Introduction
Maternal effects can have a large impact on the fitness of offspring. In plants, maternal effects in seed traits (e.g., seed mass, germination time) and offspring fitness (e.g. growth rates) have been well documented (Donohue, 2009). Maternal age at reproduction is known to affect diapause (i.e., suspended development induced by unfavorable environmental conditions) in offspring of insects (Mousseau & Dingle, 1991), and in amphibians, maternal factors have well known effects in size and rates of development (Warne et al., 2013).
Maternal effects can be the result of the direct effects of the environment on epigenetic marks, genomic imprinting, or maternal provisioning (which is influenced by both environmental and genetic effects). For example, the environment experienced by the mother can affect the expression of genes involved in germination ofArabidopsis thaliana offspring (for review, see Donohue, 2009). Genomic imprinting is the epigenetic silencing (e.g., via cytosine methylation or chromatin-mediated processes) of one of the parental chromosomes, leaving only expression from the non-silenced chromosome (Alleman & Doctor, 2000). In the case of maternal effects, only the maternal chromosomes are expressed and this can be transmitted to one or more subsequent generations (Bischoff & Müller‐Schärer, 2010). Genomic imprinting has been observed in a few insect species, plants and placental mammals (for review, see Matsuura, 2020; Thamban et al., 2020), but not in egg-laying vertebrates such as birds, monotremes and reptiles by far (Killian et al., 2001; Renfree et al., 2013).
Maternal provisioning is the supply of nutrients, resources and hormones by the mother during seed or egg development (Videvall et al., 2016). For example, the amount of stored nutrient reserves in seeds can significantly influence early seedling growth and development (Slot et al., 2013). Maternal effects can also manifest via the seed coating (which is maternally produced), the endosperm (which is a triploid tissue with two-third of genotype from the maternal parent), and/or via direct maternal effects in dispersal (Donohue, 2009). For instance, flowering time in Campanula americana determines whether the progeny will germinate in autumn or spring (Galloway & Etterson, 2007). For many marine larvae, maternal provisioning of lipids is the major source of endogenous energy and this accounts for ~40% of the metabolic needs of coral larvae (Harii et al., 2010). Maternal provisioning is affected by both the genotype and the environmental conditions experienced by the mother. For example, maternal exposure to hormones can change egg and larval morphology of reef fishes (McCormick, 1999). Maternal effects due to provisioning generally decrease over time (Roach & Wulff, 1987), but can also persist through the entire life cycle of an organism.
When different genotypes are combined to produce F1 (i.e., first generation) hybrids, maternal effects can affect the phenotypes of F1 offspring. Hybridization is the crossing between separate species or between strains/lines/populations within a species. The phenotypes of the F1 offspring may be similar to that of their maternal parents (i.e., maternal effects), intermediate between the parents (i.e., additive effects), similar to that of the dominant parent (i.e., dominance), or different to both parents (i.e., over-dominance or under-dominance) (Chen, 2013; Li et al., 2008; Lippman & Zamir, 2007). For example, environmental conditions experienced by the mother can influence the expression of genes involved of germination in progeny (Donohue, 2009). However, hybrid gene expression studies often only involve hybrids of one direction (Videvall et al., 2016), and hence are unable to distinguish between dominance effects and maternal effects.
For corals, maternal effects in morphology (Willis et al., 2006), survival (Chan et al., 2018; Isomura et al., 2013) and thermal tolerance (Dixon et al., 2015) have been reported. Chan et al. (2018) showed that interspecific hybrids of the corals Acropora tenuis andAcropora loripes had similar survival and growth to their maternal purebreds, although they exceeded parental performances in some cases. The bacterial and microalgal endosymbiont (Symbiodiniaceae spp.) communities associated with these corals did not differ between the reciprocal hybrids and their maternal and paternal purebreds (Chan et al., 2019). Since these microorganisms carry vital functions to the coral hosts and can contribute to holobiont fitness differences (Blackall et al., 2015; Rosenberg et al., 2007), this finding suggests that the microbial communities were unlikely responsible for the observed holobiont fitness differences, and that these are likely underpinned by coral host genetic and/or non-genetic transgenerational factors.
The aim of this study was to test if the phenotypic differences in reciprocal F1 hybrids of the corals A. tenuis and A. loripes could be linked to patterns of host gene expression. Four offspring groups (i.e., reciprocal F1 hybrids and two parental purebreds) were previously produced via a laboratory cross of A. tenuis and A. loripes and were exposed to seven months of ambient or elevated temperature and p CO2conditions (Chan et al., 2018). Using samples from the same experiment, we tested for maternal effects in gene expression, as observed in hybrid survival and growth. In addition, gene expression was examined between temperature/p CO2 conditions within each offspring group.