Egg production rates of the copepod Calanus marshallae in relation to seasonal and interannual variations in microplankton biomass and species composition in the coastal upwelling zone off Oregon, USA

Peterson, W. T., & Du, X. (2015). Egg production rates of the copepod Calanus marshallae in relation to seasonal and interannual variations in microplankton biomass and species composition in the coastal upwelling zone off Oregon, USA. Progress in Oceanography, 138, 32-44. doi:10.1016/j.pocean.2015.09.007

10.1016/j.pocean.2015.09.007 Elsevier

Version of Record http://cdss.library.oregonstate.edu/sa-termsofuse Progress in Oceanography 138 (2015) 32–44

Contents lists available at ScienceDirect

Progress in Oceanography

journal homepage: www.elsevier.com/locate/pocean

Egg production rates of the copepod Calanus marshallae in relation to seasonal and interannual variations in microplankton biomass and species composition in the coastal upwelling zone off Oregon, USA ⇑ William T. Peterson a, , Xiuning Du b a NOAA-Fisheries, Northwest Fisheries Science Center, Hatfield Marine Science Center, Newport, OR, United States b Cooperative Institute for Marine Resources Studies, Oregon State University, Hatfield Marine Science Center, Newport, OR, United States article info abstract

Article history: In this study, we assessed trophic interactions between microplankton and copepods by studying Received 15 May 2015 the functional response of egg production rates (EPR; eggs femaleÀ1 dayÀ1) of the copepod Calanus Received in revised form 1 August 2015 marshallae to variations in microplankton biomass, species composition and community structure. Accepted 17 September 2015 Female C. marshallae and phytoplankton water samples were collected biweekly at an inner-shelf station Available online 5 October 2015 off Newport, Oregon USA for four years, 2011–2014, during which a total of 1213 female C. marshallae were incubated in 63 experiments. On average, 80% of the females spawned with an overall mean EPR of 30.4. EPRs in spring (Apr–May, average of 40.2) were significantly higher than summer (Jun–Oct; 26.4). EPRs were intermediate in winter (Jan–Feb; 32.5). Interannually, EPRs were significantly higher in 2014 than 2011 and 2012. Total chlorophyll a (Chl a) concentration and diatom abundance both were significantly higher in summer while no seasonal differences were found in abundance of dinoflagellates, ciliates or Cryptophytes. Although total Chl a showed no interannual differences in bulk biomass of phytoplankton, community structure analysis indicated differences among years. More diverse diatom communities were observed in 2013 and 2014 compared to 2011 and 2012. Relationships between EPR and potential food variables (phytoplankton and ciliates) were significant by season: a hyperbolic functional response was found between EPR and total Chl a in winter–spring and summer, separately, and between EPR and ciliate abundance in winter–spring; a linear model fit best the functional response of EPR to diatom abundance in summer. The estimate of potential population recruitment rate (the number of females  EPR; eggs dayÀ1 mÀ2) was highest in spring (Apr–May), and annually was highest in 2013 (11,660), followed by 2011 (6209), 2012 (3172) and 2014 (1480). Our observations of in situ EPR were far higher than published laboratory rates of 23.5, calling into question our past laboratory studies that used mono-algae cultures as food for copepods. Moreover, an increasing number of studies (including this study) are showing an apparent greater importance of ciliates as a nutritious food source for Calanus species demonstrating the importance of the microbial loop to secondary production. Ó 2015 Published by Elsevier Ltd.

