earth surface processes, landforms and living organisms co-adjust according to feedback mechanisms.

ecosystem levels, self-adjustment processes are involved [8], physical ecosystem engineers are organisms that di- in the organization of structures (e.g. community assem- rectly or indirectly control the availability of resources to blages, landforms) and in the regulation of matter and other organisms by causing physical state changes in bi- energy fluxes. otic or abiotic materials. Physical ecosystem engineering by organisms is the physical modification, maintenance or The concept of ‘evolutionary geomorphology’ with explicit creation of habitats. The concept of ecosystem engineers consideration of feedback among the evolution of organ- was more recently integrated into the emerging concept isms, ecosystem structure and function, as well as land- of ‘evolutionary geomorphology’ [4]. form organization, was proposed earlier by the authors [4]. At first, the focus was on the role played by vege- We tentatively define ‘evolutionary geomorphology’ as the tation dynamics as a biotic force in addition to the main creation, modulation and adjustment of landforms involv- abiotic forces (i.e. gravity, solar energy, tectonics, chem- ing physical ecosystem engineers (sensu [8]), which im- ical weathering) in shaping the Earth’s surface, but it is plicate natural selection pressures on organisms through now being extended to other organisms. Overall, this can niche construction (sensu [6]) with positive or negative be considered as an eco-evolutionary one, i.e. a frame- evolutionary consequences. The conceptual basis of work that examines the interplay between ecology and ‘evolutionary geomorphology’ is presented in the feed- evolution at multiple scales (i.e. eco-evolutionary dynam- back model of reciprocal interactions and adjustments ics, sensu [5]), which also comprises geomorphological co- between organisms, Earth surface processes and land- adjustments. It aims (i) to describe niche constructing forms (Fig. 1). The feedback model differentiates between activities (sensu [6]) with evolutionary consequences for physico-chemical elements and controls on the one hand engineer species that modify Earth surface landforms to and living organisms and associated biotic controls on the their advantage or to the advantage of other species in the other. According to eco-evolutionary dynamics, biological ecosystem, and (ii) to explain the long-term (i.e. organ- traits and community structure and function are respec- ism adaptation) and short-term biological responses (i.e. tively selected and modulated: (i) through organism evo- community structure and function) to changes which engi- lution at the species level (e.g. genotypic mutations, co- neer species can induce in the geomorphic dimensions and evolution between predator and prey); (ii) through Earth dynamics of their niche. This biogeomorphological frame- surface processes, e.g. erosion or sedimentation, which work is therefore closely linked to the ecological concept modify the habitat and constitute environmental selection of ‘engineer species’ or ‘ecosystem engineers’ developed pressures on organisms; and (iii) through landforms which by ecologists [7, 8] and not to be confounded with ‘civil en- constitute characteristic habitats for organisms in con- gineering’, even though the authors of the concept consider junction with specific geomorphic characteristics (e.g. tex- Homo sapiens to be an ecosystem engineer. According to ture, relative altitude, slope, exposure to regular flooding).

