Vol. 556: 1–16, 2016 MARINE ECOLOGY PROGRESS SERIES Published September 8 doi: 10.3354/meps11843 Mar Ecol Prog Ser OPENPEN FEATURE ARTICLE ACCESSCCESS Temporal variation of Skeletonema community composition from a long-term time series in Narragansett Bay identified using high-throughput DNA sequencing Kelly L. Canesi1,2, Tatiana A. Rynearson1,* 1Graduate School of Oceanography, University of Rhode Island, Narragansett, RI 02882, USA 2Present Address: Gordon and Betty Moore Foundation, Palo Alto, CA 94304, USA ABSTRACT: Phytoplankton species cannot always be identified by their morphology using light microscopy, which makes inferring the ecological and biogeochem- ical importance of individual species a difficult task. Here, a combination of microscopy and high-through- put DNA sequencing was used to examine morphologi- cally cryptic and pseudo-cryptic species in the diatom genus Skeletonema from the Long-Term Plankton Time Series in Narragansett Bay (NBay), where Skele- tonema is ecologically important, comprising up to 99% of microplankton cells in surface waters. The 28S rDNA from mock phytoplankton communities compris- ing known species was amplified and sequenced using newly developed Skeletonema-specific primers. The relative abundances of species in the sequence data did not match expected abundances, suggesting that 28S copy number can vary greatly, even among closely related diatom species. The 28S rDNA was also ampli- fied from 75 field samples collected from 2008 to 2013. High-throughout DNA sequencing revealed seasonal varia- A total of 7 Skeletonema species were identified, in- tion (2008–2013) in morphologically cryptic species diversity cluding 5 newly detected species from NBay. Skele- (line, 4-point moving average) in the ecologically important diatom genus (percent composition among tonema species composition was highly seasonal and Skeletonema species shown in different colors). significantly correlated with water temperature. Winter− spring and summer−autumn communities were signifi- Image: T. A. Rynearson cantly different and characterized by low and high spe- cies richness, respectively. Species abundance during winter−spring was quantified by combining sequence KEY WORDS: Phytoplankton · Diatom · Community data with light microscopy counts, revealing Skele- composition · Diversity · Cryptic species · Time series tonema marinoi as the numerically dominant species during the winter−spring bloom. Seasonal variation in Skeletonema composition suggests that, although mor- INTRODUCTION phologically similar, species in this genus are likely adapted to different environmental conditions, raising the possibility that species composition of this important Diatoms are responsible for generating approxi- bloom-forming genus may shift as water temperatures mately 20% of global primary productivity (Nelson et in NBay increase due to anthropogenic influences. al. 1995, Field et al. 1998, Mann 1999); yet, estimates of the total number of species are rough and range be- © The authors 2016. Open Access under Creative Commons by *Corresponding author: [email protected] Attribution Licence. Use, distribution and reproduction are un - restricted. Authors and original publication must be credited. Publisher: Inter-Research · www.int-res.com 2 Mar Ecol Prog Ser 556: 1–16, 2016 tween 12 000 and 200 000 (Guiry 2012, Mann & Van- varia tion observed among cryptic species in the lab- ormelingen 2013). Recent taxonomic efforts have de- oratory, coupled with the few field studies of spatial scribed many diatom species that are morphologically and temporal variation described above, suggest that cryptic (i.e. identical) or pseudo-cryptic (i.e. small variation in cryptic species composition has impor- structural differences), including those in several eco- tant ecological and biogeochemical implications. logically important diatom genera such as Cyclotella The diatom genus Skeletonema contains both (Beszteri et al. 2005), Pseudo-nitzschia (Amato et al. cryptic and pseudo-cryptic species, and has a large 2007, Lundholm et al. 2012), Skeletonema (e.g. Ko- influence on coastal marine ecosystems. Skele- oistra et al. 2008), Chaetoceros (e.g. Beszteri et al. tonema is cosmopolitan and is known to form large 2005, Amato et al. 2007, Kooistra et al. 2010, Lund- blooms in coastal regions across the globe (e.g. Cleve holm et al. 2012), Asterionellopsis (Kaczmarska et al. 1900, Hasle 1973, Karentz & Smayda 1984, Cloern 2014) and Leptocylindrus (Nanjappa et al. 2013). et al. 1985, Castillo et al. 1995, Huang et al. 2007, Because morphologically cryptic and pseudo-cryptic Borkman & Smayda 2009a), supporting food webs species are difficult or impossible to identify using and contributing significantly to carbon flux (Dea- routine