The effects of irradiance and nutrient supply on the productivity of Arctic waters a perspective on climate change

Jean-Éric Tremblay and Jonathan Gagnon

Département de Biologie and Québec Océan, Pavillon Alexandre-Vachon, Université Laval, Québec, QC, Canada G1V OA6, [email protected]

Abstract

A previous analysis of published data suggested that annual, pelagic primary production in the Arctic Ocean is related linearly to the duration of the ice-free period, presumably through cumulative exposure to solar irradiance. However, the regions with the longest ice-free periods are located in peripheral seas and polynyas where nutrient supply by advection or the vertical mixing induced by winds and convection can be extensive. The ensuing replenishment of nutrients drives primary production to levels unattained in the strongly stratified interior (e.g. the Beaufort Sea), with the exception of upwelling areas. A reanalysis of published data showed no relation between cumulative production and incident solar radiation during the growth season. We propose that changes in annual primary production per unit area in seasonally ice-free waters are controlled primarily by the environmental forcing of nitrogen supply. Incidental changes in light regime should mostly affect the timing and, possibly, the species composition of the main production pulse(s) in the upper mixed layer and, underneath, the ability of phytoplankton to exploit nutrients in the lower euphotic zone. While the ongoing rise in the supply of heat and freshwater to the Arctic Ocean should bolster vertical stratification and further impede the mean upward supply of nutrients, episodic yet direct atmospheric forcing of the upper ocean may act in synergy with a prolonged exposure to light and greatly augment pelagic productivity.

Introduction

The physical environment of the Arctic Ocean is changing rapidly and profoundly (ACIA 2005). The most spectacular manifestation is the decline in the minimum extent of sea ice in September and the thinning of the multi-year ice that remains

(Kwok et al. 2007; Maslanik et al. 2007; Stroeve et al. 2008). The ice cover is becoming essentially seasonal and, in some region, melts increasingly early or forms increasingly late (Lemke et al. 2007). River discharge is on the rise, augmenting freshwater content and the loading of organic and inorganic matter into the coastal zone (Peterson et al. 2006). The intensity and frequency of extreme wind events increased in several regions (Zhang et al. 2004), which will impact upper ocean dynamics under a reduced-ice scenario.

Quantifying and predicting the response of marine primary production to these alterations is a prerequisite to understand their impacts on Arctic food webs, biogeochemical cycles and air-sea fluxes of climate-active gases. One obvious consequence of the ongoing changes is that the amount of light reaching the ocean surface increases. Much less obvious is how this energy subsidy interacts with concomitant perturbations of the water column to alter the magnitude and species composition of pelagic primary production. Autotrophic growth requires light, but it also necessitates nutrients whose inventories are limited in the upper ocean. The ideal mixture of nutrients for growth varies across phytoplankton taxa and the ratios of the different nutrients differ widely among the source waters that pervade the Arctic Ocean (Tremblay et al. 2002b). A fraction of these nutrients is recycled locally by biological activity (i.e. excretion, decomposition) but changes in overall productivity must be sustained by allochthonous nitrogen (N) subsidies to the euphotic zone. This input may originate from deep waters as nitrate, from the atmosphere as N2 or from precipitation and rivers as inorganic or organic N.

In the present context of rapid change, it is worth asking if generalizations can be made from the data gathered in the last 50 years to help bracket predictions and provide a perspective for the future. A few reviews of primary production have been published for Russian Seas (Vetrov and Romankevich 2004), the Arctic in general (e.g. Sakshaug 2004; Legendre et al. 1992; Smith and Sakshaug 1990) and polynyas (Arrigo 2007; Tremblay and Smith 2007). These reviews document the diversity and complexity of Arctic marine ecosystems, contrasting regions where ice dynamics and physical oceanography differ broadly (see also Carmack and Wassmann 2006).

In an effort to synthesize pan-Arctic data, Rysgaard et al. (1999) obtained a positive correlation between annual primary production and the duration of the ice-free period. Taken at face value, this relationship suggests that primary production would, in the future, increase in direct proportion with the lengthening of the ice-free season. Here we assess whether the correlation proposed by Rysgaard et al. (1999) has predictive value by investigating the causal, underlying factors. We begin by considering the role of irradiance and nutrients separately and conclude with a few perspectives and hypotheses for future research.

