Implications of leaky boundaries for compartmentalised control of pathogens: A modelling case study for bacterial kidney disease in Scottish salmon aquaculture

Abstract

Compartmentalisation is the focusing of disease control measures on valued and epidemiologically isolatable populations. However, pathogens may transmit across compartment boundaries and undermine local control measures. Bacterial kidney disease (BKD) affects both salmon and trout farms and a compartmentalised management programme has been adopted for Scotland with BKD more rigorously controlled in salmon farming regions. Undetected subclinical infection is an important uncertainty in BKD management. An existing epidemiological model has been adapted to assess the impact of external infection on the salmon compartment. The model has been modified to split infection pressure between exogenous (input) and endogenous (internal) spread, keeping the total infection pressure required to maintain the observed prevalence at steady-state. Model results suggest that BKD's persistence in salmon is dependent on input from trout or other external sources, even though these inputs are rare. Inputs also make it impossible to remove all risk of BKD occurring in salmon, although exclusion of BKD may be possible for long periods. BKD prevalence in salmon is likely to change in response to changes in other compartments, but any response will be less than linear. By far the largest predicted increases in BKD in Scottish salmon would occur if movement controls were to be lifted from salmon farms that are known to be infected. The model should have wider application to situations where pathogens are being eradicated from, or controlled within, specific compartments of multi-compartment systems and can be used to guide the appropriate balance between intensity of boundary and internal controls to minimise disease occurrence in control compartments.