377 The emergence of an ‘evolutionary geomorphology’?

From a geomorphological perspective, change induced by evolution of organisms in the Darwinian sense (natural se- organisms, notably ecosystem engineers, is (i) an active lection restricted to the organism level) on the one hand construction or modification of landforms (e.g. through and landscape adjustments on the other hand. The two- burrowing, digging, construction of structures, weather- way coupling between natural selection and landform ad- ing of rocks and substrate); and (ii) a passive modulation justment over evolutionary timescales finally leads to the of Earth surface processes through the modulation of mor- notion and concept of ‘evolutionary geomorphology’. phogenetic factors, such as gravity, water flow, wind, frost, However, we definitely refute the term ‘co-evolution’ when chemical alteration (e.g. stabilization of soil by roots, re- considering feedback between organisms and landforms, tention of fine sediment by vegetation on hillslopes) [9]. since evolution occurs in biota and not in landforms, which The term ‘modulation’ implies a modifying or controlling adjust to abiotic and biotic controls. Co-evolution sensu influence, which can be biotic or abiotic. stricto, i.e. in its biological meaning referring to species adaptations, corresponds to co-evolving traits between different living organisms connected to each other within 2. ‘Evolutionary geomorphology’ – the ecosystem (e.g. predator-prey, host-symbiont or host- parasite relationships). Landforms do not reproduce and an appropriate term? evolve in the Darwinian sense and, as pointed out by [6], abiotic components lack genes or any equivalent herita- The perspective presented here has recently contributed ble information. Furthermore, natural selection does not to the consideration of many Earth surface landforms not apply to the selection of efficient structures and functions only as abiotic structures adjusting to intrinsic and ex- above the organism level, therefore excluding communities, trinsic physico-chemical constraints but also potentially ecosystems and landforms. Organisms evolve through nat- functional components of ecosystems – ‘functional’ in its ural selection and landforms adjust to physico-chemical ecological meaning, i.e. referring to what an element and biotic constraints. However, landforms affect natural does in the context of its ecosystem - integrated in a net- selection and in return are affected by natural selection work of reciprocal interactions and adjustments between through the evolution of engineering traits of organisms. physico-chemical, ecological and evolutionary processes This feedback mechanism leads to the emergence of dy- [1, 10]. However, it could still be argued that describ- namic and adjusting biogeomorphic structures at various ing landscapes as results of evolutionary processes seems spatial scales. inappropriate and that the term ‘evolutionary geomorphol- ogy’ might be misleading or even constitute an oxymoron. Without the presence of biota, landforms clearly are phys- 3. Co-adjustment of river styles and ical features at both local and regional scales, and mor- phodynamics are expressions of a trade-off between physi- vegetation dynamics at geological cal and chemical forces of resistance [4], exhibiting change timescales over different timescales. The central question here is whether this change over time – when it is not solely con- The consideration of the relationship between vegeta- trolled by physico-chemical forces but also by evolving tion and geomorphic dynamics at aquatic-terrestrial inter- engineer species (i.e. within bioclimatic contexts favorable faces has been particularly fruitful for linking geomorphol- to living organisms) – can be referred to as ‘evolutionary ogy and biota over ecological and evolutionary timescales geomorphology’. [4, 11, 12]. Indeed, plants progressively evolved biome- Following [2], we point out that, in the presence of chanical and physiological traits and strategies that fa- biota, the consideration of the interrelationships between vored their establishment and propagation within coastal landform geomorphology and organisms on evolution- and fluvial environments [13]. The latter are subject to per- ary timescales provides an argument for employing the sistent sediment, water and energy fluxes, which require term ‘evolutionary geomorphology’. This is in the re- plants to withstand hydraulic selection pressures. Bio- stricted context proposed here, with a strict boundary be- logical traits developed to better survive under these con- tween natural selection acting at the organism level and straints are characterized by: (i) the resistance to break- other processes, such as self-adjustment, acting at higher age and uprooting, and development of root physiology ecosystem levels. Thus, the juxtaposition of the two words plasticity (e.g. large stem cross sections, strengthening ‘evolutionary’ and ‘geomorphology’, which in themselves tissues, deep rooting systems and respiration facilities have strong connotations and seemingly contradict each during floods); (ii) the passive prevention of physical dam- other initially, becomes appropriate when considering the age (e.g. stem and leave flexibility, size and shape, canopy