1. Introduction go through a brief period of diapause during autumn months after which adults re-appear in January. The awakening in winter is pre- Calanus marshallae, a sibling species of Calanus finmarchicus, sumably timed to take advantage of a winter bloom that occurs fre- Calanus sinicus and Calanus australis (Frost, 1974), occurs commonly quently in February (Feinberg et al., 2010). A cohort is born at this in continental shelf waters of the northern California Current from time, which becomes mature within 7–8 weeks (Peterson, 1986); January through August as well as in the coastal Gulf of Alaska and during the course of annual growing season, at least five genera- Bering Sea. This species was thought to be restricted to the North tions are produced (Peterson, 1980). This life history is in contrast Pacific Ocean and Bering Sea (Frost, 1974) as well as the Chukchi to that of C. marshallae populations in Dabob Bay, Washington Sea (Hopcroft et al., 2010); however, there is evidence that it has (Osgood and Frost, 1994) and the Bering Sea (Baier and Napp, invaded the Arctic Ocean, in waters near Spitsbergen (Melle and 2003; Smith and Vidal, 1984) which produce only one or two gen- Skjoldal, 1998). In the northern California Current, adults usually erations per year. The reason that C. marshallae can produce multi- ple generations in the northern California Current is almost certainly due to the availability of sufficient food resources ⇑ Corresponding author. Tel.: +1 5418670201. E-mail address: [email protected] (W.T. Peterson). throughout much of the year which in turn supports egg production http://dx.doi.org/10.1016/j.pocean.2015.09.007 0079-6611/Ó 2015 Published by Elsevier Ltd. W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 33 and subsequent development and growth of nauplii and during which females were fed a single diatom species Thalassiosira copepodites. weissflogii. We discuss reasons why the rates measured in the lab The coastal waters off Oregon are characterized by high phyto- were less than half of the rates from in situ incubation. Finally, we plankton production nearly year-round due to four independent discuss the potential reproduction rates (as the product of EPR on ‘‘primary production events” – a late winter bloom (in January/ a given day  abundance of females in situ on the same day) to February) in some years, a spring bloom (in April/May), multiple determine if there were seasonal and interannual differences in summer blooms during the upwelling season (June–September), potential ‘recruitment’. and a fall bloom after the end of the upwelling season (October/ November). These blooms have vastly different magnitudes, 2. Methods ranging from 1 to 2 lg Chl-a LÀ1 during the winter blooms to 15–30 lg Chl-a LÀ1 during summer blooms. The fall bloom, when Female C. marshallae for egg production experiments and it occurs, is on the order of 10 lg Chl-a LÀ1 (Du and Peterson, phytoplankton were collected biweekly from a station located 5 2014). The natural phytoplankton community contains a mixture nautical miles (9 km) from shore, in 62 m of water along the New- of taxa and sizes, from smaller nano- and micro-flagellates to lar- port Hydrographic line (station NH05, 124.2°W, 44.6°N) off central ger long-chain diatoms that exceed 1 mm in length. Moreover, Oregon, USA over a four-year period of 2011–2014. C. marshallae the dominant species are highly variable and ephemeral during were collected mostly at night from the upper 20 m of the water bloom events (Du and Peterson, 2014). Thus we were curious as column using a 60 cm diameter Bongo net (333 lm mesh). Live to the ability of C. marshallae to take advantage of the different samples were placed in a 20 L cooler filled with surface seawater prey types, bloom magnitude and sizes of the microplankton spe- and were transported to a temperature-controlled room in our cies as well as their functional response to phytoplankton biomass shore-based laboratory within 2–3 h of collection. To collect sample and species composition. for analysis of abundance of female C. marshallae, we used a ½ m Despite the broad distribution of C. marshallae around the mar- 202 lm mesh net hauled vertically from a few meters above the gin of the temperate/subarctic waters of the eastern North Pacific, sea floor to the sea surface. The contents of the cod-end were pre- Bering Sea and the Chukchi Sea, few studies ecological exist and served in formalin (final concentration of 5–10%). most are limited to the northern California Current. Past work on egg production rates (EPR, eggs femaleÀ1 dayÀ1)ofC. marshallae did not consider the potential effects from microplankton in terms 2.1. Copepod egg production experiments of species and community structure, except that significant rela- tionships were found between chlorophyll-a (Chl a) and EPR. Individual female C. marshallae were sorted with the aid of a Gómez-Gutiérrez and Peterson (1999) reported a mean EPR of stereomicroscope and a wide-bore pipette. Healthy-appearing 22.5 during Jul–Aug 1997. These rates were linearly correlated individuals were placed into 60 ml jars prefilled with sieved l with Chl a concentration. Similarly, Peterson et al. (2002) found a (200 m mesh) surface seawater. The goal was to pick 20–25 significant linear relationship between EPR and Chl a in June females for each experiment but this depended on the availability 1996. The range in values of EPR across continental shelf waters of female C. marshallae. Individual jars were incubated for 24 h in À ° was 20–30 when Chl a>5–10 lgL 1. Shaw et al. (2009) reported the dark at a constant temperature 10 ± 0.5 C, chosen because it an average rate of 20.3 during three cruises in summers 2005 and is the average year-around value observed for the upper 20 m of 2006. There is a somewhat controversial report of EPR by ‘‘C. mar- the water column over the inner and middle shelf off Oregon (win- ° ° shallae/glacialis” from the western Arctic Ocean by Plourde et al. ter is 10.3 C; summer 10.1 C, Peterson, unpublished data). Eggs (2005) who found average EPRs of 14 and 20 in spring and summer were first counted in the late morning, about 6–8 h after setting 2002, respectively, but Lane et al. (2008) who sampled copepods up each experiment. Individuals that had not spawned by that time on the same cruise as Plourde et al. (op.cit) claimed that there were were reexamined every 4–5 h until the end of the experiment. no C. marshallae in samples. Most females spawn eggs between 0400 and 1100 h (Peterson, This study considers several new topics related to EPR of C. mar- pers. obs. for C. marshallae), thus enumerating the eggs shortly shallae and the in situ food environment in the shelf waters off New- after they were spawned reduces the potential for egg cannibalism port, Oregon. The first topic considers reasons why a wide range of following recommendations of Runge and Roff (2000). Also EPRs were observed during incubation experiments over a four-year Peterson (1988) discussed why cannibalism is not likely to be a period and how that was affected by the potential microplankton problem in these types of experiments. Egg production rate (EPR; À1 À1 food resources. For example, in our experiments, individual female eggs female day ) is calculated from the total eggs in an exper- C. marshallae were found to be capable of producing up to 97 eggs iment divided by the number of females in that experiment, thus it per spawn with a maximum population average EPR of 68.5. These is an estimate of the ‘‘population” egg production rate, not an aver- rates are far higher than any of our previous studies (Peterson, 1988; age of only those females that spawned during each experiment. Gómez-Gutiérrez and Peterson, 1999; Peterson et al., 2002; Shaw et al., 2009). Meanwhile, the present study also observed very low 2.2. Phytoplankton sample analysis EPRs of <10. Thus we wanted to determine if such a wide range of EPR values were simply a matter of seasonal variations in On each cruise, surface water samples were collected for the microplankton biomass as opposed to variations in abundance of analyses of phytoplankton species and chlorophyll-a (Chl a). Total specific functional groups (diatoms, dinoflagellates, Cryptophytes Chl a and >5 lm size fraction subsamples (each of 100 ml) were fil- and ciliates), community structure and/or species diversity. Second, tered through GF/F filters and >5 lm polycarbonate filters, respec- on a few occasions, EPRs were depressed during certain diatom tively, extracted in 90% acetone for 24 h in the dark, and then blooms, noteworthy because past studies have indicated that some fluorescence was measured using a Turner 10-AU fluorometer. diatom blooms may be detrimental to copepod egg production and The concentration of Chl a was calculated following equations in egg hatching success (Paffenhofer, 2002; Paffenhofer et al., 2005). Strickland and Parsons (1972). The <5 lm Chl a concentration Thus we tried to determine if the reduced EPRs could be attributed was the difference between total Chl a and the >5 lm fraction. to specific diatom taxa. Third, we compared results of EPR based on Phytoplankton samples were fixed with acid Lugol’s solution in incubation of freshly-caught females (this study) with EPR 125 ml bottles (2% final concentration) immediately after collection. measured under controlled laboratory conditions (Peterson, 1988) To identify and enumerate species, a subsample of 50 ml was taken 34 W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 from the full sample and then settled in a 50 ml tissue culture flask were highly variable (Fig. 1A). The three lowest values were 3.9 (27 (Falcon) for at least 24 h prior to counting. The wide flat side of the Jul 2011), 4.0 (7 Mar 2012) and 7.7 (20 July 2012) and the three flask was placed on the stage of an inverted light microscope (Leica highest 68.5 (1 Apr 2014), 62.0 (29 Jul 2013) and 55.1 (16 Jan DM IRB); species identification and enumeration were carried out at 2014). The overall mean EPR was 30.4 ± 15 (standard error). It the magnifications of 200 or 400Â. The portion of the 50 ml sub- was unusual for all females to spawn in an experiment (only 12 sample that was examined varied with species density. To meet a out of 63 experiments). The average percentage of females which minimum total count of 500 cells, between 2 and 20 transects across spawned during the 24 h incubation was 80%, and by season, 84% the wide side of the flask were counted. For diatoms and dinoflagel- in winter–spring (Jan–May) and 75% in summer (Jun–Oct). The lates, species identification was made primarily following Horner average percentage of females that spawned was related to Chl a (2002). A group of smaller dinoflagellates (length <20–30 lm) were concentration: 81% when Chl a concentrations were high relatively consistent and abundant in our samples but difficult to be (5–30 lgLÀ1), 77% when less than 5 lg/L and 74% when below identified to species, thus we use ‘‘small dinoflagellates” to name 1 lgLÀ1. Because not all females spawned during a 24 h incuba- this group. Ciliates, a common group of microzooplankton in our tion, per capita rates (egg production rate calculated using only phytoplankton water samples, were included for this study and those females that spawned) were higher than the population rates are important because during the non-diatom bloom season they (mean 39.7 vs. 30.4, t = 5.77, p < 0.001). However, the two ‘rates’ comprised a significant fraction of total microplankton biomass. were significantly correlated (R2 = 0.87, not illustrated) suggesting For ciliates, identification to species level was not made, instead, results of comparing EPR rates in response to various food environ- the major subgroup of naked ciliates were recorded as size cate- ments are not dependent upon which method is used to calculate gories, small ciliates (length <30 lm) and large ciliates (>30 lm). EPR. We chose the ‘population’ rates as this method is most often Loricate ciliates were named as ‘‘tintinnids”. Cryptophytes were used by copepodologists. enumerated because they are morphologically easier to be recog- nized, whereas other nanoflagellates were not included for quantita- 3.2. Monthly EPR and Chl a concentration tive analysis due to the identification limitation of light microscopy. Cell abundance is expressed as cells per liter (cells LÀ1). The correspondence between monthly average EPR and Chl a was poor. Monthly average EPRs (Fig. 2) were higher in Jan–Feb and Apr–May than Aug-Oct, and lower in Mar, Jun–Jul and 2.3. Statistical analysis Nov–Dec. Total Chl a concentration ranged from 0.5 to 28.8 lgLÀ1 (Fig. 1B) and was much higher from July to September and lower in To compare interannual and seasonal variations in EPR and the other months (Fig. 2). The >5 lm Chl a size fraction followed microplankton biomass, Analysis of Variance (ANOVA) was imple- the same seasonal patterns as total Chl a, whereas monthly mented on EPR, abundance of four functional groups (diatom, averaged percentage of Chl a <5 lm size fraction was relatively dinoflagellate, ciliate and Cryptophyte) and Chl a concentration constant, ranging from 50.3% to 70.8%. data series, respectively. Tukey HSD pairwise comparisons were used to test for the differences in pairs. Phytoplankton abundance 3.3. Seasonal and interannual variations in EPR, phytoplankton and Chl a concentration were log10(x + 1) transformed before anal- biomass and functional groups yses so as to achieve normality and equal variance. The functional response of EPR (dependent variable) to the various measures of Fewer experiments were carried out in November (1 experi- microplankton biomass (taxa and chlorophyll) was characterized ment) and December (2 experiments) due to low numbers and/or by ordinary linear regression and by hyperbolic (Michaelis–Men- lack of females in the Bongo net samples. Although there was a ton) models. The equation for the hyperbolic model is y = ax/(b high number of females (average of 19) per experiment in March, + x), where a is the maximum EPR and b, the half-saturation con- the dramatically decreased monthly average EPR compared to stant value of the independent variable where one-half of the max- February and April (Fig. 2) was likely related to a small sample size imum value of the dependent variable (EPR) is reached. (2 experiments). We were concerned that the low sample size in Differences in phytoplankton community structure were exam- March, November and December might bias quantitative evalua- ined using hierarchical cluster analysis (with SIMPROF significance tion of EPR. Therefore, the three months with small sample sizes test at 0.05) and nonmetric multidimensional scaling (NMDS). were not included in the seasonal/interannual quantitative Two-way crossed ANOSIM (Analysis of Similarity) was used to test analyses. for differences in phytoplankton community structure by year and Interannual differences in EPR were significant (p = 0.023). season. The statistic R (scaled to a range of values from 0 to 1) in There were higher EPRs in 2014 than 2011 and 2012 (p = 0.05), ANOSIM is a measure of community separation between groups respectively (Table 1). Seasonal changes were also significant with 0 indicating no separation and 1 indicating complete separa- (p = 0.018) and showed higher EPRs in Apr–May (spring) than tion. Similarity of percentage (SIMPER) analysis provides a list of Jun–Oct (summer) but similar rates between winter (Jan–Feb) species in rank order of high to low species percent contribution and spring. to the overall similarity of a specific group. The length of the spe- In contrast, no significant interannual changes were shown in cies list presented in Results was determined when cumulative either total Chl a or size-fractionated Chl a concentrations individual species percent contribution added up to 80% of total. (p = 0.9), but seasonal variations were significant (p < 0.01) and The above multivariate analyses were conducted in PRIMER-E soft- the overall difference was the result of higher Chl a concentrations ware (Clarke and Gorley, 2006). Other statistical comparisons and in summer than winter. correlation analysis were performed in JMP 11 (SAS). Interannual variations were not consistent for the four func- tional groups (two-way ANOVA, year and season): diatoms were 3. Results significantly more abundant in 2013 than 2012, and Cryptophytes were more abundant in 2013 than in 2011, but interannual differ- 3.1. Egg production rate (EPR) ences were insignificant for dinoflagellates and ciliates. There were seasonal differences for diatoms only (p < 0.05): higher abundance A total of 1213 females were incubated in 63 experiments, an in summer than in spring and winter, and spring higher than win- average of 19 females per experiment. EPRs (eggs femaleÀ1 dayÀ1) ter. The abundances of dinoflagellates, ciliates and Cryptophytes W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 35

Fig. 1. Upper panel (A): Egg production rates (EPRs) during 2011 to 2014. The dotted line is the average EPR over all experiments. Lower panel (B): total Chl a concentration (lgLÀ1) on the same dates as EPR. Note the lack of correspondence between EPR and Chl a concentration.

in 2014 than 2011 and Cryptophytes were more abundant in 2013 than 2011 (p < 0.05).

