7 Risk Assessment
This chapter describes how ERAHUMED assesses the toxicological risks associated with pesticide exposure in the Albufera Natural Park. Building on the hydrological and exposure simulations described in previous chapters, the risk assessment module estimates potential impacts on aquatic organisms and the broader ecosystem.
ERAHUMED adopts the Species Sensitivity Distribution (SSD) methodology—a widely used approach in ecological risk assessment (Posthuma, Suter II, and Traas 2002). This method enables the quantification of the Potentially Affected Fraction (PAF) of species at a given chemical concentration, offering a probabilistic estimate of ecosystem-level effects. By linking simulated concentrations to biological sensitivity, this layer identifies high-risk scenarios and supports decision-making aimed at mitigating adverse outcomes.
As ERAHUMED applies predefined SSD parameters rather than computing them from toxicity data, the present chapter focuses on their application to simulation results. For details on the derivation and validation of SSDs, we refer the reader to (Amador Crespo et al. 2024).
The risk estimation approach follows the methodology described in (Amador Crespo et al. 2024), to which we refer the reader for a detailed exposition. Here, we summarize the elements that are directly relevant to ERAHUMED’s implementation.
SSDs are used to translate the pesticide concentrations computed in the Exposure module (Chapter 6) into daily values of PAF. Chemicals are grouped according to their TMoA classification (encoded by the tmoa_id chemical parameter, cf. Table 3.2), and the effects of different TMoAs are assumed to be independent.
Two distinct SSDs are provided for each chemical:
- Acute risk is evaluated using daily concentrations.
- Chronic risk is evaluated using a 21-day rolling average of concentrations, reflecting longer-term exposure effects.
The SSDs are assumed to follow a log-normal distribution and are thus parametrized in terms of their mean and standard deviations (ssd_acute_mu/sigma and ssd_chronic_mu/sigma parameters in Table 3.2).
Risk estimates are computed both at the level of individual chemicals and at the level of their associated TMoA. This dual perspective enables the identification of both compound-specific risks and mode-of-action-level cumulative effects. To obtain the overall risk, ERAHUMED combines the risks associated with each TMoA under the assumption of statistical independence, following the approach described in (Amador Crespo et al. 2024). Specifically, the overall PAF is computed as:
\[ \text{msPAF} = 1 - \prod _i (1-\text{PAF}_i) \tag{7.1}\]
where \(\text{PAF}_i\) is the risk associated with TMoA \(i\).
In addition to PAFs, ERAHUMED also computes Risk Quotients (RQs) by comparing simulated concentrations to the 5th percentile of the SSD (HC05), using the conventional ratio:
\[ \text{RQ} = \frac{\text{Concentration}}{\text{HC05}} \tag{7.2}\]
Although the HC05 is derived from the same SSDs used for PAF calculation, the Risk Quotient (RQ) itself does not have a probabilistic interpretation. Instead, it serves as a simple descriptive indicator of potential concern, offering an alternative, threshold-based perspective on risk, complementary to the PAF.