In a state of flux: The energetic pathways that move contaminants from aquatic to terrestrial environments


Thumbnail image of

Photo credit: Frank Leung, iStock Photos

Today, interest continues to grow in the potential aquatic-to-terrestrial transport of contaminants through movement of insects, birds, reptiles, and other organisms or via the energetic pathways that quintessentially link aquatic and terrestrial ecosystems.

John Muir. Henry David Thoreau. Theodore Roosevelt. Rachel Carson. Preservationists, conservationists, politicians, scientists, authors, and many others have impacted the way we view and value our environment. The 1962 publication of Carson's Silent Spring not only awakened the public to the problem of environmental contaminants, but also prompted a change in the lens through which scientists view contaminated ecosystems. Despite the debates Carson's book sparked, she raised a heightened awareness of chemicals, how they affect entire ecosystems, and our collective responsibility for their use 1.

Venturing beyond the then-dominant toxicological perspective, a wider group of scientists began to explore how ecological relationships influence the pathways and consequences of contamination. Initially, studies of contamination in aquatic ecosystems focused largely on how terrestrial contaminants move into the water as well as the bioconcentration and bioaccumulation of contaminants through aquatic food webs 2. Today, interest continues to grow in the potential aquatic-to-terrestrial transport of contaminants through movement of insects, birds, reptiles, and other organisms or via the energetic pathways that quintessentially link aquatic and terrestrial ecosystems. Food webs provide an excellent window into these pathways and can serve as vehicles that carry energy and contaminants. For example, anadromous fish (those that return to their birthplace to spawn), such as Pacific salmon (Oncorynchus spp.), transport contaminants from marine or lake ecosystems to streams, where contaminants can then be moved into terrestrial food webs by salmon-feeding grizzly bears (Ursus arctos horribilis), or even aquatic invertebrates that consume salmon carcasses (reviewed in 3). Other well-documented biovectors are seabirds and penguins, which transport significant amounts of marine-derived contaminants to terrestrial environments through guano that accumulates in their large breeding colonies (reviewed in 3). Although there is the potential for transfer of contaminants from aquatic-to-riparian-to-terrestrial ecosystems anywhere there are aquatic-to-terrestrial fluxes, most examples of biovector transport feature contaminants that originate in marine systems and are moved to fairly specific sites, such as breeding colonies or salmon runs. The extent to which this phenomenon occurs in freshwater systems and involves taxa more likely to widely broadcast contaminants within watersheds is less understood.

In this Focus article, we address ecological and societal issues related to the aquatic-to-terrestrial transport of aquatic contaminants, with the spotlight falling on flowing water ecosystems. We highlight the ways in which a new understanding of the aquatic-terrestrial interface has prompted an integrated view of cross-boundary contaminant flows within complex ecological networks. We pay particular attention to aquatic insects, which as an important source of energy for riparian consumers such as arthropods, birds, mammals, and reptiles, are especially likely to move contaminants into terrestrial ecosystems 4–6 (Fig. 1). The linkages among aquatic and terrestrial systems represent an emerging ecological and environmental issue. We believe that contextualizing contaminant fluxes within this framework will yield significant short- and long-term benefits to ecological health and human well-being.

Figure 1.

(a) Adult ebony jewelwing damselfly (Calopteryx maculata); (b) spider of the family Dolomedes, which feeds on a broad diet including aquatic insects and small fish; (c) spider of the family Tetragnathidae (a vertical orb weaving spider that feeds predominantly on emerging adult aquatic insects) in its web along with emergent mayfly and other insect prey; (d) American dipper (Cinclus mexicanus), a semi-aquatic bird that feeds primarily on aquatic insects and small fish. Photo credits: S.M.P. Sullivan (a and d), A.R. Kautza (b and c).