3.4. Phytoplankton community structure and EPR

Cluster analysis identified five different communities (Fig. 3, upper panel), among which cluster ‘‘a” represents a community type characterized by the lowest average diatom abundance (174,040 cells LÀ1) and lowest species richness (13 diatom species). Most of the samples were from January, February and April with some from September and October. Cluster ‘‘c” represents a sum- mer community type with the highest average diatom abundance (1,740,800 cells LÀ1) and species richness (31 diatom species) and includes samples mostly from May to October. Notice that nearly all samples in ‘‘c” were from 2013 and 2014. Clusters ‘‘d” and ‘‘e” Fig. 2. Monthly average egg production rates (EPR, eggs femaleÀ1 dayÀ1) and Chl a include spring and fall samplings, respectively, with moderately size-fraction concentrations (lgLÀ1). The array of numbers on the top is the average number of female copepod Calanus marshallae used in all experiments in each month for 2011–2014. Table 1 Interannual variation in egg production rate (EPR ± SE, eggs femaleÀ1 dayÀ1) of the female copepod Calanus marshallae. March, November and December are not included were similar among seasons. One-way ANOVA for testing interan- for the comparisons since fewer experiments over the four years were done. Numbers in the brackets indicate sample size. nual differences in each of the phytoplankton functional groups was also run for the winter and spring seasons, separately. In win- Year Annual Jan–Feb Apr–May Jun–Oct ter, significantly higher abundances were found for ciliates in 2014 2011 26.1 ± 3.3(18) 29.7 ± 6.3(4) 28.6 ± 7.3(3) 24.2 ± 3.8(11) than 2011 and 2012 (p = 0.02), and for dinoflagellates in 2014 than 2012 26.3 ± 3.8(13) 23.3 ± 9.4(2) 38.0 ± 6.0(5) 18.7 ± 5.0(6) 2011 (p = 0.04), but no significant differences in diatoms and Cryp- 2013 35.3 ± 3.8(14) 26.5 ± 9.0(2) 44.7 ± 5.7(5) 31.1 ± 4.8(7) 2014 40.9 ± 4.1(12) 53.4 ± 10.3(2) 44.9 ± 6.5(5) 31.9 ± 6.5(5) tophytes among years. As for spring, diatoms and ciliates were similar among years; dinoflagellates were again more abundant Mean 31.8 ± 2.0(57) 32.5 ± 4.7(10) 40.2 ± 3.0(18) 26.4 ± 2.6(29) 36 W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44

20 clusters a b c 40 d e

60 Bray-Curtis similarity

80

100 4/8/2014 2/4/2014 4/1/2013 4/8/2011 5/8/2013 5/9/2011 9/1/2011 8/4/2011 4/1/2014 1/8/2012 4/7/2012 7/6/2011 8/7/2012 7/6/2014 7/9/2013 6/7/2013 5/6/2014 6/8/2011 10/4/2013 9/13/2011 4/27/2013 7/27/2011 7/22/2011 1/10/2011 1/25/2011 2/24/2012 1/16/2014 2/23/2014 2/11/2012 5/30/2013 9/29/2011 4/16/2013 4/14/2012 1/14/2013 2/10/2011 2/26/2011 8/17/2011 7/23/2013 7/29/2013 7/20/2012 7/22/2014 8/25/2013 6/19/2013 4/25/2014 5/19/2014 2/14/2013 6/11/2014 9/11/2013 9/26/2013 5/18/2011 4/17/2011 5/19/2012 10/7/2012 10/25/2012 10/26/2011 10/15/2011 10/22/2013 Samples

2D Stress: 0.2 a 2D Stress: 0.2 a a 3 a 4 d 2 d a 2 1 a a 2 c 4 c c a a 4 c a 3 3 c a 2 4 2 4 c c d 3 c c 2 1 a a a a 4 2 c c c a 3 c a 4 3 c aa a a 4 3 2 e a 4 3 a 1 4 4 1 c 3 a a 1 2 a a 1 b e a 3 3 1 1 e a a 1 2 e a 2 e 4 2 b 3 1

e 3

Fig. 3. Phytoplankton community clusters (a–e, p = 0.05) were identified through hierarchical cluster analysis (upper panel) and were also shown in a nonmetric multidimensional scaling plot (NMDS, left lower panel). Quartile EPR (right lower panel) was overlaid on the phytoplankton community pattern in NMDS. Numbers 1–4 stand for levels of EPR (eggs femaleÀ1 dayÀ1). 1: 3.9–17.6; 2: 17.6–29.4; 3: 29.4–40.2; 4: 40.2–68.5. Note that cluster ‘‘a” is a winter community, ‘‘c” a summer community, ‘‘d” and ‘‘e” seasonal transition communities and ‘‘b” abnormal summer community.

Table 2 Summary of two-way crossed ANOSIM comparisons of phytoplankton community structure. R value (range of 0–1) in the global and pairwise tests (shown as ‘‘R” and ‘‘p” values) indicates the relative degree of difference by year and season. W: winter, Sp: spring; Su: summer.

Source of variation (Global test) Year Season R = 0.408 p < 0.001 R = 0.352 p < 0.001 Pairwise test – between year 11 vs 12 11 vs 13 11 vs 14 12 vs 13 12 vs 14 13 vs 14 0.4, 0.007 0.521, <0.001 0.513, <0.001 0.43, 0.007 0.364, 0.009 0.11, 0.2 Pairwise test – between season W vs Sp W vs Su Sp vs Su 0.423, 0.005 0.23, 0.042 0.417, <0.001

high diatom abundance (721,380 and 583,970 cells LÀ1, respec- Dictyocha speculum and a dinoflagellate Protoperidinium depressum) tively) and species richness (27 and 17 diatom species, respec- in addition to the common ones in ‘‘c”. The ordination analysis tively). The three sampling dates in ‘‘d” were from 2013 to 2014 (Fig. 3, left lower panel) also showed a clear separation of and the six sampling dates in ‘‘e” were from 2011 to 2012. The phytoplankton community types. Cluster ‘‘a” and ‘‘c” had no over- cluster ‘‘b” included two sampling dates and was distinguished lap, ‘‘b” was associated with ‘‘c” whereas ‘‘d” and ‘‘e” appeared as from ‘‘c” by some unusually abundant species (a silicoflagellate transitional communities. W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 37

Changes of phytoplankton community structure were significant Table 3 across years (ANOSIM, R = 0.408, p < 0.001, Table 2) and seasons Egg production experiments conducted in winter (Jan–Feb). Chl a concentration indicates if a small winter bloom (>1 lgLÀ1) was present or not. Abundances (R = 0.352, p < 0.001). Pair-wise year comparisons were significant À (cells L 1) of s-dino (small dinoflagellates), ciliates and Cryptophytes are also shown except 2013 vs 2014 (p = 0.2) and all the pair-wise seasonal compar- along with egg production rates (EPR) eggs femaleÀ1 dayÀ1. isons were significant as well (e.g. winter vs spring, p = 0.005). Date Chl a s-dino Ciliates Cryptophytes EPR Although there were significantly different phytoplankton community clusters, the corresponding EPRs were similar 10-Jan–11 0.6 3430 343 30,188 19.9 22-Jan–11 0.8 3773 5489 66,207 26.6 (p = 0.6, one-way ANOVA). For example, the sampling dates of 10-Feb-11 0.5 3087 686 5832 35.6 two most contrasting clusters ‘‘a” (winter) and ‘‘c” (summer) had 26-Feb-11 1.7 8233 1029 52,828 36.7 very similar EPRs on average, 32.9 ± 3.1 and 32.6 ± 4.4, respec- 8-Jan–12 0.7 3773 5832 113,890 na tively. Furthermore, patterns of the five clusters shown in the ordi- 11-Feb-12 0.9 6861 na na 37.7 nation plot were not mirrored by the values of EPR represented as 24-Feb-12 0.7 4460 686 14,751 8.9 14-Jan–13 0.7 1201 1887 6518 15.6 quartile levels (Fig. 3, right lower panel). Thus phytoplankton 24-Feb-13 0.8 8576 4803 36,019 37.4 community structure alone cannot account for variations in EPR. 16-Jan–14 0.8 14,408 13,379 145,449 55.1 4-Feb-14 1.5 42,880 9148 101,197 51.8 Average 0.9 9153 4771 60,299 33.9 3.5. Relationships between EPR and Chl a, phytoplankton functional groups

When EPRs were plotted against total Chl a concentrations in EPR in winter–spring (hyperbola, p = 0.19). Neither dinoflagellates winter–spring (Jan–May) and summer (Jun–Oct), respectively, nor Cryptophytes were correlated significantly with EPR in either hyperbolic models (Michaelis–Menton) best represented their cor- season (not illustrated). An exception for dinoflagellates is that relations at p < 0.01 (Fig. 4). The estimated maximum EPR was 50.4 the subgroup – small dinoflagellates, was found significantly in winter–spring with a 0.4 lgLÀ1 Chl a half-saturation constant, correlated with EPRs in winter–spring (R2 = 0.24, p = 0.02) and this higher than the summer maximum EPR of 39.7 with a 2.6 lgLÀ1 correlation was strengthened when only winter data were Chl a half-saturation constant. considered (R2 = 0.65, p = 0.005). In the 10 experiments in Jan–Feb (Table 3), we collected female C. marshallae either before or during the small winter bloom (Chl À 3.6. Diatom community composition and species richness a = 1–2 lgL 1). An average EPR for winter was 33.9 (maximum of 55.1). Thus it is clear that significant numbers of eggs are pro- While the abundance of dinoflagellates, ciliates and Crypto- duced annually during winter months regardless of the presence phytes were similar among years and seasons, large fluctuations of significant phytoplankton blooms. were seen within diatom functional group in terms of species rich- To further explore this idea, EPRs were regressed against ciliate ness and composition (Table 4). Species composition was generally and diatom abundance by season, respectively (Fig. 5). A significant similar between 2013 and 2014, and in both years species richness correlation was found between EPR and ciliate abundance in was far higher (17 and 18 species, respectively) than 2011 and winter–spring (hyperbolic model, p = 0.008) with an estimated 2012 (7 and 9 species, respectively).The three most abundant taxa maximum EPR of 44.6, while in summer EPR and diatom abundance contributing to the interannual differences in the diatom commu- was significantly correlated (linear model, p = 0.01, Fig. 5). A nity in 2011 were Nitzschia, Dactyliosolen fragilissimus and Pseudo- hyperbolic model was also applied for EPR versus diatom nitzschia deli/p-deli; in 2012, Chaetoceros debilis, D. fragilissimus and abundance in summer but the result was less significant (p = 0.07) Nitzschia; in 2013, three Chaetoceros species (debilis, socialis and with a maximum EPR of 40.6. contortus); in 2014, Leptocylindricus minimus, D. fragilissimus and Non-significant relationships were found between ciliates and Ch. debilis. The full list is shown in Table 3. EPR in summer (hyperbola, p = 0.64), and between diatoms and By season, species richness increased gradually from winter (4 species) to summer (15 species) with only three species common in all seasons Cylindrotheca closterium, Thalassionema nitzschioides and Asterionellopsis glacialis (Table 3). There were 9 common diatom species between spring (Apr–May) and summer and the ones more likely to be present in both seasons are the genera Chaetoceros, Pseudo-nitzschia and Thalassiosira. EPRs were significantly correlated with species richness of those species which dominate in a given sample (those species which had an abundance >104 cells LÀ1 in summer from June to August). Species richness accounted for 32% of the variance (Fig. 6); however, this relationship became insignificant (p = 0.26) when the test period was expanded from June to October (plot not shown).