Data mining

Our review of the literature considered the studies used by Rysgaard et al. (1999), Sakshaug (2004) and Vetrov and Romankevich (2004) in addition to recent work in Baffin Bay and the Beaufort Sea (Klein et al. 2002; Simpson et al. 2008). The synthesis published by Rysgaard et al. (1999) reported annual values of primary production but none of the contextual physical and chemical parameters needed for an in-depth analysis of controlling factors. The latter requires information on the bathymetry of sampling stations, nutrient supply and the date of the first and last seasonal measurement of primary production (needed to estimate incident irradiance during the measurement period).

Albeit informative in their own right, several of the studies considered were excluded from the analysis because they were deemed non comparable. Sampling sites with bottom depths and salinities of less than 5 m and 5%o, respectively (Horner and Schrader 1982; Kangas et al. 1993; Meskus 1976) belong to neritic, intertidal or fluvial ecosystems and were dismissed. Model results were also discounted (e.g. Slagstad and Stole-Hansen 1991; Taguchi 1972; Wassman et al. 2006) as they cannot be considered on the same footing as actual measurements in first analysis.

Two levels of data synthesis were produced. Level 1 concerns only those studies where populated time series (i.e. measurements made at frequent time intervals in a constrained region or water mass) of primary production were available. Time series that did not resolve the peak and tail ends of the main production pulse (Grainger 1980 - for 1968; Walsh et al. 1989; Alexander 1974; McRoy and Goering 1976; Nielsen and Hansen 1999) were excluded from the level 1 data base. For purposes of concision, clarity and comparison, the time courses are reported and compared as cumulative primary production (EP) values obtained by leap-frog integration between successive time points. Level 2 gives annual primary production values estimated by interpolation and/or extrapolation, preferably by the authors themselves or by us when the calculation was not done originally. In some cases this was done by combining data from 2 or more years. We assume that the level of confidence attached to the different estimates is much higher for level 1 than for level 2 since extrapolations do not take into account possible changes in the time courses when measurements stopped long before the ice cover was re-established. The number of studies considered for level 1 (9 studies and 13 distinct time courses) is lower than the number of studies considered for level 2 (n = 12).

In most studies, data on incident or underwater solar radiation are either not provided or not usable for comparative purposes. Daily averages of incident shortwave radiation (W m-2) corrected for cloud cover were retrieved from the NCEP/NCAR reanalysis. For each time series, cumulative radiation at the sea surface was calculated by multiplying daily averages in watts per square meter by 86,400 s and summing these daily doses over 60 days or the full time course.

Unless stated otherwise, results are expressed in GJ (Giga Joules) m-2. Information on the initial concentrations of nutrients prior to the growth season was absent from a few studies. In these cases, concentrations were taken from anterior or posterior studies in the area or from the World Ocean Database 2005 (Boyer et al. 2006).

General properties of the data set

The general location of the studies used for level 1 and 2 analyses is shown in Fig. 1 and the characteristics of the each sampling site are given in Table 1. The study sites cluster in the western Arctic, an unintended regional bias, and range broadly in latitude (60-82.5° N) depth (15-450 m, taken as the average for a given sampling region) and year day of the first measurement of primary production (from late March to early July).

Fantasy World Map
Fig. 1. Map of the sampling sites used for the analysis of detailed time series (black circles) and annual estimates (all symbols) of primary production. Numbers refer to the studies described in Table 1.
Table 1. General characteristics of the studies retained for level 1 and 2 analyses. Missing values (-) in the Start and End columns indicate that no time course data on primary production are available (only annual estimates).

ID Year

Location

Latitude Longitude Depth (°N) (°W or °E) (m)

Start End (year day) (year day)

Reference

1a

1959

Disko Bay (W. Greenland)

69.250

53.567° W 30

129

141

Petersen 1964

1b

1960

Disko Bay (W. Greenland)

69.250

53.567° W

93

165

Petersen 1964

2

1959

Dumbell Bay (Lincoln Sea)