A New Twist to an Old Problem

Fluvial ecosystems have long been considered as simply one of the many elements that comprise the landscape mosaic. Recently, a more holistic and detailed ecological view has emerged – one that reflects the diverse internal amalgam of spatiotemporal and ecological dimensions that constitute a “stream.” The complex relationships among these dimensions have led, in part, to a perspective that recognizes riverine landscapes as both internally heterogeneous and intimately linked with their surroundings via boundary dynamics. Indeed, a growing body of literature provides evidence that reciprocal exchanges of energy, matter, and organisms among streams and their adjacent riparian zones are essential for functional ecosystems (as reviewed in 5) (Fig. 2).

Figure 2.

Reciprocal food-web linkages (arrows represent energy-flow pathways) in a stream-riparian ecosystem. Solid arrows represent strong pathways (e.g., primary predator-prey relationships); dashed arrows represent weaker pathways (e.g., occasional or facultative predator-prey relationships).

A Holistic Ecological View

This conceptual shift in our understanding of aquatic-terrestrial connectivity has necessitated a more holistic ecological view of the fate of contaminants that enter aquatic systems. In particular, aquatic-to-terrestrial movement of contaminants is a topic of emerging importance because contaminant fluxes might be expected to move through the same complex ecological networks as nutrients and energy. At its core, this thesis is based largely on life history strategies of aquatic insects and their potential to serve as aquatic-to-terrestrial biovectors of contaminants. Early on, Menzie 7, for example, suggested that larvae of aquatic insects (e.g., Ephemeroptera – mayflies, Plecoptera – stoneflies, Trichoptera – caddis flies, Odonata – dragonflies and damselflies, Chironomidae – midges, Simulidae – black flies), intimately associated with aquatic sediments and water, would bioconcentrate contaminants and transport them out of the aquatic environment when the insects emerge as adults. Menzie's conceptualization has since been supported in recent work, which has illustrated that emergent aquatic insects can indeed transport aquatically derived contaminants to the terrestrial environment, where they can be subsequently accumulated by riparian consumers (Table 1). Others have shown indirect evidence linking aquatically derived contaminants with riparian and terrestrial food webs via emergent insects 4.

Table 1. Field investigations of aquatic-to-riparian exports of contaminants via emerging adult aquatic insects
  1. PCBs = Polychlorinated biphenyl.

PCBsHudson River, New York, USAContamination of PCB in nestling tree swallows (Tachycineta bicolor) reflected magnitude of contamination of emergent insects (primarily Diptera; up to 18,000 ng/g, wet wt of total PCBs in prey items collected from mouths of adults returning to nest to feed young).21
PCBsKalamazoo River Superfund Site, Michigan, USAMean total PCB concentrations (mg/kg, wet wt) of aquatic emergent insects (all orders collected) was 0.74 ± 0.56 from former impoundment and 0.13 ± 0.08 from upstream reference site.22
Inorganic mercury (Hg(II))East Fork Poplar Creek, Tennessee, USAEstimated that midges exported 4.1 g Hg(II)/yr from 2.1-km reach of industrially contaminated stream.23
PCBsTwelve mile Creek, South Carolina, USAΣPCB concentrations in riparian consumers ranged from 180–2740 ng/g in riparian spiders (Tetragnatha and Dolomedes) dependent on emergent insects; estimated that aquatic insects exported 6.13 g/yr of PCBs to 25 km of riparian zone sampled.6
PCBsLake Hartwell Superfund Site, South Carolina, USAMean PCBs in adult chironomids higher (1240 ng/g among study sites) than terrestrial insects (15.2 ng/g); similar pattern between riparian spider (820–2012 ng/g) and upland spider (30 ng/g) PCB concentrations.24
Nitrogen (N)Chikuma River, JapanAquatic subsidies transported anthropogenic N to riparian spiders; anthropogenic N transport depended on spider body size (smaller spiders used more aquatic subsidies).16
PCBsLake Erie & Detroit River, Ontario, CanadaPCBs bioamplified differently between male and female emergent maylfies (Hexagenia spp.); for males, the ratio of lipid equivalent PCB concentrations was 2.05 ± 0.38 for imagos/nymphs and 1.91 ± 0.18 for imago/subimago life stages.17
PCBsLake Hartwell, South Carolina, USAPCB transfer to spiders (via predation of aquatic emergent insects) extended a distance of ∼5 m inland.25