3.7. EPR and summer diatom blooms

Fig. 4. Egg production rates (EPR; eggs femaleÀ1 dayÀ1) of female Calanus marshal- In 5 of 21 experiments in summer (Jun–Oct), EPRs were lower lae as a function of total Chl a concentrations (lgLÀ1). Two hyperbolic curves are than expected, determined by the data points falling outside of the regressions of EPR versus total Chl a in Jan to May (green and grey symbols) and the lower 95% confidence interval (CI) band in the linear regression Jun to Oct (red symbols), respectively. Coefficients (a and b) are 50.4 and 0.4 for Jan plot (Fig. 5C). These observations raise the question as to what to May, 39.7 and 2.6 for Jun to Oct, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this might have been different in the phytoplankton community when article.) Chl a concentrations are high and/or when large diatom blooms 38 W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44

Fig. 5. Egg production rates of female Calanus marshallae as a function of total diatom (left panel A and C) and ciliate abundance (right panel B and D) by season (winter– spring, summer). In plot C, the blue lines indicate the 95% confidence intervals (CI). Below the lower 95% CI, empty circles indicate dates with very low EPRs and above the upper 95% CI, filled red circles indicate dates with very high EPRs. Note that the hyperbola model for diatom abundance versus EPR in summer was less significant (r = 0.41, p = 0.07, not illustrated here) than linear model. Coefficients (a and b) in each of the hyperbola models are: 43.7 and 61046.2 in (A); 44.6 and 544.2 in (B); 24.5 and 297.9 in (D). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

were observed in this season. Thus we inspected closely relation- diatom abundance), the same species associated with the low ships between EPRs and diatoms on those few dates (Table 5). EPR on 27 July 2011. However, for all experiments, the relation- ship between EPR and D. fragilissimus was not statistically  22 July 2011 (EPR = 3.8). Chl a concentrations were moderately significant (Hyperbolic model, n = 25, r = 0.2, p = 0.4). high (3.6 lgLÀ1) but the small <5 lm size fraction was 81% of  7 Aug 2012 (EPR = 5.7). Chl a concentration was 3.8 lgLÀ1 with total. Total diatom abundance was very low (31,903 cells LÀ1) 66% in the <5 lm size fraction. The low EPR was similar to the while the small dinoflagellates group was the highest observed earlier observation on 20 July 2012. Diatoms developed another during our study (99,767 cells LÀ1, 99% of total dinoflagellate typical bloom (324,688 cells LÀ1) with co-existed dominants abundance on this date). Since both diatoms and Chl a >5 lm such as Coscinodiscus, Pseudo-nitzschia, G. delicatula and Ch. size fraction had low concentrations, this suggests that on this debilis. However, the large dinoflagellate Protoperidinium com- date small dinoflagellates alone, though abundant, might not pressum and silicoflagellate D. speculum both were unusually be a suitable food resource to induce high EPRs. abundant at the same time (31,388 and 162,944 cells LÀ1,  27 July 2011 (EPR = 10.1). Chl a concentration was higher respectively) compared to their long term average amount in (6.3 lgLÀ1) than on 22 July but with 84% in the <5 lm size frac- our study area (510 and 3808 cells LÀ1, respectively). The low tion. A mono-species bloom of the diatom D. fragilissimus EPR might be related to these two unusually abundant species. occurred (691,827 cells LÀ1, 98% of total diatom abundance)  7 June 2013 (EPR = 13.7). Chl a concentration was again at a which might explain the low EPR. moderate level, 4.2 lgLÀ1, with 39% in the <5 lm size fraction.  20 July 2012 (EPR = 7.7). Chl a concentration was 3.0 lgLÀ1 Diatom blooms were characterized by both high species with 48% in the <5 lm size fraction. EPR should have been richness and abundance in June and July 2013. The diatom higher because a typical summer diatom bloom was observed bloom on 7 June 2013 (1,205,789 cells LÀ1) was in a similar on this date, with total diatom abundance of 960,511 cells LÀ1 magnitude to the other June and July observations and and a diverse set of dominants: D. fragilissimus, Skeletonema characterized by a common set of genera/species including costatum, Guinardia delicatula, Chaetoceros (Ch. contortus, Ch. Chaetoceros (e.g. Ch. contortus and debilis), Asterionellopsis, debilis, Ch. socialis, Ch. radicans, Ch. affinis, Ch. lorenzianus and S. costatum, Pseudo-nitzschia and C. closterium. This gives an similis) and Thalassiosira nordenskioeldii, The most abundant example of the apparent mismatch between a seemingly ideal species was again D. fragilissimus (564,302 cells LÀ1, 51% of total food supply and a low EPR. W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 39

Table 4 Comparisons of diatom community composition by year (shown as abundance) and season (as rank). The species list was determined by similarity of percentage (SIMPER) analysis (more details in Methods). Genus: A (Asterionellopsis); As (Asteromphalus); C (Cerataulina); Ch (Chaetoceros); Cylin (Cylindrotheca); D (Dactyliosolen); L (Leptocylindrus); P-n (Pseudo-nitzschia); S (Skeletonema); T (Thalassiosira).

Taxa Abundance (104 cells LÀ1) Rank 2011 2012 2013 2014 Jan–Feb Apr–May Jun–Oct A. glaciali 0.8 8.7 4 6 4 A. socialis 0.9 15 As. sarcophagus 0.2 C. pelagica 0.1 Ch. affinis 0.4 Ch. contortus 13.1 1.8 7 13 Ch. debilis 14.6 17 3.3 11 5 Ch. decipiens 0.5 Ch. didymus 0.5 Ch. laciniosus 0.4 Ch. lorenzianus 0.2 0.5 0.6 Ch. radicans 0.4 10 Ch. socialis 16.8 2.5 10 Ch. vanheurckii 0.4 Coscinodiscus 3 Cylin. closterium 0.1 0.1 0.9 2.1 2 1 3 D. fragilissimus 5.1 5.2 28.1 4 6 L. danicus 0.6 0.2 12 L. minimus 29.7 12 Navicula 0.2 0.3 9 11 Nitzschia 6.5 0.5 0.2 0.8 3 1 P-n deli/p-deli 3.4 0.1 2.5 2.1 5 2 P-n fraud/aust 3.8 14 P-n multi/pung 0.6 7 S. costatum 4.1 2.4 T. nitzschioides 0.7 0.3 0.9 1 2 9 T. nordenskioeldii 4.9 T. rotula 0.1 Thalassiosira 1.3 88

 26 Oct 2011 (EPR = 24.9). The Chl a value was very low, only 1.3 lgLÀ1 (62% for <5 lm fraction) and EPR was close to the average. Diatom abundance was similarly low, 14,065 cells LÀ1, whereas small dinoflagellates were more abundant (23,670 cells LÀ1).  23 and 29 July 2013 (EPR = 35.6 and 61.7, respectively). A bloom developed during these two sampling dates, with Chl a ranging from 7.5 lgLÀ1 to 15.9 lgLÀ1 and <5 lm size fraction compris- ing 11% and 77%, respectively. Diatom abundances were excep- tionally high (1.1 and 1.6 Â 106 cells LÀ1). Dominance within the diatom community was shared between Chaetoceros (61% and 44%, respectively) and Thalassiosira (24% and 44%, respectively).