82.517

62.167° W 25

172

66

Apollonio 1980

3

1968

Bering Sea

60.000

170.000° W 100

-

-

McRoy and Goering 1976

4

1968

N. Greenland

82.167

31.233° E 26

-

-

Andersen 1981

5a

1968

Frobisher Bay

63.667

68.450° W 50

176

110

Grainger 1980

5b

1969

Frobisher Bay

63.667

68.450° W

167

82

Grainger 1980

6

1971

Resolute Bay

74.684

94.868° W 15

181

102

Welch and Kalff 1975

7a

1973

Disko Bay (W. Greenland)

69.167

53.500° W 50

171

119

Andersen 1977

7b

1974

Disko Bay (W. Greenland)

69.167

53.500° W

138

168

Andersen 1977

7c

1975

Disko Bay (W. Greenland)

69.167

53.500° W

81

77

Andersen 1977

8a

1976

Beaufort Sea

71.000

148.000° W 30

-

-

Carey 1978

8b

1977

Beaufort Sea

71.000

148.000° W

-

-

Carey 1978

8c

1976

Chuckchi Sea

71.800

156.000° W 40

-

-

Carey 1978

8d

1977

Chuckchi Sea

71.800

156.000° W

-

-

Carey 1978

9

1982

Barents Sea

76.000

32.500° E 200

98

181

Rey et al. 1987

10

1996

Young Sound (E. Greenland)

74.310

20.251° W 35

174

100

Rysgaard et al. 1999

11

1998

NE. Baffin Bay

76.000

74.000° W 450

98

104

Klein et al. 2002

12

1993

NE. Greenland Sea

80.230

13.000° W 200

-

-

Klein et al. 2002

13

2004

SE. Beaufort Sea

70.807

125.393° W

152

84

Simpson et al. in press

Although the date of the first measurement differs broadly among studies, the different time courses (level 2) were plotted with a common time origin in order to directly compare the influence of time on EP (Fig. 2). One striking result is that in the most productive systems EP tends to reach an asymptote whereas in the others it generally increases in a roughly linear fashion. The latter observation challenges the general notion that Arctic systems are strongly pulsed in time with a peak period of production lasting only a few weeks. Another noteworthy aspect of Fig. 2 is that although the initial rates of increase vary, the time courses rapidly assume their final, relative rank. Such layering implies the absence of a single time continuum along which EP values lay as the duration of the open water period increases. In other words, the level of primary production in these systems is largely conditioned or set at the beginning and overall differences in EP are not explained by the duration of the open water period per se. As much as 85% of the variability in EP end points is predictable after only 60 days, despite measurement periods lasting from 3 to 8 months.

3 75

50 25

0 30 60 90 120 150 180 Days elapsed since first measurement

Fig. 2. Time courses of cumulative primary production (EP), starting on the first day of measurement for each time series. The different symbols refer to the ID numbers of the distinct studies and sampling years given in Table 1: 1a (open circle), 1b (open square), 2 (open diamond), 5a (open hexagon), 5b (open triangle), 6 (open inverted triangle), 7a (gray circle), 7b (gray square), 7c (gray diamond), 9 (gray hexagon), 10 (gray triangle), 11 (black circles), and 13 (black square).

0 30 60 90 120 150 180 Days elapsed since first measurement

Fig. 2. Time courses of cumulative primary production (EP), starting on the first day of measurement for each time series. The different symbols refer to the ID numbers of the distinct studies and sampling years given in Table 1: 1a (open circle), 1b (open square), 2 (open diamond), 5a (open hexagon), 5b (open triangle), 6 (open inverted triangle), 7a (gray circle), 7b (gray square), 7c (gray diamond), 9 (gray hexagon), 10 (gray triangle), 11 (black circles), and 13 (black square).

The role of irradiance

The ice cover and extreme solar cycle impose severe constraints on primary production. These limits are most severe toward the pole, where the vanishing multi-year ice conspires with the polar night to make photosynthesis impossible throughout most of the winter. Conditions are less stringent at lower latitudes where variable ice dynamics and irradiance regimes create a host of transient niches for primary producers. The replacement of multi-year ice by seasonal ice is likely to create a more favourable light climate at higher latitudes. Light plays two roles in this respect. It determines when net phytoplankton growth is possible and how much cumulative energy is available for photosynthesis during the course of the growth season.

