RIVER RESEARCH AND APPLICATIONS

River. Res. Applic. (2009) Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rra.1283

RESOURCE SUBSIDIES ACROSS THE LAND–FRESHWATER INTERFACE AND RESPONSES IN RECIPIENT COMMUNITIES JOHN S. RICHARDSON,a* YIXIN ZHANG b and LAURIE B. MARCZAK c a

Department of Forest Sciences, University of British Columbia, Vancouver, V6T 1Z4, Canada b Department of Biology, Texas State University, San Marcos, TX 78666, USA c Department of Biology and Biochemistry, University of Houston, Houston, TX 77004, USA

ABSTRACT Fluxes of resource subsidies, such as terrestrial leaf litter to streams and adult aquatic insects to riparian predators, are examples of important links between adjacent ecosystems. The importance of these cross-ecosystem resource flows from donor systems to recipient consumers is increasingly recognized. Streams, especially small streams with their high edge ratio with the terrestrial system, provide excellent models for the study of subsidies and a large portion of this literature has been produced by aquatic scientists. Field experiments manipulating flows between small streams and their riparian areas (e.g. leaf litter, terrestrial invertebrates, and adult aquatic insects to riparian areas) have indicated that consumers in streams and riparian areas are highly dependent upon such subsidies and the value of the subsidies are further modified by patterns of retention and pathways of use. Experiments typically indicate rapid growth or demographic responses by consumers, indicating these populations are resource limited or at levels of incipient population limitation, and can capitalize on short-term resource pulses. More press manipulations are still necessary to determine the dynamical consequences of subsidies for recipient communities. The nature of the subsidy (e.g. species of litter or invertebrates) and its timing are also important details that need further study. Finally, there are opportunities to consider the evolution of life cycle timing (modelling), interception strategies by recipient populations and short-term and long-term responses of communities. Copyright # 2009 John Wiley & Sons, Ltd. key words: allochthonous; consumers; dynamics; ecosystems; energy flux; linkage; nutrients; streams Received 16 October 2008; Revised 14 March 2009; Accepted 14 May 2009

INTRODUCTION Early conceptions of ecosystems held that ecological systems were self-contained entities and their dynamics were largely determined locally by processes internal to the system while external forces were relatively weak (Forbes, 1887). This concept of the relative boundedness of ecosystems has been a useful simplification in trying to understand the controls on the dynamics of ecosystems (e.g. Levin and Paine, 1974). However, over the past half century the demonstration of the importance of movements of organisms and energy across ecosystem ‘boundaries’ has changed the view of ecosystems to being considered as leaky and strongly influenced from beyond the local system (e.g. Likens and Bormann, 1974; Polis et al., 1997; Levin, 2005; Holt, 2008). One of the key processes that crosses ecosystem boundaries and links adjacent systems are resource subsidies. What is a resource subsidy? These are flows of biologically fixed energy and nutrients from one ecosystem to another, i.e. allochthonous resources produced outside of the recipient system. The consequences of these subsidies for recipient consumers can be either as a direct nutritional resource, or reducing a consumer’s costs of foraging by augmenting the local resource supply. The expected results of subsidies are increases in population productivity and enhanced population growth. Here, we distinguish active foraging across ecosystem boundaries from those of subsidies, thus mergansers or otters foraging for fish in rivers would not constitute a subsidized system in our use of the term, i.e. subsidy flows must be donor-controlled (see below). In a donor-controlled system the recipient *Correspondence to: John S. Richardson, Department of Forest Sciences, University of British Columbia, Vancouver, V6T 1Z4 Canada. E-mail: [email protected]

Copyright # 2009 John Wiley & Sons, Ltd.