Note that (Table 5) EPRs were high on days when Thalassiosira were abundant during summer. Regression analysis (Fig. 7) between EPR and Thalassiosira abundance showed a significant Fig. 6. Relationships between egg production rate (EPR) and diatom species hyperbolic relationship (n = 17, r = 0.64, p = 0.0057). richness during Jun to Aug (hyperbolic model, R2 = 0.32). Species richness used here is the number of species with cell abundance (N) >104 cells LÀ1. Coefficient a = 46.9 and b = 6.9. 4. Discussion

In comparison, five dates in summer had extraordinarily high 4.1. Spawning of Calanus copepods EPRs and fell above of the upper 95% CI. These dates were in 2011 (17 Aug, 29 Sep and 26 Oct) and 2013 (23 and 29 July). Multiple years of egg production experiment demonstrate that the C. marshallae population off Oregon produces eggs from winter À  17 Aug 2011 (EPR = 30.9). Chl a value was very high, 17.3 lgL 1 through late summer. Eggs are produced in winter months soon (76% in the <5 lm fraction). Total diatom abundance was mod- after the termination of diapause and mostly before the first phy- À erately high (571,849 cells L 1) with a 41% contribution from toplankton bloom supporting the idea that there must be triggers Thalassiosira. other than the bloom that act to awaken them from diapause À  29 Sep 2011 (EPR = 48.6). Chl a value was low as 2.6 lgL 1 (65% (Miller et al., 1991). for <5 lm fraction) but EPR was high. Diatoms were dominant Spawning before the first bloom of the year, either in winter or À (487,118 cells L 1) and Ch. debilis alone made up 83% of total spring, is not an exclusive trait for C. marshallae. Many studies of abundance. other Calanus species also found that eggs were produced before 40 W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44

Table 5 EPR (eggs femaleÀ1 dayÀ1), diatom species richness (S), diatom total cell abundance (N, cells LÀ1) and the general amount (<104 or >104 cells LÀ1) of those diatom taxa that form blooms during the summer upwelling season (Jun–Oct) in 2011–2014. S (N >104): the number of species with abundance >104 cells LÀ1. Aster: Asterionellopsis; Chae: Chaetoceros; Rhiz: Rhizosolenia + Dactyliosolen + Guinardia + Proboscia; Lept: Leptocylindrus; Pennate: Navicula + Nitzschia + Pleurosigma + Thalassionema; P-n: Pseudo-nitzschia; Thal: Thalas- siosira. ‘‘ + ” means N <104; ‘‘++” means N >104.

Samples SS(N >104) N EPR Aster Chae Rhiz Lept Pennate P-n Thal 6/8/2011 9 3 100,854 14 0 ++ + + ++ + 0 7/6/2011 11 1 35,676 10 + + 0 + ++ + + 7/22/2011 6 2 31,903 4 0 0 ++ 0 + 0 + 7/27/2011 4 1 702,376 10 + 0 ++ 0 + 0 + 8/4/2011 8 6 4,077,042 40 ++ 0 ++ 0 ++ ++ ++ 8/17/2011 20 10 571,849 31 ++ ++ ++ 0 ++ ++ ++ 9/1/2011 8 5 1,034,268 25 + 0 ++ 0 ++ ++ ++ 9/13/11 3 3 512,846 20 0 0 0 0 ++ ++ 0 9/29/11 7 3 487,118 43 ++ ++ 0 0 + ++ 0 10/26/11 10 0 14,065 25 + + + 0 + + + 7/20/2012 24 12 960,515 8 + ++ ++ + ++ + ++ 8/7/2012 19 6 324,688 6 0 ++ ++ + + ++ ++ 6/7/2013 25 14 1,205,789 14 ++ ++ ++ + ++ ++ 0 6/19/2013 27 13 2,181,740 31 ++ ++ + ++ + ++ ++ 7/9/2013 21 7 874,183 24 ++ ++ ++ + ++ ++ ++ 7/23/2013 24 16 1,129,462 36 ++ ++ + + ++ ++ ++ 7/29/2013 20 11 1,602,001 62 ++ ++ 0 0 ++ ++ ++ 9/26/13 17 6 1,154,561 24 0 ++ ++ + + ++ + 10/22/13 28 20 1,306,986 28 ++ ++ + ++ ++ ++ ++ 6/11/2014 21 7 373,572 18 0 ++ ++ 0 ++ ++ ++ 7/22/2014 26 23 8,694,373 41 ++ ++ ++ ++ ++ ++ ++

EPRs were often higher in winter before any blooms and lower dur- ing summer when large diatom blooms were frequent. As for Calanus species in the Pacific Ocean, Baier and Napp (2003) reported that C. marshallae in the Bering Sea also spawned well before the spring bloom, as early as February based on back-calculated temperature-dependent development rates of the age-structure observed in April and May. The same is true for C. pacificus in Dabob Bay, WA (Leising et al., 2005; Pierson et al., 2005), and for C. pacificus in the coastal waters off Oregon (Peter- son, unpublished data). Spawning by C. chilensis in the southern/ central Chile (Hidalgo and Escribano, 2007) is seasonal whereas C. chilensis spawn year round in waters off northern Chile (Hidalgo and Escribano, 2008). C. sinicus can spawn year around in the Yellow Sea, P.R. China, although most females do not produce eggs during summer months but go through ‘over-summering’ at depth because temperature of the upper Fig. 7. Relationship (hyperbolic model, coefficients a = 40.6, b = 24794.9) between mixed layer is very high, exceeding 18 °C(Huo et al., 2008; Kang egg production rate (EPR) and diatom genus Thalassiosira abundance (cells LÀ1). et al., 2011; Wang et al., 2009; Zhang et al., 2005). Uye and Murase (1997) reported maximum EPRs of 22–24 for C. sinicus from coastal waters off Japan (Seto Inland Sea) in spring, similar the first bloom of the year, although the numbers of eggs were sub- to the rates reported by the above cited studies in the Yellow Sea. stantially lower than those observed during or post-blooms. As examples, Plourde and Runge (1993) working on C. finmarchicus in the Gulf of St. Lawrence reported EPR <10 pre-bloom, 22–82 4.2. EPR and phytoplankton concentrations during the bloom and 30–50 post-bloom. Niehoff et al. (1999) for C. finmarchicus in the Norwegian Sea reported EPR of 8 Significant functional relationships were found between EPR pre-bloom and 44 during the spring bloom. Gislason and and total Chl a concentration in our study. Our past work in June Astthorsson (2000) reported a maximum EPR for C. finmarchicus 1996 and July–September 1997 also found strong correlations of 7 pre-bloom and 46 during the bloom in waters off Iceland. A between EPR and Chl a (Gómez-Gutiérrez and Peterson, 1999; similar result for C. glacialis (Melle and Skjoldal, 1998; Niehoff Peterson et al., 2002). This result is common to studies of other and Hirche, 2005) was found in a Norwegian fjord (see Table 2 in species in the genus of Calanus. However, the uniqueness of our Niehoff and Hirche, 2005). Stenevik et al. (2007) reported for findings is that we noted differences in season-specific functional C. finmarchicus in the Norwegian Sea EPR of 10, 22 and 18 with response of C. marshallae EPR to phytoplankton availability. Hyper- pre-, during and post-bloom conditions, respectively. Madsen bolic models provided the best fit for both winter–spring and sum- et al. (2008) found both C. finmarchicus and C. glacialis spawned mer and the parameters from the models inform clearly different before the spring bloom in Disco Bay Greenland, as did Runge biological behaviors and are consistent with in situ observations. et al. (2006) for C. finmarchicus on Georges Bank, U.S. Conversely, The very high EPRs in winter and spring were not simply a function Head et al. (2013) found that EPRs of C. finmarchicus in the central of autotrophic phytoplankton as a food source; rather, they were Labrador Sea were not different under pre-, during, and likely the result of multiple feeding strategies that female C. mar- post-bloom conditions. Most of these observations cited for other shallae use for producing eggs. In summer, the half-saturation con- Calanus species are in stark contrast to our work off Oregon where stant (Chl a = 2.6 lgLÀ1) for reaching the maximum EPR was much W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 41 lower than the magnitude of summer blooms (10–20 lgLÀ1) fatty acid markers were correlated with EPRs. Vargas et al. suggesting the amount of phytoplankton (mostly diatoms) was (2010) reached the same conclusion for copepods from the upwel- not limiting egg production and instead other factors (prey types ling zone off Dichato, Chile. Ceballos and Alvarez-Marques (2006) or environment) caused the high variability in EPR. The significant also found significant correlations between diatom concentration correlation between EPR and Chl a were obscured when the data and egg production of both C. helgolandicus and Calanoides carina- from all seasons were combined for the test (r = 0.17, p = 0.2), tus, in the coastal upwelling zone off northern Spain (Cantabrian which again highlights the different functional response in egg Sea). EPR of C. pacificus in Dabob Bay, WA was significantly corre- production at different times of the year. lated with diatom concentration in winter and spring (Pierson The significance of EPRs for C. marshallae responding to the et al., 2005; Leising et al., 2005). The study by Ianora et al. abundance of different microplankton functional groups also dif- (2015) observed higher in situ EPRs of Acartia clausi and fered by season: EPRs in winter/spring were best correlated with C. helgolandicus in the years when diatoms were more productive ciliate abundance, whereas in summer were best correlated with in Adriatic Sea. diatom abundance. Similar findings were reported by Vargas et al. (2006) for the copepods Acartia tonsa and Centropages 4.3. EPR and phytoplankton community structure brachiatus from the coastal upwelling zone off Dichato, Chile where copepods fed on dinoflagellates and ciliates in winter but diatoms We had hypothesized that the patterns of phytoplankton com- in summer. munity structure from traditional cluster and ordination analysis Some of our winter EPRs were the highest in a given year and would explain the variation in EPR due to the combined informa- occurred before the winter bloom. We thus hypothesize that in win- tion on species composition and abundance, functional groups ter energetic requirements for egg production and embryo develop- and species richness. However, our results did not show any ment may come from a combination of lipid reserves and from covariance between the five significantly different community feeding omnivorously on both phytoplankton and microzooplank- types and EPR values (as compared in Fig. 3). On the other hand, ton. At this time of the year, EPR ranged from 4 to 68.5 over a range species richness alone seemed to inform better variation in EPR of diatom abundance of 0–180,000 cells LÀ1 (mean 28,650; median which supports the hypothesis presented by others (e.g., Irigoien 3002 cells LÀ1) and ciliate abundance of 343–13,379 cells LÀ1 (mean et al., 2000b; Leising et al., 2005; Paffenhofer et al., 2005 and refer- 3345; median 4328 cells LÀ1). We attribute the extraordinary EPR in ence therein) that mono-specific blooms (lower species diversity) winter 2014 (that were nearly double that in the other winters may result in a decline in EPR. In contrast, a mixture of microplank- and summers) to the abundance of ciliates (naked ciliates, ton species may provide a more balanced diet for copepods (Poulet 11,263 cells LÀ1, Table 3) and dinoflagellates (mostly small dinoflag- et al., 2006; Uye, 1996; Vehmaa et al., 2011). ellates, 28,644 cells LÀ1). Jonasdottir et al. (2002) also reported unusually high clutch size of C. finmarchicus (75–110 eggs per 4.4. Why were there some cases of exceptionally low EPR in summer? female) in April when Chl a concentrations were below 1 lgLÀ1; however, they did not report ciliate/dinoflagellate information. Average EPRs were lower in summer as compared to spring or There are only a few studies about relationships between Cala- winter. Our analyses of individual cases in summer (see Section 3.7) nus egg production rate and concentration of taxa/species of provide the following empirical understanding. First, in the trivial microplankton, but most of them, such as Ohman and Runge case, when both Chl a and diatom concentration were low, EPRs (1994) for C. finmarchicus in the Gulf of St. Lawrence, reported sim- were low; second, mono-species blooms of certain diatoms and ilarly that EPRs were high when feeding on ciliates and/or dinoflag- dinoflagellates were associated with low EPR; third, high abun- ellates under post-spring bloom conditions. Another study by dance of Thalassiosira along with a diverse set of other abundant Irigoien et al. (2000a) in a 7-year study of egg production by diatoms were associated with high EPRs. The second and third C. helgolandicus in the English Channel at Station L4 off Plymouth points are consistent with positive association of higher species showed that flagellates and ciliates were the most important richness with higher EPR (Fig. 6). Limited evidence was shown factors explaining variations in egg production. Further evidence about the possibility of detrimental effects caused by mono- that ciliates are important prey comes from feeding studies – the species blooms of certain phytoplankton species (as in 2 of 21 Nejstgaard et al. (2001) mesocosm experiments showed that experiments). Recall cases on both 27 July 2011 and 20 July 2012, C. helgolandicus generally preferred non-diatom foods such as cili- the dominant diatom was D. fragilissimus whereas EPRs were excep- ates and metazoans. Jonasdottir et al. (2011), in another mesocosm tionally low. The same observation was made by Ceballos and study, reported significant correlations between egg production of Álvarez-Marqués (2006) that EPRs were decreased during a bloom C. finmarchicus and ciliate biomass. Moreover, Jonasdottir of D. fragilissimus for both C. helgolandicus and Calanoides carinatus and Koski (2011) showed that EPR of C. finmarchicus and in July 2002. Further, the bloom observed by us on 7 Aug 2012 C. helgolandicus from the North Sea in July and August were best showed that an unusually high concentration of the silicoflagellate explained by the abundance of dinoflagellates, flagellates and D. speculum was associated with very low EPR, and this too has been ciliates. Finally, Fessenden and Cowles (1994) who conducted feed- found previously by Nejstgaard et al. (2001) in their mesocosm ing experiments on copepods collected from the same station off experiments with C. helgolandicus. The dinoflagellate Gyrodinium Newport, Oregon and as our work, showed that during summer aureleum was shown to have a negative influence on egg production diatom blooms (when Chl a reached 45 lgLÀ1), C. pacificus fed on of C. helgolandicus (Irigoien et al., 2000a) as was the haptophyte diatoms but switched to ciliates in winter. Phaeocystis sp. (Bautista et al., 1994) on egg production of the same Our results support the importance of diatom diets to species. Others have shown reduced EPRs during dense diatom C. marshallae egg production during summer upwelling months. blooms but either species were not enumerated or there were no This has been acknowledged by many studies for different species. dominant species that could be associated with reduced EPR For example, Irigoien et al. (2000b) measured both EPRs and inges- (Nejstgaard et al., 2001; Paffenhofer, 2002; Poulet et al., 2007, tion rates of C. helgolandicus on various phytoplankton taxa and his 2006). Pierson et al. (2005) noted that EPRs of C. pacificus in Dabob results showed little selective feeding but a slight preference for Bay, WA were reduced during a bloom of Thalassiosira; however, we diatoms. Jonasdottir et al. (2002) compared egg production of C. did not detect such negative effects (Fig. 7) by this species. Some finmarchicus to quality of phytoplankton (indexed through fatty studies did not report negative effects from mono-species blooms acid analysis) in shelf waters off Iceland and found that diatom on egg production: these include blooms of Phaeocystis (Irigoien 42 W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 et al., 2000b; Jonasdottir et al., 2011) and Skeletonema (Irigoien et al., 2005). Many factors are suspected for the depressed EPR during dia- tom blooms, such as growth rate and phase of the bloom/taxa dur- ing a bloom, fatty acid composition and content of diatoms, ratios of protein: carbohydrate, toxin effects associated with the presence of certain aldehydes, etc. and each of these factors have been reviewed succinctly in Paffenhofer et al. (2005) thus need not be restated here. Beyond the potential effects of the factors related to food quality, there are also response differences by the copepods themselves, for example, their susceptibility to toxins and detoxi- fication capability (Wichard et al., 2008). Another potential reason for the reduced EPR is that during the summer upwelling season in the northern California Current the blooms are often caused by long-chain diatom species easily extending beyond 1 mm in length, and those species are from the genera Chatoceros, Thalassiosira, Leptocylindrus, Pseudo- nitzschia, Skeletonema, Guinardia, etc. We speculate that these spe- cies may be too large to be handled easily and thus would be less actively selected by copepods and is based on our observations that when blooms of large-cell diatoms were present, EPR were low (e.g. 7 June 2013). Studies by Peterson et al. (2002) and Vargas et al. (2007) both argued for low efficiency of primary production transfer through the classical diatom food chain in the productive coastal upwelling system and this may be partially related to inedible diatom species.