378 Johannes Steiger, Dov Corenblit

reconfiguration abilities to reduce the hydrodynamic force scale. However, even though certain landforms, such as during floods, brittle twig bases that enable living stems dunes, alluvial bars, single meandering channels, rounded to break free); and (iii) the resilience to mechanical de- hillslopes and hydrological networks, would exist with- struction or sediment burial (e.g. resprouting from both out life [17], many structural and dynamic properties of roots and damaged shoots, clonal growth, buoyant seeds these landforms are modulated by the effects of engineer and fragments) [14]. By favoring vegetation establish- species. ment within river margins, these resistance and resilience strategies contribute to increase river bank cohesiveness and fine sediment and nutrient retention within stable is- 5. Constructed landforms as lands and floodplains [15]. An extensive and decisive mod- ification of fluvial dynamics and river styles at geological adaptive functional components of timescales associated with the evolution of plants occurred ecosystems during the Early Palaeozoic period [12, 16]. Evolving plants and large woody debris converted shallow braided Even if we accept the conservative delineation of Dietrich river styles into single-thread meandering channels and and Perron [17], we suggest that the role of ecosystem into island braided and anastomosing channels. Concomi- engineers played in landform evolution can be significant. tantly, engineering populations of plants evolved new life In biotic contexts, within river corridors [18], dune systems history traits (e.g. new timing of their life cycle and mor- [19], and salt marshes [20], the texture, size, shape and phology) as a result of the selection pressures they in- temporal changes of alluvial or intertidal bars and dunes duced via their hydrogeomorphic modifications within flu- are modified by the engineering morphological, biome- vial corridors (e.g. decrease of exposure to hydraulic con- chanical and life history attributes of pioneer plants. The straints). above-mentioned fact is also illustrated by animals influ- encing river systems and saltmarshes, for example dam constructing beavers [21], crayfish constructing pits and 4. Biotic signatures in the Earth’s mounds on gravel substrates [22], aquatic mollusks adding physical structure to the environment via shell production topography and resulting reefs [23] or burrowing crabs [24]. These modulations divert landform texture, morphology and dy- As suggested by Dietrich and Perron [17], the success in namics from the state they would reveal in strictly abiotic identifying unique signatures of life on topography, i.e. conditions. We point out that in certain cases these land- a landform that could only exist in the presence of life, form modulations are favorable for the engineer species certainly depends on the occurrence scale and frequency themselves and at the same time for many other species of certain landform properties and the technical possi- which have adapted to the constructed ecosystem [2, 7]. bility to detect them through high-resolution topographic We suppose that ecosystem engineer dynamics may have data. Unique signatures, such as animal and insect nests, positive, negative or no specific effects on the engineer- mounds, burrows and galleries, very frequently occur at ing organism and/or other organisms. The examples dis- the sub-metre scale. A significant proportion of the Earth cussed here are positive ecosystem engineer effects, where surface is subject to such activities with long-term cumula- ecosystem engineering provides a benefit to the engineer tive impacts. Furthermore, life itself can constitute unique or other species [8]. Nevertheless, positive niche con- signatures on the topography at large scales, through au- struction as defined by [6] – i.e. on average increasing togenic bioconstruction activities. For example, Rudist the fitness of the niche-constructing organism itself – can and coral reefs respectively represent fossil and exist- have more or less negative impacts on other species in ing topographic signatures of life on the Earth surface the ecosystem. These impacts may enhance evolutionary at a large scale and have varied in the extent of their responses of the affected species, or lead to their exclu- development at different periods in Earth history. Di- sion from the ecosystem if they lack the capacity to adapt etrich and Perron [17] excluded such constructional fea- to the modified environmental conditions. The construc- tures from their approach, defining these structures as tion of a beaver dam, for example, shows such negative ‘life itself’ rather than the influence of life on topography impacts when terrestrial organisms are submerged by the through the mediation of sediment erosion, transport and newly formed pond or lake. However, the question of to deposition processes. In line with that restriction, they which extent ecosystem engineering has negative effects argued that unique signatures of life on the topography on the ecosystem engineer itself seems more difficult to of the Earth’s surface may not exist above the sub-metre answer. [6] defines negative niche construction as niche-