methodology such as light microscopy, they son 1980, Biddanda & Benner 1997). Historically, a represent a major challenge in both documenting the single species, Skeletonema costatum, was thought taxonomic diversity of marine waters (Sims et al. to be the most ecologically important, but the genus 2006) and understanding the impact of species com- has since been subdivided into at least 10 marine position on marine ecosystem function and biogeo- species (Guillard et al. 1974, Medlin et al. 1991, Sarno chemical cycles. Molecular characterizations have et al. 2005, 2007, Zingone et al. 2005), leaving open provided insights into the ecology of cryptic and the question of how variation among species influ- pseudo-cryptic species (denoted hereafter as cryptic ences the ecology of coastal regions around the world. species), though studies of natural samples are lim- Here, we examined Skeletonema species composi- ited. For example, targeted molecular analyses re - tion from Narragansett Bay (NBay), RI, USA, where vealed that cryptic species in the genus Pseudo- Skeletonema has an enormous impact on phyto- nitzschia sampled from Puget Sound, Washington, plankton productivity. It is found in the bay through- varied in their relative abundances over space and out the year and can dominate phytoplankton blooms time, and shifts in species composition were corre- in every season (Borkman & Smayda 2009a). On lated with different environmental conditions (Hub- average, Skeletonema represents 49% of the phyto- bard et al. 2014). Similarly, cryptic species of Pseudo- plankton community (>~10 µm) in NBay and can nitzschia in the Gulf of Naples showed distinct comprise up to 99% during blooms (Windecker patterns of species occurrence over time (Ruggiero et 2010). Previous studies have identified a high degree al. 2015). In the diatom family Leptocylindraceae, a of both genetic and physiological variability in Skele- targeted analysis of small subunit rDNA revealed tonema sampled from NBay. For example, protein unique temporal and geographical distributions of variation among isolates revealed seasonal variation individual cryptic species sampled from the north- in the genetic composition of what was then con - eastern Atlantic and Mediterranean coasts (Nanjappa sidered to be S. costatum (Gallagher 1980). In light et al. 2014). On a broader geographic scale, a global of the newly identified species in Skeletonema, we biogeography of species in the genus Skeletonema, hypothesized that the genetic variation observed in assessed using the large subunit of the rDNA, re - NBay previously may have instead reflected the dif- vealed large-scale geographic distributions for some fering compositions and activities of multiple Skele- species and, for others, ranges that appeared restricted tonema species. To test this hypothesis we addressed by temperature (Kooistra et al. 2008). 3 questions in this study: (1) How many species of Field observations of species distributions have Skeletonema exist in NBay? (2) Does Skeletonema been supported by physiological studies showing community composition vary seasonally? (3) Do shifts different growth rates among cryptic species in in Skeletonema species composition correspond to response to environmental conditions such as tem- seasonal fluctuations in environmental variables? We perature and light (Kaeriyama et al. 2011), salinity predicted that multiple cryptic species would be (Jackson et al. 1992) and nutrients (Maldonado et al. identified in NBay, given that 2 had previously been 2002). Studies of Pseudo-nitzschia have also de - identified (Kooistra et al. 2008), and that there would scribed how cryptic species respond to unique envi- be a high correlation between environmental vari- ronmental cues that induce sexual reproduction ables and community composition over a multi-year (Quijano-Scheggia et al. 2009). The physiological time series. Canesi & Rynearson: Skeletonema community composition 3 MATERIALS AND METHODS brightwellii (isolated from Puget Sound on 12 May 2007), Heterocapsa triquetra (CCMP448), Thalassio - Sample collection sira pseudonana (CCMP1335) and T. rotula (CCMP - 3096). Each mock community comprised 4 × 106 cells NBay is a temperate estuary located on the east and was filtered and frozen as described in ‘Sample coast of the USA and has a mean depth of 9 m. Be- collection’. tween 2008 and 2013, surface water samples were collected weekly from NBay (41° 34.2’ N, 71° 23.4’ W) and 200 to 250 ml seawater were filtered in triplicate DNA extraction and amplification onto 25 mm diameter, 0.22 µm pore size filters (EMD Millipore), flash-frozen in liquid nitrogen and stored For both mock community and field samples, DNA at −80°C. These filtered field samples were collected was extracted