Introduction
Parental effects occur when maternal, paternal or both parental phenotypes affect offspring phenotypes (Bonduriansky & Day 2018; Uller et al. 2013). Such effects occur in a wide range of taxa (Uller 2008) via different pre- and post-natal routes (e. g. microhabitat selection for eggs, reproductive investment, intrauterine environment, parental care). Parental experiences can affect offspring fitness (Burton & Metcalfe 2014), although are not necessarily adaptive (Bonduriansky & Day 2018). For example, maternal undernourishment is associated with the development of diabetes and obesity in the progeny (Hales & Barker 2001), while paternal undernutrition in mice results in altered glucose metabolism and growth in the offspring (Anderson et al.2006).
The knowledge about the molecular mechanisms of the pre-natal parental effects is still limited (Gluckmanet al. 2005; Jensen et al.2014). The transmission of some parental effects via germline has been related to genetic mechanisms, such as the association between the frequency of some deleterious mutations in sperm and increasing male’s age (Wyrobek et al. 2006). However, it is likely that non-genetic mechanisms also play a major role in parent-offspring information transfer (Danchin et al. 2011; Jablonka & Raz 2009), as genetic-based inheritance solely cannot fully explain the variation of offspring phenotypes (Danchin et al. 2011). Epigenetic modifications, such as DNA methylation, histone modifications and microRNAs, mediate rapid changes in transcription influenced by environmental changes (Richards 2006) that can affect phenotypes (Richardset al. 2017; Verhoeven et al. 2016). Among the epigenetic mechanisms, DNA methylation is the best characterized, being important on several biological processes, from genomic imprinting to cell differentiation (Jones 2012; Lea et al. 2017). DNA methylation on regulatory regions generally supresses gene expression (Moore et al. 2013), whereas methylation in gene bodies contributes to reducing transcriptional noise (Huh et al. 2013). Thus, differential methylation can affect gene expression and result in phenotypic plasticity (Baerwald et al. 2016; Herman & Sultan 2016). However, while the transmission of environmentally-induced epialleles via DNA methylation from parents to offspring has been identified in plants, whether epigenetic mechanisms can provide a heritable memory of environmental influence in animals remains controversial (Heard & Martienssen 2014), as well as the potential adaptive value of this type of transmission (Perez & Lehner 2019).
The parental rearing environment can induce phenotypic modifications during early development which can be long-lasting and potentially intergenerational (Burton & Metcalfe 2014). A well known example is the effect of structural environmental complexity on behaviour (Braithwaite & Salvanes 2005; Roberts et al.2011), physiology (Näslund et al.2013), cognitive capacity (Salvaneset al. 2013) and brain structure (Kihslinger et al. 2006) in fish. Physical structures are critical for most fish at different points of their life cycle (e. g. for spawning, sheltering, foraging), suggesting that structural complexity is an important ecological factor of their natural environment (Näslund & Johnsson 2016). Captive fish reared in enriched environments have shown increased survival in the wild compared to those reared in impoverished environments (D’Anna et al. 2012; Roberts et al. 2014), as well as enhanced cognitive capacity and behavioural flexibility (Salvanes et al. 2013; Spence et al. 2011; Strand et al. 2010). However, little is known about the molecular mechanisms underlying plastic responses to environmental enrichment, or whether these changes could be transmitted across generations (Näslundet al. 2012; Näslund & Johnsson 2016).
Kyrptolebias marmoratus (Poey 1880) is a predominantly self-fertilising fish living in mangrove forests in North and Central America (Tatarenkov et al. 2017), occupying a varied range of mangrove fossorial microhabitats influenced by periodical tide variation (Ellisonet al. 2012b). Its naturally inbred nature makes K. marmoratus populations particularly suited to assess the influence of the environment on behaviour (Ellisonet al. 2013; Ellison et al.2012b), phenotypic plasticity (Earleyet al. 2012) and epigenetics (Ellison et al. 2015). In their natural environment, the species inhabits inherently heterogenous mangrove habitats, with different selfing lineages coexisting in the same microhabitat (Ellison et al.2012b), and displays aggression towards conspecifics (Taylor 2000) that vary depending on kinship relationship (Edenbrow & Croft 2012; Ellison et al. 2013). These fish ermerse to forage or in response to intraspecific aggression or poor water quality (Turko et al.2011), suggesting that environmental complexity may play an important role on their ecology and behaviour.
We reared two generations of genetically-identical K. marmoratus in matched and mismatched environments with different levels of structural complexity to examine the intergenerational influence of environmental enrichment on individual physiology and behaviour, and the potential role of epigenetic mechanisms (brain DNA methylation) to mediate environmentally-induced parental effects.