Once light has reached the water its availability for phytoplankton is determined by a combination of factors that includes mixing, which sets the mean irradiance seen by the cells and affects water transparency via sediment re-suspension on shallow shelves (e.g. Forest et al. 2008), the biomass of phytoplankton and horizontal inputs of dissolved and particulate organic matter in coastal regions. The interplay between incident irradiance, water transparency and mixing depth in determining the potential for phytoplankton growth has been the subject of empirical considerations that will not be repeated here (Nelson and Smith 1991; Sakshaug and Slagstad 1991). It is noteworthy that the data sets used in our analysis seldom include the information needed to estimate mean irradiance in the mixed layer.

Onset of the productive season

Net primary production is considered possible only when phytoplankton have enough light for positive growth, generally in accordance with Sverdrup's critical depth model as revised by Smetacek and Passow (1990). Although communities located at the ice-water interface or at the top of the euphotic zone may not be vigorously mixed, they require enough light to outgrow losses, i.e. irradiance must be higher than the compensation light intensity below which the sum of phytoplankton respiration and other losses (grazing, lysis and sinking) exceeds the gain of carbon by photosynthesis. Recent empirical estimates indicate that the compensation irradiance of diatom communities in the Arctic is similar to the mean for the north Atlantic (1.3-1.9 mol quanta m-2 day1; Tremblay et al. 2006b).

Evm Scatter Block Wlan
Fig. 3. Relationship between the mean, downwelling short-wave radiation at the sea surface during the first week of measurements and (A) the year-day of the first measurement, and (B) the latitude of the sampling site.

Not surprisingly, the locations that open early during the year receive much less irradiance initially (Fig. 3A). The mean incident short-wave radiation at the surface during the first week of measurement is independent of latitude (Fig. 3B), indicating that the onset of the productive period is controlled essentially by regional oceanic processes and ice dynamics instead of the solar cycle. Once there is enough light for net growth the initial rate of development of a bloom should depend on the exposure of micro-algae to irradiance and their ability to harvest this light efficiently, which is influenced by the physiological state of the cells and, possibly, temperature (Tremblay et al. 2006b).

Cumulative irradiance

The pattern shown in Fig. 4A is counter-intuitive. The systems with the highest levels of primary production after 60 days (the point where all systems can be compared) received much less light than the least productive ones. This results is consistent with the inability of diatoms to harvest light immediately after an abrupt switch from near complete obscurity to full sunlight after rapid ice ablation (Tremblay et al. 2006b), a problem that is likely to be most acute in systems that open late and close to the summer solstice.

How the algae respond to light at the onset of the growth season presumably depends on how fast the transition occurs, but also on the different photoadaptation and photoacclimation strategies of the seed organisms present (Sakshaug and Slagstad 1991). The photoinhibition of planktonic algae has sometimes been regarded as an experimental artefact of incubations, whereby the cells are contrived to remain under high light for longer than they do under mixed, in situ conditions. The advent of rapid, non-intrusive active fluorescence techniques showed that photoinhibition is often severe in these communities, even when exposed to light flashes of a few micro-seconds. In fact, this short-term photoinhibition is often more acute than is apparent from incubations spanning a few hours to a day (either in situ or simulated in-situ), suggesting that in the latter, the algae have some time to adjust and recover, which does not intervene during near-instantaneous measurements.

As expected, the apparent negative effect of irradiance disappears when considering time courses in their entirety. Nevertheless, the total amount of radiation (EE) received during the time courses bears no relationship to EP, either for the duration of each time course (Fig. 4B) or on an annual basis (Fig. 4C). Note also that in Fig. 4B, EP ranges by an order of magnitude for EE values of 1.5-2.1 GJ m-2. Although rigorous comparisons based on mean irradiance in the mixed layer are not possible in retrospect, it is clear that differences in irradiance dose are by far insufficient to account for the range of EP across seasonally-open systems.

Fig. 4. Relationships between cumulative primary production (EP) and cumulative downwelling, short-wave radiation at the sea surface (EE) for (A) the first 60 days of each time course, (B) the total duration of each time course, and (C) 1 year for EP with EE calculated for the time course only (B). Symbols as per Fig. 2.

Fig. 4. Relationships between cumulative primary production (EP) and cumulative downwelling, short-wave radiation at the sea surface (EE) for (A) the first 60 days of each time course, (B) the total duration of each time course, and (C) 1 year for EP with EE calculated for the time course only (B). Symbols as per Fig. 2.

