Hydrol. Earth Syst. Sci., 10, 967–979, 2006 www.hydrol-earth-syst-sci.net/10/967/2006/ Hydrology and © Author(s) 2006. This work is licensed Earth System under a Creative Commons License. Sciences Pattern, process, and function in landscape ecology and catchment hydrology – how can quantitative landscape ecology support predictions in ungauged basins? B. Schroder¨ University of Potsdam, Institute of Geoecology, Potsdam, Germany Received: 5 May 2006 – Published in Hydrol. Earth Syst. Sci. Discuss.: 29 June 2006 Revised: 20 November 2006 – Accepted: 6 December 2006 – Published: 19 December 2006 Abstract. The understanding of landscape controls on the In the following, the terms landscape and catchment are used natural variability of hydrologic processes is an important re- interchangeably, but the first represents the ecological the lat- search question of the PUB (Predictions in Ungauged Basins) ter the hydrological perspective. initiative. Quantitative landscape ecology, which aims at un- In this context, “patterns” are defined as observations ex- derstanding the relationships of patterns and processes in dy- hibiting a spatial or temporal structure that is significantly namic heterogeneous landscapes, may greatly contribute to different from a random process realisation. These patterns this research effort by assisting the coupling of ecological contain information on the mechanisms which they emerge and hydrological models. from (Grimm et al., 2005). “Processes” are understood as the The present paper reviews the currently emerging rap- interactions of different objects in an environment. “Func- prochement between ecological and hydrological research. It tion”, however, has different meanings in environmental sci- points out some common concepts and future research needs ences, denoting either processes, roles, services or the “func- in both areas in terms of pattern, process and function analy- tioning” of a whole system with perspectives focusing either sis and modelling. Focusing on riverine as well as semi-arid on the performance of specific objects or on their importance landscapes, the interrelations between ecological and hydro- for a specific system (Jax, 2005). I will use the latter defini- logical processes are illustrated. Three complementary ex- tion here. amples show how both disciplines can provide valuable in- formation for each other. I close with some visions about Based on hierarchy theory and further developments, land- promising (landscape) ecological concepts that may help ad- scapes have been referred to as complex adaptive systems, in vancing one of the most challenging tasks in catchment hy- which patterns at higher levels emerge from localised inter- drology: Predictions in ungauged basins. actions at lower levels (Levin, 1998). Complexity arises from the interplay between intraspecific and interspecific biotic in- teractions and from different abiotic constraints and interact- ing driving forces and disturbances – all of them acting on 1 Introduction a hierarchy of spatial and temporal scales. Understanding of Landscape ecology and catchment hydrology, both disci- these complex interactions, identifying the underlying driv- plines deal with patterns and processes as well as their in- ing forces and the reliable prediction of resulting system’s teractions and functional implications on a variety of scales responses are the main objectives of environmental research. (Turner, 2005b; Sivapalan, 2005). Thus, it is reasonable to In this context, a typical area of landscape ecological re- study the interplay between ecological and hydrological pat- search is the analysis of the effect of spatiotemporal resource terns and processes and to seek for mutual possibilities to as- distribution on the persistence, distribution and richness of sist either discipline in dealing with their respective research species (e.g. Wiens, 2002a). Catchment hydrology typically questions. Each discipline has developed its own theories focuses on the understanding of the effect of the spatiotem- and methodologies; an interdisciplinary approach assembles poral distribution of soil and topographical properties on the the respective benefits and simultaneously provides an alter- soil moisture pattern or on the runoff response (e.g. Wilson native viewpoint on the same complex system: landscapes. et al., 2005). Both objects of interest, species, soil moisture, and runoff, play important roles in specific landscape func- Correspondence to: B. Schroder¨ tions as for instance biomass production, nutrient cycling, or ([email protected]) groundwater recharge. Published by Copernicus GmbH on behalf of the European Geosciences Union. 