Here we present three organizing themes, each of which operates at an increasingly larger spatial scale, as a conceptual framework to address key considerations in understanding and predicting aquatic-to-terrestrial contaminant flows: population and community dynamics, landscape alterations, and climate change.

Population and community dynamics

Because species differ in bioaccumulation kinetics, movement patterns, and their interactions with other species in food webs or otherwise, community structure is expected to be one of the strongest mediators of aquatic-to-terrestrial contaminant fluxes. Even within species, population structure (e.g., age and size distribution), foraging plasticity, and behavior have been correlated with contaminant loads.

From the Pond to the Web to the Nest

Examples of how species-specific information can influence aquatic-to-terrestrial movement of contaminants come from a wide range of taxa, including:

Bivalves. Because mercury accumulation rates differ among bivalve species according to different feeding strategies and assimilation efficiencies 15, changes in the way bivalve communities are composed can conceivably influence how mercury moves through food webs.

Spiders. Spiders had more of the highly bioavailable methyl mercury than other invertebrates (e.g., lepidopterans and orthopterans) at the terrestrial-aquatic interface; therefore, they were thought to be responsible for transporting aquatic mercury into terrestrial food webs 4. The implication is that changes in spider populations might alter mercury transport into food webs. Even within groups of spiders, size and guild (relative to web building) constrains consumption rates of aquatic insects, and subsequently, the uptake of aquatically-derived contaminants 16.

Aquatic Insects. In a lake ecosystem, Daley et al. 17 attributed differences in PCB concentrations in adult male and female mayflies to the differential loss of lipid content during emergence. This suggested that transferring PCBs to wildlife consumers will vary based on the sex ratios of adult emergent insects. The authors estimated that differences in sex-ratio composition of mayflies in the diets of animals feeding on this insect could account for daily intake of PCBs that vary by 36%.

Fish. With their slower rates of elimination, longer exposure time of older individuals, and consumption of prey at higher-trophic levels by older and larger individuals, Hg concentrations increase in fish populations with the age and size of the individuals 2. Rypel et al. 18 found significant differences in PCB concentrations between sexes for channel catfish (Ictalurus punctatus), largemouth bass (Micropterus salmoides), and spotted bass (Micropterus punctulatus). The authors speculated this might be due to differences in reproduction, lipid deposition, and motility. Therefore, fish population structure is a critical factor in the subsequent transfer of contaminants to piscivorous birds and mammals, including humans.

Birds. Morrissey et al. 19 found different contaminant profiles (total organochlorines [OCs], PCBs, Hg) between resident and migratory American dippers (Cinclus mexicanus), a semi-aquatic songbird that feeds on aquatic benthic macroinvertebrates and small fish. Resident dippers breed on the main river and altitudinal migrants breed on tributaries, illustrating the importance of understanding and recognizing species-specific ecology in predicting contaminant pathways.

Mammals. Exposure to localized persistent contaminants (e.g., OCPs, PCBs, and PBDEs) can differ among individual river otters (Lontra canadensis) within the same population, based on individual movement and landscape use 20. Thus, understanding the behavioral ecology of species is likely to prove useful in understanding their role in contaminant pathways.

In some cases, the presence of the contaminant itself may elicit ecological changes through direct lethal and non-lethal effects on organisms in ways that shape the pathways and magnitudes of aquatic-to-terrestrial fluxes of contaminants. Among these mechanisms are trophic cascades (e.g., contaminants eliminating sensitive grazers, thereby causing the number of primary producers to increase); altered competitive interactions (e.g., direct effects of contaminants on sensitive species can alter interactions of tolerant species), and changes in the relative abundances of species within a community following a direct contaminant effect on a keystone species 8.

