Spermatozoa exhibit a high degree of variation in morphology and velocity among animals [1]. Several factors have been suggested to contribute to this variation, such as fertilization modes (i.e. external or internal fertilization [2, 3]), phylogeny [4], post-mating sexual selection, including sperm competition [5], and cryptic female choice [1]. The evolution of fertilization modes and sperm competition are thought to be evolutionary forces that generate sperm diversity [2, 6, 7]. However, these factors confound each other if phylogenetically distant species are compared, and the relationship between them and sperm characteristics is obscure in many cases [1].
Generally, sperm of internally fertilizing species are believed to be longer than those of externally fertilizing species [8–10]. In addition, the relationship between the sperm components (i.e. sperm head, midpiece, and flagellum) and sperm velocity differed between species with external and internal fertilization [11]. Sperm motility of internal fertilizers in fish is similar to that of amniotic tetrapods, which are also internal fertilizers [12], but there are significant differences in motility between sperm of externally fertilizing fish and amphibians [13]. However, studies exploring the influence of fertilization modes are not as robust, possibly due to the small number of taxa that contain both externally and internally fertilizing species, and the evidence that fertilization modes affect sperm characteristics is still limited [8, 14]. In addition, the comparison of phylogenetically distant species would involve a high risk of confounding factors other than the fertilization mode [15].
Over the past 50 years, many studies on sperm competition have been documented and shown that sperm competition affects sperm morphology and speed within intra- and interspecies [1, 16]. Multiple paternity rates caused by different mating systems generate different levels of sperm competition [17–19]: low levels in monogamous species which present a dyadic relationship, and high levels in promiscuous or polyandrous species which mate with multiple partners [20, 21]. The presence of sneaker males is also a factor that increases the level of sperm competition [22, 23]. Comparative studies across taxa have shown that species with a high risk of sperm competition have longer sperm with faster swimming capabilities (e.g. fish [24], mammals [25], birds [26–28]). However, a number of studies have also shown that sperm competition level is negatively associated or not associated with sperm total length and velocity [29–32]; thus, the theory of sperm competition is a topic worthy of discussion.
In fish, while most teleost species exhibit external fertilization, a few taxa have internal fertilization with copulatory behaviour, which may have evolved multiple times [33, 34]. Fish with internal fertilization tend to possess elongated sperm heads [2, 29, 35]. Furthermore, several groups possess a unique fertilization mode called internal gametic association (IGA) with copulation [36]. The IGA process is similar to internal fertilization; however, sperm-egg fusion occurs after eggs are released into seawater. Like the internal fertilizing species, for example, some of the marine Cottidae [14, 37, 38], the sperm head of IGA species is longer than that of species with external fertilization. However, several internal fertilizing fish have oval-headed sperm [15, 39, 40], and several externally fertilizing fish have sperm with an elongated head [41–43]. Therefore, the evidence of how the fertilization mode affects sperm characteristics is still unclear, even in fish, and a comparison between close relatives is required to elucidate the effect of fertilization mode.
Furthermore, fish exhibit various mating systems and a wide range of sperm competition levels [30]. A recent meta-analysis indicated that the multiple paternity rates of internal fertilizers are higher than those of external fertilizers, suggesting that internal fertilizers can also possess high levels of sperm competition [44]. In addition, a high population density of fish increases sperm competition levels [45, 46]. High levels of sperm competition increase sperm total length and velocity in Tanganyikan cichlid fishes [24, 47]. At the species level, sneaker males have faster sperm than territorial males in several fish (e.g. [48–51]). Furthermore, one study reported that different tactics (i.e. sneaker and courting) affect sperm midpiece morphology in internally fertilized Xiphophorus nigrensis [52]. Thus, bony fish are an optimal group for elucidating sperm evolution associated with different fertilization modes and sperm competition.