379 The emergence of an ‘evolutionary geomorphology’?

constructing activities that change environments in such action to the effects and responses of engineer organisms, a way that engineer specie fitness is reduced, e.g. by modulated landforms become adaptive functional compo- inducing a discordance between the organism and its en- nents embedded within the biogeomorphic system (i.e. an vironment. abiotic landform and its associated community, including engineer species), e.g. by way of positive niche construc- We shall now attempt to discuss this point by using the ex- tion sensu [6], become adaptive functional components em- ample of pioneer riparian trees as ecosystem engineers in bedded within the biogeomorphic system (i.e. an abiotic fluvial corridors [25]. In order to colonize and reproduce landform and its associated community, including engineer within hydrogeomorphologically disturbed river systems, species). Ultimately, these modulations can become real pioneer riparian trees (e.g. cottonwood) evolved biologi- extended phenotypes of the constructing organism, sensu cal adaptations, such as high bending stability and strong Dawkins [29] as suggested for ponds created by beavers. anchorage, allowing them to withstand floods, or rapid root According to [29], ecosystem engineering traits and their elongation after germination, reducing drought mortality effect on the physical environment and geomorphology can during flood recession and causing rapid substrate stabi- thus, over evolutionary time, be encoded in the genetics lization after germination [26]. Once colonized, these pio- of the engineer. Biogeomorphic structures (e.g. fluvial is- neer trees increase surface roughness and favor sediment lands, galleries in soil, ant or termite mounds, coral reefs) accretion, thus actually enhancing the biogeomorphologi- are not explicitly written in the genotype of the engineer cal succession [27]. This in turn eventually leads to the ex- species, but they emerge from the interactions between clusion of this particular pioneer engineer species, which the organisms and the physical environment [2]. In our depends on hydrogeomorphic disturbances for recruitment. opinion, the feedback leading to extended phenotypes of However, the disturbance regime will decrease and even- organisms provides a further argument for suggesting the tually disappear because under biotic (here: vegetation) term of ‘evolutionary geomorphology’ in the context pre- control, the fluvial corridor tends to raise its topography sented here. and stabilize. If river systems were not naturally dynamic systems or if the resistance of engineer plants to mechan- ical constraints were too strong, it could be argued that this particular engineer species would be excluded from 6. Perspectives wide areas or could even become extinct due to its own ecosystem engineering within fluvial corridors. We sug- Even if landforms engineered by life are not ’unique’ to- gest that this is not the case as long as river systems pographic signatures, the emerging landform properties remain dynamic, providing suitable areas for colonization adjusted to organisms’ activities must be considered to be by new generations of riparian pioneer trees, and as long of crucial importance because they contribute to complex as the resistance traits of these riparian plants do not be- interactions and reciprocal adjustments between ecologi- come so strong as to completely stabilize the river system. cal and evolutionary processes within the ecosystem. We Furthermore, certain riparian trees, such as cottonwood, therefore suggest that focusing on these geomorphologic evolve traits related to the modifications they induce in modulations, rather than on the search of a wholly unique the fluvial environment. The stabilization and construc- signature of life on topography, will constitute a major tion of landforms combined with local accumulation of re- issue for geomorphologists working on the Critical Zone. sources lead to development as a feedback of competitive We are only just beginning to understand how life af- adult traits, providing advantages for resource uptake (e.g. fects landform dynamics over geological timescales and high size and deep root system). Cottonwood currently how, in turn, landforms or properties of landforms mod- combines ‘r’ (ruderal opportunistic) traits at the juvenile ify ecosystem structure and function through the modifi- stage with K (competitive) traits at the adult stage when cation of organism fitness. Conceptual advances [4] and it reaches its sexual maturity on stabilized surfaces. The recent findings showing that engineered alluvial facies initial abiotic conditions selected the r traits in the long are the signature of feedback between hydrogeomorphic term, whereas the K traits may correspond to an evo- processes and plant evolution [12, 16] support the emer- lutionary feedback associated with engineer effects [28]. gence of what we have called ‘evolutionary geomorphol- Thus, even if the described engineer effects may seem to ogy’. Nevertheless, we are aware that further exchanges be negative in the short term, they appear to be rather and collaborations among geomorphologists, notably bio- positive in the long term. geomorphologists, ecologists and especially evolutionary As in the case of riparian trees, the properties of many biologists are required to support our as yet limited con- landforms do not only adjust to external physical and ceptual framework. As suggested by [9], inter-disciplinary chemical forces but are also engineered by life. As a re- studies have to be designed to explicitly examine the re-

