Table 2. Future directions under which our model may be expanded to better capture the complexity of biological systems or to make further predictions about the effects of EDCs on systems of sexual selection.

Effect/Factor	Details	Model Predictions	Some Key References
Mate Choice	Female choice may be separated in to three ‘components’, 1) responsiveness,2) discrimination and 3) a preference function. We only varied responsiveness.Future models may address the other components, or allow females to vary onecomponent in relation to available males.	Flexible mate choice wouldincrease population resilienceto extinction resulting from awallflower effect.	[19], [61]
Heritability	We assume preference is heritable and genetically controlled. Studies havefound low heritability of preference in guppies, suggesting an environmentalcomponent in mate choice. In further models, past experiences or the suiteof available males may influence mate choice.	Environmental component tomate choice would decreasethe signal disruption associatedextinction the and loss ofpreference we report.	[19], [66]
Preference Costs	In our model, ‘opportunity cost’ of not mating is the only cost to preference.Other models have also imposed direct costs on female fitness; e.g. decreasedfecundity. Models of disrupted sexual selection may wish to include such costs.	Direct costs may result in lowerpreference levels. Weakerpreferences are less likely tobe susceptible to disruptedmating signals.	[44], [61], [63], [67]
Good-Genes and/ or Sexy-Sons	In our model, breeding values of males result from improved offspring survivaland a ‘sexy-sons’ mechanism. Future work may model disruption of sexualselection operating purely via sexy sons. Wherein, breeding values associatedwith signalling males are purely the result of the procreation of ‘sexy’ sons.	Signal was maintained duringdisruption via its associationwith viability. Therefore, undera sexy-son’s-only mechanism,signal may be lost completely.	[44], [68]–[72]
Genetic Control of Signal and Predation	We assume that signals occur via an innate mechanism. In guppies, genes for colourare maintained by interacting sex linkage and predation. Models may also wish toinclude genetic suppression/promotion of signal and predation.	In the presence of disruptionand predation, alleles fornon-expression of colourmay proliferate becauseenvironmental disruptionsmarginalise the advantage ofcolour expression.	[59], [73]
Multiple Sexually Selected Traits	In many species, females select a male based on multiple traits; e.g. colorationand display rate in guppies. Our models contain a single trait. Further modelsof disrupted sexual selection may fit multiple traits under sexual selection.	One/multiple genes underlinkage may control preferencefor several traits. Hence,preference for a disrupted signalmay be maintained by selectionon preference for undisruptedtraits.	[12], [19], [21], [74]
The Lek Paradox	Here, high mutation rates counter the erosive effect of preference on geneticdiversity. Models may wish to explore the relative importance of other mechanismsthat maintain genetic variation with regard to disruption of mating signals. Forexample, multiple interacting loci may control preference/survival. Withadditional loci, the mutation rate at a single locus can be reduced, yet geneticvariation still maintained. Note the same outcome would be observedunder assumptions of polyploidy, rather than diploidy.	Unknown.	[43], [45], [46], [75]

As an example, consideration is given to processes operating guppies.