In this study, we investigated the effect of fertilization modes on sperm characteristics, taking into account the different mating systems and sperm competition levels in marine fish. A comparative study within the family or genus is essential for considering phylogenetic effects (e.g. [53, 54]). However, few taxa have both external and internal fertilization [15]. Thus, we compared pairs of closely related taxa from different families [47]. Three groups from a broad taxonomic range in marine fish (Group I; Pomacentridae vs Embiotocidae, Group II; Sebastinae vs Scorpaeninae, and Group III; Aulorhynchidae vs Hypoptychidae) were used in this study to assess the relationship between fertilization modes or sperm competition and sperm characteristics (Fig. 1, Additional file 1: Table S1). As data from disparate sources may be different in quality and may be erroneous [55], we caught all fish in the field (Additional file 1: Table S1) and used for the following analyses. According to the previous studies, we determined fertilization mode and estimated sperm competition levels of each species based on not only mating system, presence or absence of sneakers, the density at mating sites, and frequency of multiple mating, but also relative testes mass (RTM) we calculated (Additional file 1: Table 1). We measured sperm morphology, velocity, and motility in different solutions using six species with external fertilization and four species with internal fertilization, including species with IGA (Additional file 2: Table S2). We also examined the male intromittent organs to determine their relationship to the fertilization mode (Additional file 3: Table S3), as intromittent organs are typically projected in copulatory species (e.g. mammals, cartilaginous fish, and reptiles). We compared these characteristics within closely related species to control for phylogenetic effects, and analysed the relationships between RTM and sperm components.
Table 1 Family and species names of fish used in this study, their fertilization modes, mating systems, estimation of relative sperm competition level (ESC), and relative testes mass (RTM).
Group
|
Family (subfamily)
|
Species
|
Fertilization mode
|
Mating System
|
Sneaker
|
Density of mating site
|
Frequency of Multiple mating
|
ESC
|
Reference
|
RTM
|
I
|
Pomacentridae
|
Amphiprion clarkii
|
External
|
Monogamy
|
Absent
|
Low
|
Low (dyadic)
|
Low
|
[97–99]
|
-0.44 ± 0.05
|
|
|
Chromis notata
|
External
|
MTV-polygamy2
|
Present?
|
High
|
Low (dyadic)
|
Medium
|
[100–102]
|
0.43 ± 0.27
|
|
|
Pomacentrus nagasakiensis
|
External
|
MTV-polygamy
|
Present?
|
High
|
Low (dyadic)
|
Medium
|
[80]
|
0.33 ± 0.09
|
|
Embiotocidae
|
Ditrema temmincki temmincki
|
Internal
|
MTV-polygamy
|
Present?
|
High
|
High (multiple paternity)
|
High
|
[103–105].
|
0.16 ± 0.26
|
II
|
Scorpaenidae (Scorpaeninae)
|
Dendrochirus zebra
|
External
|
Promiscuity
|
Absent
|
Low
|
Low (dyadic)
|
Low5
|
[106]
|
-0.36 ± 0.13
|
|
Scorpaenidae (Sebastinae)
|
Sebastes cheni
|
Internal
|
MTV-polygamy3?
|
Present
|
High
|
High (multiple paternity)3
|
High
|
[107–110]
|
-0.19 ± 0.37
|
|
|
Sebastiscus marmoratus
|
Internal
|
MTV-polygamy (Promiscuity)
|
Absent
|
Low
|
Low (dyadic)
|
Low
|
[111, 112]
|
-0.26 ± 0.13
|
III
|
Aulorhynchidae
|
Aulorhynchus flavidus
|
External
|
MTV-polygamy
|
Absent4
|
High
|
No data
|
Medium
|
[65, 113]
|
0.08
|
|
Hypoptychidae
|
Hypoptychus dybowskii
|
External
|
MTV-polygamy
|
Present
|
High
|
High (multiple)
|
High
|
[64, 114]
|
0.19 ± 0.1
|
|
Aulorhynchidae
|
Aulichthys japonicus
|
IGA1
|
MTV-polygamy
|
Absent
|
High
|
High (multiple)
|
Medium
|
[89, 115]
|
0.06 ± 0.11
|
1Internal gametic association [88]. 2Male-territory-visiting polygamy. 3Closely related species S. inermis and S. atrovirens showed male-territory-visiting polygamy and multiple paternity. 4Aquarium observations found the appearance of sneaker male [116], but we referred to the field studies to arrange the condition among species, as the other studies were conducted in the field. 5The mating system of D. zebra is reported to be promiscuous, but mating occurs between dyadic relationships (i.e. monogamous). Therefore, the level of sperm competition is predicted to be low.