Habitat loss and alteration can cause various species to be added or removed from an environment and/or changes in the relative abundance of a species within a community. Such dynamics can fundamentally alter communities in ways that affect trophic transfers of contaminants. For example, predatory birds can induce top-down shifts in the community structure of insectivorous fish and the aquatic insects they consume. These changes can affect the magnitude of emergence events and thus the potential to transfer contaminants to terrestrial systems.

Nonnative and invasive species also figure prominently in community-mediated aquatic-to-terrestrial contaminant fluxes (Fig. 3). For example, the invasive round goby (Neogobius melanostomus) is thought to help mobilize contaminants in food webs and increase exposure to humans. This is the case because its persistence in contaminated environments attracts predatory fish, which are also popular game species, into polluted habitats 9. Baxter et al. 10 showed that when nonnative rainbow trout (Oncorhynchus mykiss) invaded a stream in northern Japan, they outcompeted the native Dolly Varden charr (Salvelinus malma) for the terrestrial arthropods that fell into the stream from riparian vegetation. In turn, Dolly Varden shifted their foraging to benthic invertebrate grazers, which indirectly increased algal biomass and decreased biomass of adult aquatic insects emerging from the stream, leading to a 65% decrease in the density of spiders of the family Tetragnathidae (horizontal orb-weavers whose diet consists mainly of adult aquatic insects).

Figure 3.

(a) Aquatic-to-terrestrial contaminant transfers (arrows) in a stream-riparian ecosystem and potential shifts in aquatic-to-terrestrial transfers due to the (b) invasion by high numbers of rusty crayfish (Orconectes rusticus). The rusty crayfish, a native to the Ohio River drainage, has expanded its range significantly into the upper Midwest, New England, and Ontario, where it can decrease the abundance of aquatic macroinvertebrate larvae and increase primary productivity (e.g., algae). Red arrows represent contaminant transfer from sediment and primary producers (e.g., periphyton) to consumers. Solid black arrows represent strong pathways (e.g., primary predator-prey relationships); dashed arrows represent weaker pathways (e.g., occasional or facultative predator-prey relationships). Changes in the magnitude of anticipated contaminant flux are represented by the thickness of the arrows. For illustrative purposes, this figure and Figure 2 represent simplified conceptualizations of a network of complex species interactions and subsequent aquatic-to-terrestrial contaminant fluxes. Larger fluvial systems would also include a planktonic component as part of the trophic links that would also be expected to affect the transport of contaminants from aquatic-to-terrestrial ecosystems.

Landscape alterations

Landscape changes might be expected to alter species-habitat relationships, species interactions, and network dynamics profoundly and in turn alter the linkages and magnitude of contaminants transported from aquatic-to-terrestrial ecosystems. Land use within a watershed can interact with local riparian conditions to affect the condition of the aquatic ecosystem. This may include changes in the water quality and a decoupling of aquatic-terrestrial systems that subsequently severs cross-boundary flows of reciprocal resources and significantly changes community structure and function 11. For aquatic-to-terrestrial contaminant fluxes, the configuration and complexity of the riverine landscape mosaic is thought to be highly influential. Changes in hydrology, channel morphology, riparian vegetation, and floodplain structure cause distinct riverine landscapes to develop in different land-use types. In turn, the land-use types might be expected to differ from one another in patterns of aquatic-to-terrestrial contaminant fluxes. Using examples from the Scioto River (OH, USA) we illustrate how aquatic-to-terrestrial contaminant fluxes might be expected to differ in riverine landscapes in predominantly forested, agricultural, and urban matrices and the potential implications on aquatic-to-terrestrial contaminant fluxes (see Landscapes of the Scioto River, Ohio, USA).

