DISCUSSION
Using a parentage-based tagging approach we have provided the first description of the genetic mating system for Yellowstone Cutthroat Trout, a species of conservation concern and recreational angling importance. We documented multiple forms of mating among adults including monogamy, polygyny, and polyandry, and offspring production was found to be heavily right skewed with few adults producing many and most producing few or no offspring. The variance in relative reproductive success was higher among males than females as measured both by the number of mates acquired and number of offspring produced. We observed a positive correlation between the number of mates acquired and relative number of offspring produced; however, the strength of this correlation was lower for females than for males. On average, male adults had a greater number of mates and produced a higher number of offspring than their female counterparts. We failed to detect a signal of inbreeding avoidance, although the group of parents that produced progeny were on average slightly less related than adults that did not produce progeny. Both arrival date and total length were identified as significant predictors of relative reproductive success for both female and male Yellowstone Cutthroat Trout. Lastly, we identified that the effective number of breeders was lower than effective populations and these two parameter estimates had non-overlapping confidence intervals. Combined, the current study of Yellowstone Cutthroat Trout mating system helps to fill several gaps in knowledge regarding life history and these data may be useful moving forward for conservation and management efforts.
Social mating systems and spawning behaviors are well studied within salmonids (see Esteve 2005 for review); however, there remain taxa for which there is little or no information, including Cutthroat Trout. Several traits observed in our study of Yellowstone Cutthroat Trout were in line with patterns described for other members of the Salmonidae family. For example, we found evidence of both polygamous and monogamous matings, an observation that has been noted for Steelhead (Seamons et al. 2004a), Brown Trout (Salmo trutta , Serbezov et al. 2010), and Brook Trout (Kanno et al. 2011). In Chinook Salmon (O. tshawytscha ) and Atlantic Salmon, there is evidence of polyandry but not polygyny (Bentzen et al. 2001; Fleming 1998). Theoretical expectations are that polygamy would be a common occurrence among salmonids because one sex is free of parental care (males) and intrasexual selection in the form of male-male competition is strong (de Gaudemar et al. 1998). Thus, it was not necessarily unexpected for us to observe polygamy and monogamy in Yellowstone Cutthroat Trout. From a management perspective, these findings are important as multiple paternity (or polyandry) has been shown to increase effective population size (N e) relative to systems with strict monogamy (Pearse and Anderson 2010). It is worth noting that the parent pair-offspring relationships identified via PBT which were used to infer mating systems displayed a high level of concordance with results from sibship analysis. Our findings mirror previous efforts to resolve full-sibling families via sibship methods (Ackerman et al. 2016), and highlight that it is possible to accurately estimate sibship andN b without parental genotypes (Waples and Waples 2011). The ability to infer family structure using only offspring from a single generation is significant as it can be applied in scenarios where sampling both parents and their offspring is either logistically difficult or not possible.
Another trait observed in Yellowstone Cutthroat Trout that mirrors observations from other salmonids was protandry, whereby males arrived earlier than females to spawning grounds. Although male Yellowstone Cutthroat Trout arrived earlier to spawning grounds than females (Figure 4A ), the cumulative number of offspring assigned to males was highest among those that arrived midseason (Figure 4B ). Protandry has been widely documented in Pacific salmon (Morbey 2000) and Steelhead (McMillan et al. 2007, although see Seamons et al. 2004b) and it has been hypothesized that males arrive early to establish dominant access to spawning females (e.g., Healey and Prince 1998). The ability for males to establish dominance over competing males can be influenced by body size (Quinn & Foote1994; Fleming 1998; Dickerson et al. 2002), prior residence (Foote 1990), and the ratio of sexually active females to males (e.g., operational sex ratio; OSR). Results from our generalized linear models identified both arrival date and total length as significant predictors of relative reproductive success on an individual basis. In other words, while the bulk of offspring were collectively assigned to males that arrived mid-season (Figure 4B ), early arriving males that were larger in size had the highest individual relative offspring production (Figure 5 ). There are numerous examples in the salmon literature of reproductive success being positively related to body size (e.g., Schroeder 1981; Fleming and Gross 1992; Quinn and Foote 1994; Fleming 1996; Seamons et al. 2004b; Berejikian et al. 2009; Anderson et al. 2012); however, there are other instances where the relationship is either weak, non-existent, or affected by other forms of selection (e.g., Seamons et al. 2004b; Dickerson et al. 2005; Seamons et al. 2007). Whereas different life history strategies (e.g., precocial males and jacks), behavioral strategies (e.g., sneaker or satellite males), and forms of selection (e.g., predation by terrestrial animals) may lessen the strength of the relationship between arrival time and body size on reproductive success (Dickerson et al. 2002; Dickerson et al. 2005; MacMillan et al. 2007), in our system the largest, earliest arriving males experienced the highest levels of relative reproductive success.
Pairwise genetic relatedness among adult Yellowstone Cutthroat Trout that successfully mated was not significantly different relative to adults that did not successfully reproduce, and this trend has been noted for other salmonids. Several studies have identified mate choice among salmonids, specifically at loci associated with the major histocompatibility class (MHC). In particular, both Atlantic Salmon and Chinook Salmon have been shown to select mates that increase heterozygosity at the MHC loci, but not at putatively neutral loci (Landry et al. 2001; Neff et al. 2008). Because we only surveyed Yellowstone Cutthroat Trout at putatively neutral SNP loci we were unable to test for mate association at MHC, but nonetheless, we showed patterns of genetic relatedness among parents supported a random mating scheme.
Lastly, we generated estimates of N b andN e and showed that theN b/N e was less than one for Yellowstone Cutthroat Trout. Ratios ofN b/N e vary across freshwater fish (Brown Trout: 0.48, Lake Trout, Salvelinus namaycush : 1.212, Razorback Sucker, Xyrauchen texanus : 1.004) and simple life history characteristics such as age at maturity and adult lifespan have been shown to explain up to 2/3 of the variation in observed N b/N e ratios (Waples et al. 2013). In the case of Yellowstone Cutthroat Trout in the South Fork Snake River, there are two pools of potential parents capable of contributing to annual reproduction: tributary residents and mainstem adults that perform fluvial spawning migrations (Thurow et al. 1988). That we observed N b/N eratios less than one could be explained by only a subset of the potential parent population contributing to reproduction (i.e., reduced resident or fluvial contributions), an unusually large variance in reproductive success, or a highly skewed operational sex ratio that occurred in 2016 relative to other spawn years. Because we only have data for a singular spawn year it is difficult to identify specific mechanisms to explain our observedN b/N e ratio, but nonetheless, both these values are sufficiently high to suggest that Yellowstone Cutthroat Trout in Burns Creek represent a genetically stable and diverse population.
In conclusion, we have described the mating systems of a species of conservation concern and identified important predictors of relative reproductive success. Moving forward, these data can serve as a baseline for future monitoring efforts and a template for investigations into the reproductive ecology of other freshwater salmonids. Additionally, the observation that the largest adults were responsible for producing a disproportionate amount of offspring may serve as a basis for angling restrictions. For example, Gwinn et al. (2015) identified that protection of the largest adults in the population via harvest slots (i.e., harvest is restricted to fish of intermediate length) consistently produced greater numbers of fish harvested and greater catches of trophy fish while conserving reproductive biomass and a more natural population age-structure. Throughout much of their range in Idaho (including the South Fork Snake River and its tributaries), the harvest of Yellowstone Cutthroat Trout is prohibited, but in areas where harvest is still allowed or is being considered, data generated here may be of direct relevance.