968 B. Schroder:¨ Landscape ecology meets catchment hydrology Thanks to the substantial methodological advances in the 2.1 Landscape ecological concepts – applicable to catch- area of observation (e.g. remote sensing), analysis (e.g. ge- ment hydrology? ographical information systems, spatial statistics), and mod- elling (e.g. digital terrain modelling, physically-based simu- Landscape ecologists describe heterogeneity in landscapes lation modelling), the availability of computer power and the in terms of two concepts: patch-matrix and gradients (e.g. development of theories no longer neglecting space (Kareiva, Turner et al., 2001; Wagner and Fortin, 2005). The first re- 1994), ecologists as well as hydrologists turned to a spatial lates to island-biogeography (MacArthur and Wilson, 1967) paradigm – considering spatial and spatiotemporal patterns, and metapopulation theory (Hanski and Gilpin, 1997), the relationships, and processes (Grayson and Bloschl¨ , 2000a). second to niche theory (Hutchinson, 1957) and community Accordingly, recent scientific questions in landscape ecol- ecology (e.g. Austin, 1985). Patches are defined depending ogy and hydrology focus on the interactions of patterns and on the scale and the research question (Addicott et al., 1987); processes and their functional implications. Not only catch- patches differ in patch quality, their boundaries affect flows ment hydrologists but also landscape ecologists apply mod- of energy, material, and species; patch context matters, and elling approaches to tackle this task. Based on their respec- composition and configuration of patches affect local and re- tive theoretical backgrounds (cf. Beven, 2002; Reggiani and gional processes (Wiens, 2002a). Schellekens, 2003; Wiens, 2002b; Levin, 1992), phenomeno- Wu and Levin (1997) describe ecological systems as hier- logical models (sensu Bolker, 2006) are used for pattern de- archical dynamic mosaics of patches (cf. Poole et al., 2004). scriptions, whereas mechanistic models are used for process Local patch dynamics can constitute shifting mosaics – so- description and pattern generation in an adaptive cycle of called mosaic cycles – if the patches exhibit similar but out- inference – i.e. formulating, testing, and rejecting hypothe- of-phase dynamics (e.g. Olff et al., 1999; Remmert, 1991; ses on the basis of comparisons between observed and sim- Watt, 1947). Unsurprisingly, this kind of shifting landscape ulated patterns (Holling and Allen, 2002). Recent develop- mosaics is also found in hydrologically controlled systems ments in ecological and hydrological modelling emphasize (Bornette and Amoros, 1996; Malard et al., 1999; Latterell the use of multiple patterns providing insight into different et al., 2006). aspects of the studied system for model building and calibra- Riverine landscape ecology continues the success story of tion (Grimm et al., 2005; Wiegand et al., 2004; Beven, 2006). the landscape ecological framework focusing on the inter- The present paper reviews the currently emerging rap- face of terrestrial and aquatic systems (Ward et al., 2002a; prochement between ecological and hydrological research. It Tockner et al., 2002). According to Wiens (2002a), all rele- points out some common concepts and future research needs vant concepts derived from landscape ecological theory can in both areas in terms of pattern, process and function analy- be exemplified within riverine landscapes – and vice versa, sis and modelling. After presenting some already realised or riverine systems provide good opportunities to test this the- realisable collaborations, I close with some visions regarding ory. promising concepts from (landscape) ecology that may help Since organisms determine the structure and functioning advancing one of the most challenging tasks in catchment of landscapes (Covich et al., 2004), many landscape ecol- hydrology: Predictions in ungauged basins (PUB). ogists follow an organisms-centred perspective (e.g. Wiens et al., 1993). One aim of quantitative landscape ecol- ogy is the understanding of species-habitat relationships and the prediction of the spatio-temporal species distribu- tion by means of species distribution modelling (Guisan and 2 Interplay between ecology and hydrology Thuiller, 2005). Species habitat selection is controlled by environmental resources on a hierarchy of spatial and tem- poral scales (Mackey and Lindenmayer, 2001). On small The interplay between ecological and hydrological research scales, selective forces are mainly biotic interactions such as commences on different levels and scales. Several stud- predation and competition; on larger scales, the abiotic en- ies present a growing number of emerging rapprochements vironment and related disturbance regimes are more impor- between ecological and hydrological research in different tant (Biggs et al., 2005).