Landscapes of the Scioto River (Ohio, USA)

Thumbnail image of

Three Scioto River (Ohio, USA) riverine landscapes: (a) forested/natural landscape, (b) agricultural landscape, and (c) urban landscape. The Scioto River is a mixed-use tributary of the Ohio River and represents a valuable resource to the residents of Ohio, providing multiple ecosystem services including drinking water, irrigation, recreation, and biodiversity. Sediment contamination within the Scioto River is an issue that both federal and state agencies recognize. The Ohio Environmental Protection Agency has posted fish consumption advisories recommending that humans limit their consumption of fish from the Scioto due to mercury and PCB concentrations found in fish tissue.

Forested landscapes

Forested/natural riverine landscapes (a) are characterized by a high diversity of landscape elements, including geomorphic features (bars and islands, complex channel networks), riparian systems (wetlands, alluvial forests), and an active and connected floodplain.

Implications for aquatic-to-terrestrial contaminant fluxes

  • Wetlands mitigate contaminant transfers.

  • Sediment depositson the floodplain bioaccumulate in riparian food webs and beyond.

  • High aquatic-terrestrial connectivity leads to high aquatic-to-terrestrial contaminant flux.

  • Greater internal feedback loops may exist due to high patchiness and greater diversity of aquatic and riparian consumers and lower concentrations of contaminants due to lower use in the watershed.

Agricultural landscapes

River reaches in agricultural landscapes (b) typically range from highly channelized ditches to partially constrained reaches (as shown here). Agricultural riverine landscapes are commonly associated with loss of riparian vegetation, simplified channel structure, and a lack of or a low-quality floodplain. This example shows no floodplain connectivity on one side and poor-quality, ephemeral slack waters (i.e., wetted habitats where flow is minimal to non-existent except during high water) on the other.

Implications for aquatic-to-terrestrial contaminant fluxes

  • Contaminated sediments are deposited directly on agricultural fields.

  • High levels of contaminants as a result of agricultural activities.

  • Highly productive systems (due to nutrient enrichment) lead to greater biotransport potential by emergent aquatic insects.

Urban landscapes

Urban riverine landscapes tend to be highly controlled to prevent flooding, inhibiting the development of an active floodplain. Riparian zones are commonly narrow to non-existent, and active land management typically occurs up to the water's edge. Rivers are often dammed or partially impounded, resulting in lower flows and a shift from flowing (lotic) to lake (lentic) aquatic communities.

Implications for aquatic-to-terrestrial contaminant fluxes

  • Limited to no sediment deposition in floodplains.

  • No wetlands to mitigate contaminant transfers.

  • Low aquatic-terrestrial connectivity decreases aquatic-to-terrestrial contaminant flux, but contaminant levels are usually high due to surrounding land use.

  • Increased resident time of sediment due to altered flows (e.g., impoundments) increase bioaccumulation and concentrations in emergent insects.

  • Sensitive aquatic macroinvertebrate taxa (e.g., Ephemeroptera, Plecoptera, Trichoptera) fall out and are replaced with non-emergent (Oligochaeta, limpets [Ferrissia spp.], pond snails [Physella spp.] and small-bodied families [Simulidae, Chironomidae]) potentially weakening the aquatic-to-terrestrial contaminant pathway.

Climate change

One of the most obvious ways that climate change can influence aquatic-to-terrestrial contaminant fluxes is by altering community and population structure via shifts in species distributions and changes in abundance. As species distributions shift, expanding in some cases and contracting in others, communities will change, and this shift has potential to affect interactions in ways that may increase or reduce the extent of reterrestrialization and exposure. As with many other taxa, stream macroinvertebrates and predatory fish are likely to shift distributions in response to climate change, and this response can prompt fundamental shifts in trophic structure and, potentially, pathways of contaminant transfers. Changes in community structure are perhaps most likely to occur in high-altitude and high-latitude environments, which based on their simpler food webs, may be more sensitive to disruptions by contaminants than temperate or tropical systems. In the absence of range shifts, migratory species could still alter seasonal movement patterns in ways that affect contaminant transport.