À1 À2 4.5. Potential recruitment rates (eggs day m ) Fig. 8. Interannual and seasonal changes of potential recruitment rate of Calanus marshallae during 2011–2014. Ceballos and Alvarez-Marques (2006) may have been the first to coin the term ‘‘potential reproduction rate” (PRR) for the calcula- The PRR calculation may be considered as estimate of potential tion of in situ female abundance  egg production rate. In our ‘‘recruitment rate”, but obviously do not account for egg mortality study, the highest PRR was found in 2013 whereas PRRs were very (Ohman et al., 2002; Peterson and Kimmerer, 1994) nor for variable low throughout 2014 and as well in 2012 (Fig. 8). When shown as egg hatching or naupliar survival rates. However, the PRR calcula- climatology, rates were highest in spring and lower during the tion does illustrate the total number of egg produced by all females summer upwelling season. Maximum rates were on a given day thus is an estimate of secondary production by the 174,977 eggs dayÀ1 mÀ2 (the same units are used hereafter) with ‘population’, which is especially relevant because even if the an average of 30,433. No recruitment was apparent after August females in an incubation produce high numbers of eggs, if there of any year, because there were no females in the quantitative ver- are few-to-no females in the ocean, then the data from incubations tical towed net since most of them had entered diapause by that must be valued differently. In this regard, we have clearly shown time (although in 2011 and 2013, there were enough females in that there was no recruitment after August in any of the four years the Bongo net tow in fall and winter to be used for the EPR of data reported here, a fact that adds value to the calculation of PRR. incubations). Niehoff et al. (1999) reported nearly the same result of PRR for 4.6. Laboratory estimates of EPR compared to in situ results C. finmarchicus from the North Atlantic at Weathership M – up to 200,000 but with an average of 34,187 pre-spring bloom and In an extensive set of laboratory experiments, Peterson (1988) 34,855 during the bloom. Stenevik et al. (2007) working on the estimated the functional response of feeding rates, fecal pellet pro- same species in the Norwegian Sea as part of the ESSAS study duction rates and egg production rates by female C. marshallae fed reported an average PRR of 146,926 before the spring bloom period a wide range of concentrations of T. weissflogii. The maximum EPR and 220,742 during the spring bloom. Gislason and Astthorsson estimated from an Ivlev model was only 23.5. This is similar to (for (2000) reported much lower PRRs, noting a maximum of 16,000 some cases) but mostly lower than values observed in this study for C. finmarchicus in oceanic waters but only 1000 in shelf during summer, and far less than rates observed in winter (e.g., waters. Harris et al. (2000) reported a maximum of 180,000 for an EPR of 53.4 in 2014 and four-year mean of 32.5) and spring C. finmarchicus in the Norwegian Sea and an average of about (EPR of 44.7 in 2013 and four-year mean of 40.2). The reason for 10,000 for C. helgolandicus from the English Channel (a coastal sta- these large differences cannot be known but we are concerned that tion at a similar water depth as Newport Line). Bonnet et al. (2005) T. weissflogii, being an easily-grown ‘laboratory weed’, was not compared data on C. helgolandicus from Stonehaven (North Sea, nutritionally adequate during those experiments despite every 50 m depth), English Channel (Plymouth, 55 m depth) and the Can- attempt to maintain cells in log-growth phase, as described in tabrian Sea (N. Spain). Maximum values were 256,000 (Cantabrian Peterson (1988). We suggest that some zooplanktologists may Sea), 35,146 (English Channel) and 11,764 (Stonehaven). Ceballos need to re-think past lab work which used easily cultured diatoms and Alvarez-Marques (2006) reported maximum PRR of 56,140 (‘weeds’) and re-do important rate measurements, especially if for C. helgolandicus and 57,444 for Calanoides carinatus in coastal reported rates are less than rates known from in situ incubations. waters off Spain. Jonasdottir and Koski (2011) reported similar This is especially true for C. marshallae development time for indi- maxima for Calanus helgolandicus in the North Sea (Dogger Bank) vidual life stages from N3 to adult – in the experiments reported by during summer with a range of 443–49,355 and for C. finmarchicus, Peterson (1986), the C5 stage required several weeks to mature; 162–49,954. moreover, only females were produced in those experiments. W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44 43