380 Johannes Steiger, Dov Corenblit

cursive influences among genes, phenotypes and geomor- Races in the Struggle for Life. John Murray, London, phic processes and structures on the Earth. 1859 We suggest that to identify the role of intraspecific ge- [4] Corenblit D., Steiger J., Vegetation as a major con- netic variability of organism populations related to en- ductor of geomorphic changes on the Earth surface: gineer traits in order to quantify the variability in abi- toward evolutionary geomorphology. Earth Surface otic responses and resulting feedbacks on engineer pop- Processes and Landforms, 2009, 34, 891-896 ulations and communities will be a future challenge (see [5] Pelletier F., Garant D., Hendry A.P., Eco-evolutionary also [2]). This will require a formal examination of how, dynamics: Introduction. Philosophical Transactions of and to what degree, genetic variation in engineer species the Royal Society B-Biological Sciences, 2009, 364, may influence physical habitat properties and thus com- 1483-1489 munity structure and function in the short term, as well [6] Odling-Smee F.J., Laland K.N., Feldman M.W., Niche as adaptation and speciation over evolutionary timescales, Construction: the Neglected Process in Evolution. especially based on paleontological evidence. Genetically Princeton University Press, Princeton, NJ, 2003 variable engineer traits that could be associated with cru- [7] Jones C.G., Lawton J.H., Shachak M., Organisms as cial biogeomorphologic properties have to be identified by ecosystem engineers. Oikos, 1994, 69, 373-386 combining field or laboratory studies. The links among [8] Jones C.G., Lawton J.H., Shachak M., Positive and specific variations of these traits, specific landform mod- negative effects of organisms as physical ecosystem ulations, associated community structures and ecological engineers. Ecology, 1997, 78, 1946-1957 responses have yet to be established. [9] Corenblit D., Gurnell A.M., Steiger J., Tabacchi E., Reciprocal adjustments between landforms and liv- We are aware that these suggestions - ranging from ing organisms: Extended geomorphic evolutionary in- designing inter-disciplinary studies to linking particular sights. Catena, 2008, 73, 261-273 landform modulations with intraspecific genetic variabil- [10] Fisher S.G., Heffernan J.B., Sponseller R.A., Welter ity of engineer ecosystems and biological life traits – are J.R., Functional ecomorphology: Feedbacks between still in the theoretical stage. We are confident, however, form and function in fluvial landscape ecosystems. that more specific and more practical recommendations Geomorphology, 2007, 89, 84-96 will emerge from inter-disciplinary research efforts. [11] Murray A.B., Knaapen M.A.F., Tal M., Kirwan M.L., Biomorphodynamics: Physical-biological feedbacks that shape landscapes. Water Resources Research, 7. Acknowledgments 2008, 44, W11301. doi:10.1029/2007WR006410 [12] Davies N.S., Gibling M.R., Cambrian to Devonian The authors wish to thank all four reviewers for their help- evolution of alluvial systems: The sedimentological ful and constructive suggestions and discussions of this impact of the earliest land plants. Earth-Science Re- commentary, which helped us advance our own thoughts views, 2010, 98, 171-200 and reflections. We also thank S. Lühmann for English [13] Gibling M.R., Davies N.S., Palaeozoic landscapes proofreading of the manuscript. shaped by plant evolution. Nature Geoscience, 2012, 5, 99-105 [14] Lytle D.A., Poff N.L., Adaptation to natural flow regimes. Trends in Ecology & Evolution, 2004, 19, References 94-100 [15] Collins B.D., Montgomery D.R., Fetherston K.L., Abbe [1] Meysman F.J.R., Middelburg J.J., Heip C.H.R., Biotur- T.B., The floodplain large-wood cycle hypothesis: A bation: a fresh look at Darwin’s last idea. Trends in mechanism for the physical and biotic structuring of Ecology & Evolution, 2006, 21, 688-695 temperate forested alluvial valleys in the North Pa- [2] Corenblit D., Baas A.C.W., Bornette G., Darrozes J., cific coastal ecoregion. Geomorphology, 2012, 139, Delmotte S., Francis R.A., Gurnell A.M., Julien F., 460-470 Naiman R.J., Steiger J., Feedbacks between geo- [16] Davies N.S., Gibling M.R., Evolution of fixed-channel morphology and biota controlling Earth surface pro- alluvial plains in response to Carboniferous vegeta- cesses and landforms: A review of foundation con- tion. Nature Geoscience, 2011, 4, 629-633 cepts and current understandings. Earth-Science Re- [17] Dietrich W.E., Perron J.T., The search for a topo- views, 2011, 106, 307-331 graphic signature of life. Nature, 2006, 439, 411-418 [3] Darwin C., On the Origin of Species by Means of [18] Bertoldi W., Gurnell A.M., Drake N.A., The topo- Natural Selection, or the Preservation of Favoured

