INTRODUCTION
Cold-water corals are important ecosystem engineers that provide habitats to a diversity of species including invertebrates, vertebrates and commercially important fish and crustaceans (Roberts & Hirshfield Reference Roberts and Hirshfield2004; Roberts et al., Reference Roberts, Wheeler and Freiwald2006; Brancato et al., Reference Brancato, Bowlby, Hyland, Intelmann and Brenkman2007; Stone & Shotwell Reference Stone, Shotwell, Lumsden, Hourigan, Bruckner and Dorr2007; Kahng et al., Reference Kahng, Benayahu and Lasker2011; Braga-Henriques et al., Reference Braga-Henriques, Porteiro, Ribeiro, De Matos, Sampaio, Ocaña and Santos2013). Even though cold-water corals have greater species diversity than shallow water corals, their biology and ecology is considerably less understood (Roberts et al., Reference Roberts, Wheeler, Freiwald and Cairns2009). Cold-water corals were first discovered in the 1800s, yet general awareness of these organisms has only recently been brought to the forefront of conservation and ecological efforts. This effort is primarily due to increasing concern for cold-water corals and the species dependent on the habitats they form, as commercial fisheries exploit greater depths to maintain catch rates during the last two decades (Roberts et al., Reference Roberts, Wheeler and Freiwald2006). This is in addition to the growing concerns of climate change and its unequivocal effects on the world's oceans. The general consensus is that cold-water corals grow exceptionally slowly and thus a single trawling event could be fatal to a diverse deep-sea community and prior to recent advances in cold-water coral research and conservation, these organisms were principally out of sight and out of mind (Roberts et al., Reference Roberts, Wheeler and Freiwald2006).
In general, the reproductive biology of cold-water corals has been poorly studied (Roberts et al., Reference Roberts, Wheeler and Freiwald2006; Watling et al., Reference Watling, France, Pante and Simpson2011). There are over 3000 known species of cold-water corals (those occurring below 50 m) worldwide (Cairns, Reference Cairns2007), but of these, the reproductive biology is known for fewer than 40 species. Reproductive ecology seeks to identify sexuality (gonochoristic or hermaphroditic), reproductive mode (i.e. broadcast spawning, brooding larvae and budding), oocyte size and fecundity of a given species. Both population dynamics and recovery rates of benthic epifaunal species such as corals are dependent on several factors, one of which is reproductive ecology (Waller, Reference Waller, Freiwald and Roberts2005; Watling et al., Reference Watling, France, Pante and Simpson2011). Therefore understanding reproductive ecology is an imperative base framework from which the knowledge and understanding of population dynamics and recovery rates can branch.
The most common reproductive mode among cold-water coral species is gonochorism, though hermaphroditic species have been identified (Waller et al. Reference Waller, Tyler and Gage2005; Cairns, Reference Cairns2007; Orejas et al., Reference Orejas, Gili, López-González, Hasemann and Arntz2007; Kahng et al., Reference Kahng, Benayahu and Lasker2011). The two major subclasses of cold-water corals are Hexacorallia and Octocorallia, also known as hard and soft corals respectively. In octocorals there are two modes of sexual reproduction: broadcast spawning, where fertilization and development occur within the water column and secondly, brooding larvae internally or externally (Excoffon et al., Reference Excoffon, Navella, Acuña and Garese2011; Kahng et al., Reference Kahng, Benayahu and Lasker2011).
Reproduction is also a sensitive indicator of environmental stress (Van Veghel & Bak, Reference Van Veghel and Bak1994; Richmond, Reference Richmond and Birkeland1997; Zakai et al., Reference Zakai, Levy and Chadwick-Furman2000; Harrison & Ward, Reference Harrison and Ward2001; Bongiorni et al., Reference Bongiorni, Shafir, Angel and Rinkevich2003; Waller & Tyler, Reference Waller and Tyler2011). A current concern, aside from benthic disturbances, is the response of cold-water corals to anthropogenic and climate change impacts. It is now thought cold-water corals may be more vulnerable to climate change than shallow water corals (Kleypas et al., Reference Kleypas, Feely, Fabry, Langdon, Sabine and Robbins2006; Turley et al., Reference Turley, Roberts and Guinotte2007). However, due to the lack of physiological data, our ability to understand and predict the potential recovery of cold-water corals is limited.