4.7. Limitation of our approach References

We did not measure ingestion rates of the copepods in our Baier, C.T., Napp, J.M., 2003. Climate-induced variability in Calanus marshallae populations. Journal of Plankton Research 25, 771–782. study so strictly speaking, any statement about the importance of Bonnet, D., Richardson, A., Harris, R., Hirst, A., Beaugrand, G., Edwards, M., Ceballos, ciliates or diatoms should be stated as a hypothesis. Based on their S., Diekman, R., Lopez-Urrutia, A., Valdes, L., Carlotti, F., Molinero, J.C., Weikert, herculean work on C. pacificus and Pseudocalanus in Dabob Bay H., Greve, W., Lucic, D., Albaina, A., Yahia, N.D., Umani, S.F., Miranda, A., Santos, (Leising et al., 2005; Pierson et al., 2005), a sobering discussion is A.d., Cook, K., Robinson, S., Fernandez de Puelles, M.L., 2005. An overview of Calanus helgolandicus ecology in European waters. Progress in Oceanography 65, provided by them on the problems and pitfalls associated with 1–53. trying to relate data on phytoplankton species composition, Calbet, A., 2008. The trophic roles of microzooplankton in marine systems. ICES C. pacificus ingestion and egg production rates to selective feeding, Journal of Marine Science 65, 325–331. Ceballos, S., Alvarez-Marques, F., 2006. Reproductive activity and physiological and inferences that one species or another affected EPR. Another status of the calanoid copepods Calanus helgolandicus and Calanoides carinatus extensive effort was shown by Jonasdottir et al. (2011) in a under food-limiting conditions. Journal of Experimental Marine Biology and mesocosm study and again the problems with determining which Ecology 339, 189–203. Clarke, K.R., Gorley, R.N., 2006. PRIMER v6: User Manual/Tutorial. PRIMER-E Ltd, factors control copepod egg production were discussed. However, Plymouth. the fact remains that not all factors related to food quality can be Du, X., Peterson, W., 2014. Seasonal cycle of phytoplankton community composition controlled in a laboratory experiment, making it nearly impossible in the coastal upwelling system off central Oregon in 2009. Estuaries and Coasts 37, 299–311. to demonstrate mechanism(s) that account for the greatest Feinberg, L.R., Peterson, W.T., Shaw, T.C., 2010. The timing and location of spawning cumulative amount of variance in EPR. This problem and potential for the Euphausiid Thysanoessa spinifera off the Oregon coast, USA. Deep-Sea solutions need not to be discussed here since many ideas and Research II 57, 572–583. Fessenden, L., Cowles, T., 1994. Copepod predation on phagotrophic ciliates in hypotheses are succinctly articulated in a workshop held in Naples Oregon coastal waters. Marine Ecology Progress Series 107, 103–111. in November 2002 (summarized in Paffenhofer et al., 2005). Frost, B.W., 1974. Calanus marshallae, a new species of Calanoid Copepod closely However, if there were ever a near-perfect experiment, the two allied to the sibling species C. finmarchicus and C. glacialis. Marine Biology 26, studies cited above come to mind: the in situ study of C. pacificus 77–99. Gislason, A., Astthorsson, O.S., 2000. Winter distribution, ontogenetic migration, from Dabob Bay and the mesocosm study in Norway. and rates of egg production of Calanus finmarchicus southwest of Iceland. ICES Journal of Marine Science 57, 1727–1739. Gómez-Gutiérrez, J., Peterson, W.T., 1999. Egg production rates of eight calanoid copepod species during summer 1997 off Newport, Oregon, USA. Journal of 5. Conclusions Plankton Research 21, 637–657. Harris, R.P., Irigoien, X., Head, R.N., Rey, C., Hygum, B.H., Hansen, B.W., Niehoff, B., We have shown that EPRs were highest in spring and winter Meyer-Harms, B., Carlotti, F., 2000. Feeding, growth, and reproduction in the and lowest during the summer upwelling season. Significant func- genus Calanus. ICES Journal of Marine Science 57, 1708–1726. Head, E.J.H., Melle, W., Pepin, P., Bagoien, E., Broms, C., 2013. On the ecology of tional relationships were shown between EPR and Chl a, but only Calanus finmarchicus in the Subarctic North Atlantic: a comparison of when the dates were parsed into winter–spring and summer. High population dynamics and environmental conditions in areas of the Labrador egg production in winter and spring were attributed to a greater Sea-Labrador/Newfoundland Shelf and Norwegian Sea Atlantic and Coastal Waters. Progress in Oceanography 114, 46–63. abundance of ciliates and likely dinoflagellates at that time. Varia- Hidalgo, P., Escribano, R., 2007. Coupling of life cycles of the copepods Calanus tions in EPR during summer remain enigmatic in that on some chilensis and Centropages brachiatus to upwelling induced variability in the occasions, extremely low rates were measured whereas at other central-southern region of Chile. Progress in Oceanography 75, 501–517. Hidalgo, P., Escribano, R., 2008. The life cycles of two coexisting copepods, Calanus times, extremely high rates were measured; based on Chl a con- chilensis and Centropages brachiatus, in the upwelling zone off northern Chile centrations, EPR should not have been food limited. We suggested (23°S). Marine Biology 155, 429–442. that during certain mono-specific diatom blooms, EPRs were Hopcroft, R.R., Kosobokova, K.N., Pinchuk, A.I., 2010. Zooplankton community patterns in the Chukchi Sea during summer 2004. Deep Sea Research Part II 57, depressed whereas when diatom species richness was high during 27–39. some blooms in summer, EPRs were high. Since we did not conduct Horner, R.A., 2002. A Taxonomic Guide to Some Common Marine Phytoplankton. copepod feeding experiments on natural particle assemblages, we Biopress Ltd, Bristol, England. Huo, Y.-Z., Wang, S.-W., Sun, S., Li, C.-L., Liu, M.-T., 2008. Feeding and egg production do not know if ciliates continue to contribute as an important com- of the planktonic copepod Calanus sinicus in spring and autumn in the Yellow ponent of copepod diets in summer, but clearly, diatom blooms are Sea, China. Journal of Plankton Research 30, 723–734. important to sustain high EPRs although they do not always result Ianora, A., Bastianini, M., Carotenuto, Y., Casotti, R., Roncalli, V., Miralto, A., Romano, in high EPRs. Finally, Vargas et al. (2006) remarked ‘‘that the clas- G., Gerecht, A., Fontana, A., Turner, J.T., 2015. Non-volatile oxylipins can render some diatom blooms more toxic for copepod reproduction. Harmful Algae 44, sical marine food web model does not apply to some coastal 1–7. upwelling systems”. Further, Calbet (2008) reminds us that the Irigoien, X., Harris, R.P., Head, R.N., Harbour, D., 2000a. The influence of diatom ‘‘strength of the mesozooplankton–microzooplankton link is tradi- abundance on the egg production rate of Calanus helgolandicus in the English Channel. Limnology and Oceanography 45, 1433–1439. tionally overlooked in plankton studies”. We could not agree more. Irigoien, X., Head, R.N., Harris, R.P., Cummings, D., Harbour, D., Meyer-Harms, B., 2000b. Feeding selectivity and egg production of Calanus helgolandicus in the English Channel. Limnology and Oceanography 45, 44–54. Acknowledgements Irigoien, X., Verheye, H.M., Harris, R.P., Harbour, D., 2005. Effect of food composition on egg production and hatching success rate of two copepod species (Calanoides carinatus and Rhincalanus nasutus) in the Benguela upwelling system. Journal of We thank C. Tracy Shaw, Jay Peterson (JP), Jennifer Fisher (JF) and Plankton Research 27, 735–742. the captains of the R/V ELAKHA for collecting all the samples for this Jonasdottir, S., Dutz, J., Koski, M., Yebra, L., Jakobsen, H., Vidoudez, C., Pohnert, G., Nejstgaard, J., 2011. Extensive cross-disciplinary analysis of biological and study along the Newport Hydrographic Line. Jay and Jennifer chemical control of Calanus finmarchicus reproduction during an aldehyde conducted many of the egg production incubation experiments as forming diatom bloom in mesocosms. Marine Biology 158, 1943–1963. well. JP and JF provided comments on various earlier drafts of this Jonasdottir, S., Gudfinnsson, H., Gislason, A., Astthorsson, O., 2002. Diet composition manuscript. Post-doctoral support for X. Du was provided by a grant and quality for Calanus finmarchicus egg production and hatching success off south-west Iceland. Marine Biology 140, 1195–1206. from the NOAA-MERHAB program (NA07NOS4780195). Funding for Jonasdottir, S.H., Koski, M., 2011. Biological processes in the North Sea: comparison collection of zooplankton samples from 2011 onwards was provided of Calanus helgolandicus and Calanus finmarchicus vertical distribution and by the NOAA Stock Assessment Improvement Program, NOAA Fish- production. Journal of Plankton Research 33, 85–103. Kang, H.-K., Lee, C.-R., Choi, K.-H., 2011. Egg production rate of the copepod Calanus eries and the Environment Program, as well as residual funds from sinicus off the Korean coast of the Yellow Sea during spring. Ocean Science the U.S. GLOBEC/Synthesis program. Journal 46, 133–143. 44 W.T. Peterson, X. Du / Progress in Oceanography 138 (2015) 32–44