381 The emergence of an ‘evolutionary geomorphology’?

graphic signature of vegetation development along namics modulated by burrowing crab activities in con- a braided river: Results of a combined analysis of trasting SW Atlantic intertidal habitats. Estuarine airborne lidar, color air photographs, and ground Coastal and Shelf Science, 2008, 80, 365-373 measurements. Water Resources Research, 2011, 47, [25] Gurnell A., Petts G., Trees as riparian engineers: the W06525, doi:10.1029/2010WR010319 Tagliamento river, Italy. Earth Surface Processes and [19] Baas A.C.W., Complex systems in aeolian geomor- Landforms, 2006, 31, 1558-1574 phology. Geomorphology, 2007, 91, 311-331 [26] Corenblit D., Steiger J., Gurnell A., Naiman R.J., [20] Van Hulzen J.B., Van Soelen J., Bouma T.J., Mor- Plants intertwine fluvial landform dynamics with eco- phological variation and habitat modification are logical succession and natural selection: a niche strongly correlated for the autogenic ecosystem engi- construction perspective for riparian systems. Global neer Spartina anglica (Common cordgrass). Estuaries Ecology and Biogeography, 2009, 18, 507- 520 and Coasts, 2007, 30, 3-11 [27] Corenblit D., Tabacchi E., Steiger J., Gurnell A., Re- [21] Naiman R.J., Animal influences on ecosystem dynam- ciprocal interactions and adjustments between flu- ics. BioScience, 1988, 38, 750-752 vial landforms and vegetation dynamics in river corri- [22] Johnson M.F., Rice S.P., Reid I., Increase in coarse dors: a review of complementary approaches. Earth- sediment transport associated with disturbance of Science Reviews, 2007, 84, 56-86 gravel river beds by signal crayfish (Pacifastacus le- [28] Francis R., Corenblit D., Edwards P., Perspectives niusculus). Earth Surface Processes and Landforms, on biogeomorphology, ecosystem engineering and 2011, 36, 1680-1692 self-organisation in island-braided fluvial ecosys- [23] Gutierrez J.L., Jones C.G., Strayer D.L., Iribarne O.O., tems. Aquatic Sciences - Research Across Bound- Mollusks as ecosystem engineers: the role of shell aries, 2009, 71, 290-304 production in aquatic habitats. Oikos, 2003, 101, 79- [29] Dawkins R., The Extended Phenotype. Freeman, Ox- 90 ford, UK, 1982 [24] Escapa M., Perillo G.M.E., Iribarne O., Sediment dy-

382