In this study we examined the reproductive habits of eight Eastern Pacific cold-water corals, collected from Washington to Southern California coasts. The Eastern Pacific coast of North America has a narrow continental shelf as well as expansive latitude range which provides a unique habitat for cold-water coral species. The eight species of cold-water corals examined are: Primnoa notialis Cairns & Bayer, 2005, Primnoa pacifica Kinoshita, 1907, Swiftia beringi (Nutting, 1912), Swiftia kofoidi (Nutting, 1909), Swiftia pacifica (Nutting 1912), Swiftia simplex (Nutting, 1909), Swiftia spauldingi (Nutting, 1909) and Swiftia torreyi (Nutting, 1909). The objectives of this study were to enhance ecological understanding of cold-water octocorals by determining sexuality, reproductive mode, oocyte size, fecundity (reproductive potential) and larval formation. This study will establish a biological framework so that future studies can be used to compare fecundity or larval formations between cold-water octocoral species and examine seasonal influences. These data may ultimately aid scientists and conservation managers in protecting cold-water corals and the highly diverse communities they support.
MATERIALS AND METHODS
Previously collected and curated cold-water octocorals were acquired at the Smithsonian National Museum of Natural History in September 2012 (Table 1). Samples were preserved in 70% ethanol at room temperature until the start of histological processing. For each individual species three polyps were dissected out from the organism for histological fixation and ultimately fecundity analysis.
Table 1. List of species names, Smithsonian identification numbers and collection location of samples analysed. Summarized result for sex, fecundity (per polyp) and maximum oocyte size (feret diameter) (μm) for each sample.
The dissected tissue was decalcified with rapid decalcifier, 5-min immersion for Swiftia beringi, S. pacifica and S. simplex, and 20-min immersion for Primnoa pacifica and P. notialis. Tissue samples were then progressively dehydrated in a graded ethanol series. Swiftia kofoidi, S. spauldingi and S. torreyi samples were decalficied with VWR fixative-decalcifier for 8 h prior to serial dehydration. All species samples were then cleared with toluene in three 10-min immersions and then placed in a paraffin wax bath for approximately 24 h at 56°C.
Tissues were then embedded into paraffin wax blocks and left to cool overnight. Wax blocks were placed in a freezer 1 h prior to sectioning. Prior to serial sectioning, approximately 5–10 slides from each species were taken to measure nucleus diameter from 50–100 random oocytes. The average nucleus diameter from each species was then calculated and used to assign slide intervals for serial sectioning. All samples were serially sectioned at 5 μm thickness with intervals unique to each species oocyte nucleus diameter. Tissue sections were mounted onto glass slides, dried, and then stained with Mason's Trichrome.
Slides were examined using an Olympus (CX31) compound microscope with a Motic video camera attachment. Images were captured using Motic Image Plus and analysed using ImageJ software (NIH) to calculate oocyte diameter (a ‘feret’ diameter function was used to determine the area of the oocyte as if it were a perfect circle).
Approximately 100 random oocytes (or as close as possible) were measured using this method and oocyte size–frequency diagrams were constructed from results. Total fecundity was calculated by counting all previtellogenic and vitellogenic oocytes (that had a nucleus present) within the three polyps (Figure 1). To determine potential fecundity, the number of oocytes per polyp were averaged and multiplied by number of polyps counted within a 5 cm length of the sampled individual. For male specimens, spermatogenesis was determined by staging up to 100 random spermatocysts within a serially sectioned sample. Spermatocyst stages (I–IV) indicate maturity, stage IV being the most mature and stage I being immature (Figure 1). For example a coral with a high percentage of stage IV spermatocysts may be ready to spawn.
Fig. 1. Histological sections of gametogenesis in two Swiftia simplex individuals. (A) Oogenesis in USNM 57229 V: vitellogenic oocyte and pv: previtellogenic oocyte. (B) Spermatogenesis stages in USNM 57225. I: Stage one, II: Stage two, III: Stage three, and IV: Stage four.
RESULTS
All eight species of Eastern Pacific octocorals were gonochoric and no brooding or larvae were observed in this histological study, suggesting free spawning as the mode of reproduction (Table 1).
Primnoa notialis
Primnoa notialis individual (USNM 58171) oocyte diameters ranged from under 100–690 μm (max = 687.99 μm) (Figure 2). The mean oocyte diameter was 247.81 μm. Primnoa notialis had a total fecundity of 18 (±4.51 SE) oocytes per polyp (Figure 3). The majority (more than half) of the male spermatocytes were in stage I of spermatogenesis (Figure 4).