Even in cases where communities retain similar composition, changes in temperature and precipitation can influence resources in ways that shift diets and alter food webs, as shown with American dippers (Cinclus mexicanus, 12). Increased temperatures could also affect foraging behavior and diet via changes to basal metabolic rate and energetic requirements. These changes, consequently, affect the rate and pathways of the uptake and elimination of contaminants.

Understanding Processes Through Space and Time

Predicting terrestrial-to-aquatic contaminant fluxes requires considering systems-level effects across multiple spatiotemporal scales. This includes knowledge about the ecosystem processes (e.g., hydrology and geomorphology), organismal physiology, community organization, and species interactions. In many ways, this approach is distinct from predicting terrestrial-to-aquatic contaminant fluxes that can be tracked largely through the chemical properties of contaminants and the physical routes that contaminants enter aquatic systems.

Although methodological constraints have limited our ability to understand the inherent complexity of natural systems, a suite of emerging tools and approaches is improving our ability to track aquatic-to-terrestrial energy fluxes and quantify their magnitude. For example, using natural abundances of stable isotopes now allows scientists to track contaminant fluxes within and between ecosystems 13 and to predict exposure to riparian and terrestrial consumers 6. Scientists also use the isotopic compositions of contaminants themselves to track the contribution of anthropogenic sources of contaminants in aquatic consumers. Food web and bioaccumulation models largely applied within single-system contexts may also be appropriate for the aquatic-terrestrial interface and beyond. In this vein, Blais et al. 3 proposed a mass balance model that incorporates an abiotic component, a simplified food web, and a spawning salmon component to understand biovector transport. Ultimately, multiscale approaches with temporal considerations (e.g., the potential of aquatic insects as biovectors will vary by season) will be valuable in targeting and predicting potential “transfer hot spots” and critical in understanding and predicting exposure risk. A variety of tools may aid this effort, including radio-telemetry and satellite technology to track the movements of animals that may act as biovectors and remotely sensed data and applications including geographic information systems (GIS) and light detection and ranging (LIDAR, including new aquatic-terrestrial, water-penetrating green LIDAR).

With the number of chemicals being registered for use continuing to increase, the imperative to understand contaminant pathways is stronger than ever. Not only may these contaminants collectively produce significant impacts on aquatic ecosystems and wildlife populations, but biologically mediated pathways also may result in contaminant transfers to human systems (see How contaminants may reach our food supply). Although exposure risk can be mitigated several ways, including public health and consumption advisories, exposure is linked tightly to socioeconomic and demographic attributes, such as education and income. Moreover, food choice can also be highly embedded in the cultural milieu of a society, as is the case with traditional diets of indigenous communities in the Arctic, which are linked tightly to nutritional, cultural, spiritual, and economic benefits. As such, the issue goes beyond the domain of public health and cannot be resolved by risk-based health advisories or dietary replacements alone 14. Thus, environmental justice issues might be anticipated to apply to aquatic-to-terrestrial contaminant fluxes as they have to other pathways.

How Contaminants May Reach Our Food Supply

Thumbnail image of

Aquatic-to-terrestrial transport of aquatic contaminants is likely to have far-reaching consequences, both ecologically and for human health and society, including transfers of zoonotic diseases, which, although not discussed in the present article, are an important consideration. Understanding the potential impacts of aquatic-to-terrestrial contaminant transfers requires additional research on known and potential biovectors, especially for species expected to “broadcast” contaminants widely and in spatiotemporally disjunct and/or unpredictable patterns.


We extend our thanks to J. Alberts, A. Kautza, and P. Tagwireyi for their assistance in the field and with geographic information systems work. This work was funded in part by The Ohio State University and McIntyre-Stennis funds.

Supplemental Data

Suggested reading list. (20 KB DOC)