J. S. RICHARDSON, Y. ZHANG AND L. B. MARCZAK

consumer cannot influence the supply rate of resources directly, i.e. they are dependent on the resource supply, such as the input of terrestrial leaf litter to streams. Resource subsidies have been known for decades, for example the importance to consumers of leaf litter inputs to streams or to soils (e.g. Lloyd, 1921; Hynes, 1941). The synthetic approach developed by Polis et al. (1997) allowed for the integration of these qualitatively different forms of donorcontrolled resource subsidies and facilitated progress on conceptually unifying these processes. There is considerable variation in the rates of subsidy fluxes in terms of magnitude, duration, amplitude and quality, with a spectrum of possible dynamical responses of consumers (Holt, 2008). The availability of theory linking these processes has resulted in formulation of a range of testable hypotheses (e.g. Polis et al., 1997; Loreau and Holt, 2004). Freshwater scientists have repeatedly demonstrated the utility of aquatic systems as model systems for studying resource subsidies. Since organisms that are mostly aquatic have specializations that identify them as distinct from terrestrial organisms, and vice versa, the distinction across these boundaries is relatively clear. It is from this physically distinct boundary at the ocean-shore interface that Polis and Hurd (1995, 1996) developed their empirical and theoretical work providing a modern synthesis of the importance of cross-ecosystem subsidies (Polis et al., 2004). From Polis’ examples of marine wrack on fringes of oceanic desert islands, to Dissolved organic carbon (DOC) inputs to freshwaters, leaf litter to streams, etc., aquatic systems in general, and streams in particular, have been forefront in the development and testing of these ideas. Oceans and lakes, while bounded by shorelines, are now known to be more heterotrophic than previously considered, i.e. their productivity is largely dependent on externally produced organic carbon (Cole et al., 2006; Cole et al., 2007). DOC, much of it from the terrestrial landscape, plays a central role in the functioning of lake ecosystems and support of aquatic food webs (Pace et al., 2004). By studying whole-lake ecosystems, Carpenter et al. (2005) found that carbon cycles relating to dissolved and particulate organic carbon, zooplankton and fishes are substantially subsidized by flows of terrestrial organic carbon from surrounding watersheds. Even more so than in lakes, the strong linkages of streams to the terrestrial realm through resource subsidies have contributed some of the best demonstrations of the importance of these cross-ecosystem fluxes. Streams provide a useful model system because they are gravitational attractors for material and so tend to passively accumulate resources from adjacent terrestrial systems (Leroux and Loreau, 2008) and also because of their high ratio of edge with the surrounding terrestrial environment. This is particularly true for small streams that have a high edge to area ratio; systems with this kind of geometry receive larger amounts of allochthonous material and energy as inputs into the food web. In this review we concentrate on examples of cross-ecosystem resource subsidies involving streams, and to some extent lakes, and particularly on examples from the past decade (Figure 1). Previous reviews by Polis et al. (1997) and Baxter et al. (2005) described additional examples of resource subsidies. Details of some of the consequences for consumers of resource flows to and from streams are available and hence we emphasize those examples. There can be little doubt that freshwater scientists have made important contributions to the development and testing of the significance of the dynamics of cross-ecosystem resource subsidies.

HOW HAS AQUATIC SCIENCE CONTRIBUTED TO THE DEVELOPMENT OF ECOLOGICAL CONCEPTS? Freshwater science has been at the forefront of testing of the roles of resource subsidies in recipient communities. Many of the best examples of resource subsidies come from the interface of aquatic-terrestrial systems, probably because of the relatively clear physical boundaries between adjacent ecosystems. Organisms primarily living in aquatic or terrestrial systems possess a suite of traits that are suited more to one system than another. However, there are also many organisms capable of moving across the aquatic-terrestrial boundary as part of their foraging behaviour or as a consequence of ontogenetic niche shifts of complex life cycles. Decades of experimental and descriptive work have demonstrated that cross-ecosystem resource subsidies across the diffusive interface between freshwater and terrestrial realms can have large consequences for consumerresource dynamics and the resulting population responses. These studies in freshwater have also contributed to our understanding of some of the controlling processes on populations and their communities, the development of spatially explicit models for communities and the theory of trophic cascades. Copyright # 2009 John Wiley & Sons, Ltd.

River. Res. Applic. (2009) DOI: 10.1002/rra

RESOURCE SUBSIDIES ACROSS THE LAND–FRESHWATER INTERFACE

Figure 1. A schematic illustration of some of the major flows of biologically fixed energy across the stream-terrestrial interface and along the fluvial network. Widths of arrows do not imply magnitudes of fluxes

EXAMPLES OF CROSS-ECOSYSTEM RESOURCE SUBSIDIES INVOLVING FRESHWATERS Leaf litter and other terrestrial inputs to aquatic ecosystems One of the best studied of all cross-ecosystem subsidies is the input of organic matter (Figure 1), particularly leaf litter, to streams (e.g. Hynes, 1941; Fisher and Likens, 1972; Richardson, 1991; Wallace et al., 1999; Webster et al., 1999). Many taxa feed more-or-less exclusively on decaying leaf litter (including surface biofilms and leaf tissue) in streams and lakes, a functional group referred to as detritivores or ‘shredders’. Amounts of 400–700 g ash-free dry mass m 2 year 1 of leaf litter can fall into forested streams (Richardson et al., 2005). Detailed experimental work over the past two decades has demonstrated through additions and interception of leaf litter inputs that consumer populations are strongly limited by these resource subsidies (e.g. Richardson, 1991; Dobson and Hildrew, 1992; Wallace et al., 1999). Populations of stream detritivores feeding on terrestrial litter have a large scope for rapid increase as resource abundance increases, and respond quickly to depletion of litter standing stocks (Wallace et al., 1999; Rowe and Richardson, 2001). Wallace et al. (1999) have demonstrated that in the absence of litter inputs, the productivity of an entire stream system, including predators, declined dramatically. Copyright # 2009 John Wiley & Sons, Ltd.

River. Res. Applic. (2009) DOI: 10.1002/rra

J. S. RICHARDSON, Y. ZHANG AND L. B. MARCZAK

Dissolved organic carbon (DOC: organic matter