Fig. 2. Per cent frequency of oocyte diameters in all female individuals analysed. Left column down: Pacifica notialis individual USNM 58171. Serial sections taken at 100 μm intervals. Swiftia beringi individual USNM 1116868. Serial sections taken at 90 μm intervals. Swiftia kofoidi individual USNM 59817. Serial sections taken at 60 μm intervals. Middle column down: Swiftia simplex individual USNM 57229. Serial sections taken at 50 μm intervals. Mean per cent frequency of oocyte diameter of two Swiftia simplex individuals USNM 57226 and USNM 57230. Both were collected in September in 1964. Swiftia pacifica individual USNM 57221. Right column down: Mean per cent frequency of oocyte diameter of two Swiftia pacifica individuals 1 USNM 116867 and USNM 1116868. USNM 1116867 and USNM 1116868 individuals were collected off the Washington coast in June 2006. Serial sections taken at 70 μm intervals. Swiftia torreyi individual USNM 49538. Serial sections taken at 50 μm intervals. Swiftia torreyi individual USNM 1027078. Serial sections taken at 50 μm intervals. Error bars indicate standard error (SE). N = sample size and n = no. of oocytes measured. Black arrows indicate mean oocyte diameter.
Fig. 3. Mean fecundity per polyp of six species of Eastern Pacific cold-water coral (from largest to smallest fecundity). Error bars indicate standard error (SE).
Fig. 4. Percent of spermatocyst stages of male Primnoa notialis individual USNM 87627 (N = 1), Primnoa pacifica individuals USNM 57557 and USNM 1122474 (N = 2), Swiftia kofoidi individual USNM 57235 (N = 1), Swiftia simplex individuals USNM 57224, 57225, 57227 A, and 57227 B (N = 4), Swiftia spauldingi individual USNM 57162 (N = 1), and Swiftia torreyi individual USNM 56988 (N = 1).
Primnoa pacifica
No female individuals were found in this study. Mean per cent of spermatocyst stages of the two P. pacifica males show presence of all four stages of spermatogenesis. The majority of spermatocysts were in stage II of spermatogenesis.
Swiftia beringi
Swiftia beringi had the highest mean and maximum oocyte diameter of all the species in this study (Table 1 and Figure 2). In addition, S. beringi females had the third highest total fecundity: 13.6 (±2.85 SE) oocytes per polyp (Figure 3). No male individuals were found in this study.
Swiftia kofoidi
The mean oocyte diameter in females was approximately 435 μm (Figure 2). The female S. kofoidi individual had the lowest fecundity of all eight species, an average of 3 (±1.53 SE) oocytes per polyp (Figure 3). Swiftia kofoidi male individual USNM 57235 had all stages of spermatogenesis present (Figure 4). There is an even distribution of spermatocysts in stages I, II and III, with a slight decrease in frequency of stage IV spermatocysts.
Swiftia pacifica
Swiftia pacifica females show a difference in mean oocyte diameter between months (Figure 2). A lower average oocyte diameter (150 μm) was seen in the March individual (USNM 57221) compared with approximately double the oocyte diameter (282 μm) observed in the June S. pacifica samples. Averaged between the three females S. pacifica had the second lowest total fecundity of all species: 4.6 (±2.06 SE) oocytes per polyp (Figure 3). No male S. pacifica individuals were found in this study.
Swiftia simplex
Little difference was seen in mean oocyte diameter between the months of May (269 μm) and September (298 μm) in female individuals (Figure 2). Swiftia simplex females had the highest total fecundity at 42.53 (±9.82 SE) oocytes per polyp (Figure 3). Averaged over the four male individuals just over half of the spermatocysts were in the mature stages of III and IV spermatogenesis (Figure 4).
Swiftia spauldingi
No female individuals were found in this study. Only one male individual (USNM 57162) was present for histological analysis. All four stages of spermatogenesis were identified the S. spauldingi male (Figure 4). The majority of spermatocysts however, were in stage III of spermatogenesis.