Lane, P.V.Z., Llinas, L., Smith, S.L., Pilz, D., 2008. Zooplankton distribution in the Beaufort Seas during spring and summer, 2002. Deep Sea Research Part II: western Arctic during summer 2002: hydrographic habitats and implications Topical Studies in Oceanography 52, 3411–3426. for food chain dynamics. Journal of Marine Systems 70, 97–133. Plourde, S., Runge, J.A., 1993. Reproduction of the planktonic copepod Calanus Leising, A.W., Pierson, J.J., Halsband-Lenk, C., Horner, R., Postel, J., 2005. Copepod finmarchicus in the lower St. lawrence Estuary: relation to the cycle of grazing during spring blooms: does Calanus pacificus avoid harmful diatoms? phytoplankton production and evidence for a Calanus pump. Marine Ecology Progress in Oceanography 67, 384–405. Progress Series 102, 217–227. Madsen, S.J., Nielsen, T.G., Tervo, O.M., Soderkvist, J., 2008. Importance of feeding for Poulet, S.A., Escribano, R., Hidalgo, P., Cueff, A., Wichard, T., Aguilera, V., Vargas, C.A., egg production in Calanus finmarchicus and C. glacialis during the Arctic spring. Pohnert, G., 2007. Collapse of Calanus chilensis reproduction in a marine Marine Ecology Progress Series 353, 177–190. environment with high diatom concentration. Journal of Experimental Marine Melle, W., Skjoldal, H.R., 1998. Reproduction and development of Calanus Biology and Ecology 352, 187–199. finmarchicus, C. glacialis and C. hyperboreus in the Barents Sea. Marine Ecology Poulet, S.A., Wichard, T., Ledoux, J.B., Lebreton, B., 2006. Influences of diatoms on Progress Series 169, 211–228. copepod reproduction. I Field and laboratory observations related to Calanus Miller, C.B., Cowles, T.J., Wiebe, P.H., Copley, N.J., Grigg, H., 1991. Phenology in helgolandicus egg production. Marine Ecology Progress Series 308, 129–142. Calanus finmarchicus: hypotheses about control mechanism. Marine Ecology Runge, J.A., Plourde, S., Joly, P., Niehoff, B., Durbin, E., 2006. Characteristics of egg Progress Series 72, 79–91. production of the planktonic copepod, Calanus finmarchicus, on Georges Bank: Nejstgaard, J.C., Hygum, B.H., Naustvoll, L.H., Bamstedt, U., 2001. Zooplankton 1994–1999. Deep Sea Research Part II: Topical Studies in Oceanography 53, growth, diet and reproductive success compared in simultaneous diatom- and 2618–2631. flagellate-microzooplankton-dominated plankton blooms. Marine Ecology Runge, J.A., Roff, J.C., 2000. The measurement of growth and reproductive rates. In: Progress Series 221, 77–91. Harris, R.P., Wiebe, P.H., Lenz, J., Skjoldal, H.R., Huntley, M. (Eds.), ICES Niehoff, B., Hirche, H.J., 2005. Reproduction of Calanus glacialis in the Lurefjord Zooplankton Methodology Manual. Academic Press, pp. 401–444. (western Norway): indication for temperature-induced female dormancy. Shaw, C.T., Feinberg, L.R., Peterson, W.T., 2009. Interannual variations in vital rates Marine Ecology Progress Series 285, 107–115. of copepods and euphausiids during the RISE study 2004–2006. Journal of Niehoff, B., Klenke, U., Hirche, H.J., Irigoien, X., Head, R., Harris, R., 1999. A high Geophysical Research: Oceans 114, C00B08. frequency time series at Weathership M, Norwegian Sea, during the 1997 spring Smith, S.L., Vidal, J., 1984. Spatial and temporal effects of salinity, temperature and bloom: the reproductive biology of Calanus finmarchicus. Marine Ecology chlorophyll on the communities of zooplankton in the southeastern Bering Sea. Progress Series 176, 81–92. Journal of Marine Research 42, 221–257. Ohman, M.D., Runge, J.A., 1994. Sustained fecundity when phytoplankton resources Stenevik, E.K., Melle, W., Gaard, E., Gislason, A., Broms, C.T.A., Prokopchuk, I., are in short supply: Omnivory by Calanus finmarchicus in the Gulf of St. Ellertsen, B., 2007. Egg production of Calanus finmarchicus-A basin-scale Lawrence. Limnology and Oceanography 39, 21–36. study. Deep Sea Research Part II: Topical Studies in Oceanography 54, 2672– Ohman, M.D., Runge, J.A., Durbin, E.G., Field, D.B., Niehoff, B., 2002. On birth and 2685. death in the sea. Hydrobiologia 480, 55–68. Strickland, J.D.H., Parsons, T.R., 1972. A practical Handbook of Sea-water Analysis: Osgood, K.E., Frost, B.W., 1994. Comparative life histories of three species of Ottawa, 12. planktonic calanoid copepods in Dabob Bay, Washington. Marine Biology 118, Uye, S., 1996. Induction of reproductive failure in the planktonic copepod Calanus 627–636. pacificus by diatoms. Marine Ecology Progress Series 133, 89–97. Paffenhofer, G.-A., 2002. An assessment of the effects of diatoms on planktonic Uye, S., Murase, A., 1997. Relationships of egg production rates of the copepod copepods. Marine Ecology Progress Series 227, 305–310. Calanus sinicus to phytoplankton availability in the Inland Sea of Japan. Plankton Paffenhofer, G.A., Ianora, A., Miralto, A., Turner, J.T., Kleppel, G.S., d’Alcala, M.R., of Biology and Ecology 44, 3–11. Casotti, R., Caldwell, G.S., Pohnert, G., Fontana, A., Muller-Navarra, D., Vargas, C.A., Escribano, R., Poulet, S., 2006. Phytoplankton food quality determines Jonasdottir, S., Armbrust, V., Bamstedt, U., Ban, S., Bentley, M.G., Boersma, M., time windows for successful zooplankton reproductive pulses. Ecology 87, Bundy, M., Buttino, I., Calbet, A., Carlotti, F., Carotenuto, Y., d’Ippolito, G., Frost, 2992–2999. B., Guisande, C., Lampert, W., Lee, R.F., Mazza, S., Mazzocchi, M.G., Nejstgaard, J. Vargas, C.A., Martínez, R.A., Cuevas, L.A., Pavez, M.A., Cartes, C., GonzÁlez, H.E., C., Poulet, S.A., Romano, G., Smetacek, V., Uye, S., Wakeham, S., Watson, S., Escribano, R., Daneri, G., 2007. The relative importance of microbial and Wichard, T., 2005. Colloquium on diatom-copepod interactions. Marine Ecology classical food webs in a highly productive coastal upwelling area. Limnology Progress Series 286, 293–305. and Oceanography 52, 1495–1510. Peterson, W., Gomez-Gutierrez, J., Morgan, C.A., 2002. Cross-shelf variation in Vargas, C.A., Martinez, R.A., Escribano, R., Lagos, N.A., 2010. Seasonal relative Calanoid copepod production during summer 1996 off the Oregon coast, USA. influence of food quantity, quality, and feeding behaviour on zooplankton Marine Biology 141, 353–365. growth regulation in coastal food webs. Journal of the Marine Biological Peterson, W.T., 1980. Life History and Ecology of Calanus marshallae Frost in the Association of the United Kingdom 90, 1189–1201. Oregon Upwelling Zone. PhD. Thesis. Oregon State University, 200pp. Vehmaa, A., Larsson, P., Vidoudez, C., Pohnert, G., Reinikainen, M., Engstrom-ost, J., Peterson, W.T., 1986. Development, growth, and survivorship of the copepod 2011. How will increased dinoflagellate:diatom ratios affect copepod egg Calanus marshallae in the laboratory. Marine Ecology Progress Series 29, 61–72. production?– A case study from the Baltic Sea. Journal of Experimental Marine Peterson, W.T., 1988. Rates of egg production by the copepod Calanus marshallae in Biology and Ecology 401, 134–140. the laboratory and in the sea off Oregon, USA. Marine Ecology Progress Series Wang, S., Li, C., Sun, S., Ning, X., Zhang, W., 2009. Spring and autumn reproduction 47, 229–237. of Calanus sinicus in the Yellow Sea. Marine Ecology Progress Series 379, 123– Peterson, W.T., Kimmerer, W.J., 1994. Processes controlling recruitment of the 133. marine Calanoid copepod Temora longicornis in Long Island Sound: Egg Wichard, T., Poulet, S.A., Boulesteix, A.-L., Ledoux, J.B., Lebreton, B., Marchetti, J., production, egg mortality, and cohort survival rates. Limnology and Pohnert, G., 2008. Influence of diatoms on copepod reproduction. II. Oceanography 39, 1594–1605. Uncorrelated effects of diatom-derived a, b, c, d-unsaturated aldehydes and Pierson, J.J., Halsband-Lenk, C., Leising, A.W., 2005. Reproductive success of Calanus polyunsaturated fatty acids on Calanus helgolandicus in the field. Progress in pacificus during diatom blooms in Dabob Bay, Washington. Progress in Oceanography 77, 30–44. Oceanography 67, 314–331. Zhang, G.-T., Sun, S., Zhang, F., 2005. Seasonal variation of reproduction rates and Plourde, S., Campbell, R.G., Ashjian, C.J., Stockwell, D.A., 2005. Seasonal and regional body size of Calanus sinicus in the Southern Yellow Sea, China. Journal of patterns in egg production of Calanus glacialis/marshallae in the Chukchi and Plankton Research 27, 135–143.