A RQUITECTURA DE LA SOLUCIÓN

Lake (2000) defined three types of disturbance and ecological responses to them in stream ecosystems to characterize flow generated disturbances and the response of the biota to them; presses, pulses and ramps. These types of disturbance and ecological response can be applied to anthropogenic pollution and provide an explicit framework to characterize their subsequent temporal patterns. Presses are disturbances that arise sharply and then reach a constant and maintained level, while pulses are short-term and clearly delineated disturbances (Lake, 2000). Ramps are disturbances whose strength increases steadily over time that may or may not reach an asymptote after an extended period of time (Lake, 2000).

Figure 2.1. Temporal pattern of effluent discharge on a hypothetical yearly timescale. The solid line represents the press disturbance generated by nutrients, BOD and solids and the dashed lines account temporal variation in their magnitude. Solid triangles represent pulses of veterinary medicines released within the effluent as discrete events in time that vary in their duration (horizontal bars) and magnitude (displacement along the ordinate) (adpated with permission from Lake, 2000).

The temporal pattern of effluent discharge from land-based salmonid farms can be best defined as a press disturbance with nested pulse disturbances (Figure 2.1).

Although the concentrations of BOD, nutrients and solids in the effluents vary with time (e.g., daily pulses and annual cycles) due to factors such as feeding and excretion, the rearing stage of the fish, temperature and unpredictable events (e.g., system failure), their release is continuous on a yearly timescale and can be considered to fluctuate within a flexible range of values (e.g., Table 2.1).

Alternatively, the use of veterinary medicines on a therapeutic basis is discrete and its frequency of occurrence (i.e., number of events from t1 - tn) is related to epidemiological factors of the disease being treated. Their prophylactic use should follow a similar discrete pattern, as preventive treatments are likely to be associated to specific rearing stages and/or times of the year. The occurrence of other chemical compounds, such as disinfectants, will also be discrete as they are associated to non-continuous management actions (e.g., tank cleaning).

The solid line in Figure 2.1 represents the effluent discharge of BOD, nutrients and solids and the dashed lines account for temporal variability in their concentrations that are not monotonically increasing or decreasing. The pulse disturbances generated by the discharge of veterinary medicines nested within the press disturbance generated by BOD, nutrients and solids are represented by black triangles in Figure 2.1. Pulses can vary in their duration and concentration from time to time, represented by the triangle's horizontal bars and the displacement of the triangles along the ordinate in Figure 2.1, respectively. However, the fact that certain veterinary medicines might persist in stream sediments (e.g., OTC - Rose and Pedersen, 2005) implies that their frequency of occurrence as a disturbance in the environment may not be synchronous with their frequency of occurrence in the effluent. If a compound remains bioactive or regains bioactivity through e.g., sediment resuspension, its frequency of occurrence may result in pulses that exceed its frequency of discharge in the effluent or effectively in a press type disturbance.

In continuous, multiple cohort production programmes where specific life stages (e.g., alevins) are successively and continuously produced, specific compounds may be continuously used and discharged within the effluents (e.g., to treat Saprolegnia sp.).

The response of ecological endpoints to disturbance (i.e., press, pulse or ramp) (Figure 2.2) is dependent on the resolution of the ecological metric used to assess

the response. A ramp response, as shown in Figure 2.2, can eventually lead to the extinction of an ecological endpoint and is most likely to be observed when considering structural endpoints of low biological organization. A press response in community level parameters of benthic invertebrates, for example, can imply a ramp response (i.e., local extinction) from one or several species. The disappearance of shredders from the stream segment affected by trout farm effluents in the study of Camargo et al. (1992) (see section 5.3) is an example of such a case. The studies reviewed in section 5 suggest that several structural and functional endpoints at the community and ecosystem level of organization in stream reaches receiving salmonid farm effluents respond as a press (i.e., they have shifted to new perturbed states). This is particularly apparent in the case of the benthic macroinvertebrate community, whose consistent response across studies indicates a shift towards taxonomic and functional groups tolerant of altered physicochemical variables and able to efficiently exploit the niches created by disturbance. The degree to which pulse responses of particular ecological endpoints might be nested within press responses and the role of multiple stressors in eliciting these nested responses, however, is not clearly understood.

Figure 2.2. Temporal response pattern of ecological endpoints to disturbance. (A) pulse, (B) press and (C) ramp. The x-axis is on a generic timescale (adapted with permission from Lake, 2000).

The long term effects of a continuous pollution source can be assessed by considering the ecosystem's resilience once the polluting events finish (e.g., once the salmonid farm stops operating). Streams have the ability to displace the effects of stressors by diluting them and exporting them downstream (Pratt and Cairns, 1996), giving them a remarkable natural ability to recuperate from stress. The time-scales involved in stream recovery from land-based salmonid farm effluent disturbances once they have ceased to operate are relatively unknown. Hurd et al.

(2008) recently showed using ¹³C that the sediments of a spring creek remained enriched with hatchery derived C for several years after hatchery closure. To our knowledge, this is the only study that has made a partial assessment of the recovery of stream food webs after the cessation of effluent discharge.