Swiftia torreyi
A small variation in mean oocyte diameters was seen in S. torreyi females between the months of June (244 μm) and October (241 μm) (Figure 2). The total fecundity for the two female samples is 8 (±1.15 SE) (Figure 3). Over 70% of spermatocysts were in stages I and II in the single S. torreyi male sample (Figure 4)
DISCUSSION
All eight Eastern Pacific cold-water coral species are gonochoric and no brooded larvae were observed in any of these species, suggesting they likely use a free spawning reproductive mode (though this cannot be determined conclusively until a full seasonal study is conducted). This is consistent with similar cold-water coral reproductive studies (Waller & Tyler, Reference Waller, Tyler and Gage2002, Reference Waller and Tyler2011; Waller, Reference Waller, Freiwald and Roberts2005; Roberts et al., Reference Roberts, Wheeler and Freiwald2006; Mercier & Hamel, Reference Mercier and Hamel2011; Waller & Tyler, Reference Waller and Tyler2011; Watling et al., Reference Watling, France, Pante and Simpson2011; Brooke & Järnegren, Reference Brooke and Järnegren2013; Waller & Feehan, Reference Waller and Feehan2013; Baillon et al., Reference Baillon, Hamel, Wareham and Mercier2014; Waller et al., Reference Waller, Stone, Johnstone and Mondragon2014). Total fecundities ranged from 3 to 42.53 oocytes per polyp among the six species with female individuals. Within the genus Swiftia there was a large range in fecundities, with Swiftia simplex having the highest average fecundity of all females analysed: 42.53 (±9.82 SE). Of the five Swiftia species, Swiftia kofoidi and Swiftia pacifica had the lowest average fecundities: 3 (±1.53 SE) and 4.6 (±1.19 SE) oocytes per polyp, respectively. The large variation in fecundity within Swiftia species is most likely due to low sample sizes and missing months where fecundity could have increased prior to spawning. However, the collection date should also be noted, as extended preservation times can deteriorate tissue over time, and since the authors did not personally collect material, initial preservation may not have been ideal for histology. These factors could be seen in the tissue of a majority of the samples, and could attribute to a source of error within the data.
Though samples for true seasonal studies were not available, there are some indications of seasonality trends within species. Of the six species that had male samples four species, Primnoa pacifica, S. kofoidi, S. simplex and S. spauldingi had all four stages of spermatocytes present. There seems to be a trend of increasing frequencies of stage III and IV spermatocysts from March to July in P. pacifica. This seem to indicate an approaching spawning event post July but as both March and July have stage IV spermatocysts present they have the ability to spawn at both time periods. These observations are congruent to what was observed in the seasonal male spermatogenesis of P. pacifica from a fjord ecosystem in Alaska (Waller et al., Reference Waller, Stone, Johnstone and Mondragon2014). Variation in spermatocyst stages between this study and Waller et al. (Reference Waller, Stone, Johnstone and Mondragon2014) may be due to a variety of factors such as changes in latitude, depth, food availability and/or hydrodynamic flow between collection sites.
Swiftia simplex females may have different average oocytes sizes between May and September but oocyte diameter ranges were approximately the same (<99–700 μm). This may indicate a non-seasonal trend, as females always seem to have an available supply of vitellogenic oocytes. Swiftia pacifica females also indicate a difference in oocyte diameter between the month of March and June, with a higher frequency of previtellogenic oocytes in March and higher frequency of vitellogenic oocytes in June. However due to the low number of oocytes, lack of true seasonal samples and tissue deterioration there are insufficient data to determine a true seasonal trend.
In June, S. torreyi has a decreased fecundity but higher average oocyte diameter. In October, there is an increase in fecundity and a decrease in average oocyte size. This could potentially show a seasonal reproductive pattern, with oocytes being formed in October and spawned around June. These differences could also be due to the energy trade-offs between fecundity and oocyte diameter. Swiftia torreyi females may either invest their energy in a lot of oocytes with less lipids (yolk) storage or produce limited oocytes with a large lipid storage. Differences in collection depth could also be the explanation for the differences seem in S. torreyi females. There is nearly a 10-fold difference in collection depth between S. torreyi females. Individual USNM 1027078 was collected in 2003 at 1029 m where as individual USNM 49538 was collected in 1889 at 117 m.
This study, though small, increases knowledge on the basic life histories of cold-water coral species and provides a starting point for future reproductive studies on these habitat-forming organisms. Additional research is needed to understand species-specific reproductive seasonality as well as the reproductive potential between Eastern Pacific cold-water coral species.
ACKNOWLEDGEMENTS
We would like to thank Steve Cairns and William Moser of the Smithsonian Museum of Natural History for aid in obtaining samples for this project. We also thank Paul Tyler and one anonymous reviewer for their helpful comments in improving this manuscript.
FINANCIAL SUPPORT
Funding from this project came from a NOAA Deep Sea Coral Research and Technology Program and the West Coast and Polar Regions Undersea Research Center at the University of Alaska Fairbanks, part of NOAA's Office of Ocean Exploration and Research (UAF12-0070 to R. Waller). K.A. Feehan was partially supported by the Dearborn Scholarship, University of Maine.