The role of nutrients

Allochthonous nitrate

Nutrients are supplied to the upper Arctic Ocean by a variety of processes that operate at different scales of time and space. Horizontal inputs are provided by rivers and currents originating from adjacent seas. Advection is especially important immediately downstream of gateways to the Pacific and Atlantic, where it can support intense primary production, especially in the Chukchi Sea. Surface waters reside in the central Arctic for ca. 10 years and, unless their transit occurs exclusively under multi-year ice, labile nutrients are readily consumed at the periphery during the first year. This scenario also applies to the inorganic nutrients delivered by major rivers. For example, the Mackenzie River supplies a lot of silicate to the Arctic Ocean, but any residual, inorganic phosphate and nitrogen is exhausted before the freshwater plume advances into the sea (Emmerton et al. 2008; Simpson et al. 2008).

Denitrification over the shallow shelves of the subarctic Pacific Ocean and the Arctic, in combination with high silicate loading by rivers insures that phosphorus and silicon remain in excess, especially where Pacific waters dominate. Nitrogen is the primary yield-limiting nutrient in the Arctic as it was exhausted first in all blooms investigated so far (Kattner and Budeus 1997; Simpson et al. 2008; Tremblay et al. 2002b, 2008), even in the Barents Sea where concentrations of silicate and phosphate are relatively low (Reigstad et al. 2002).

Physical singularities episodically subsidize a given region with nutrients. These singularities can take the form of internal waves, storms that erode the halocline, shelf-break upwelling and dynamic instabilities caused by bathymetry or convection (Mathis et al. 2007; Williams et al. 2006; Zhang et al. 2004; Tremblay et al. 2002a). The incidence and strength of upwelling events and halocline perturbations are presumably increasing with the rising frequency and intensity of cyclones (Yang et al. 2004) and the retreat of the perennial ice pack beyond the shelf break (Carmack and Chapman 2003). It is not currently possible to assess or forecast the net result of changes in the mean versus episodic deliveries of nutrients on the magnitude and species composition of primary production.

The total availability of nutrients to primary producers is the sum of the initial inventory in the euphotic zone (end of winter) and any inputs that occur once the growth season is initiated. Because these subsidies are difficult to quantify and were seldom considered in the papers used for level 1 and 2 analyses, we evaluated only the initial inventories of nutrients present at or prior to the onset of the growth season.

It is clear from the depth distribution of studies (Table 1) that vertical integration must be standardized for comparison. Time series obtained over deep waters indicate that pronounced seasonal nutrient deficits frequently extend to 50-75 m (Tremblay et al. 2002a; Smith et al. 1997), due to either episodic vertical mixing during the growth season or the ability of phytoplankton to thrive within subsurface chlorophyll maxima (Tremblay et al. 2008). In this view, systems with bottom depths of less than 50 m cannot yield the same new production because of their limited nutrient inventories and the vertical proximity or overlap between sediment denitrification and the euphotic zone. In Young Sound, as example, the shallow bottom (35 m) and associated denitrification limit nitrate concentrations to 2 |M during winter (total spring time inventory of 70 mmol m-2) (Rysgaard et al. 2004). Phytoplankton in waters less than 40 m in depth typically have access to initial nitrate inventories of 150 ± 135 mmol m-2, whereas those in deeper waters can potentially tap into as much as 1,000 mmol m-2 (Fig. 5). In order to perform a legitimate comparison of the nutrients available to phytoplankton, late-winter inventories were estimated for the upper 75 m over deep waters or for the whole water column at sites shallower than 75 m.

Fig. 5. Relationship between the initial concentration of nitrate at the onset of the growth season and (A) the year-day of the first measurement for level 1 and 2 data, and (B) the estimated duration of the ice-free period for all data. Closed and open circles indicate bottom depths of less or more than 30 m, respectively.

Fig. 5. Relationship between the initial concentration of nitrate at the onset of the growth season and (A) the year-day of the first measurement for level 1 and 2 data, and (B) the estimated duration of the ice-free period for all data. Closed and open circles indicate bottom depths of less or more than 30 m, respectively.

Although the shallowest waters have the smallest nitrate inventories, bathymetry alone explains 30% (p < .05) of the overall variability in the initial availability of nitrate, indicating that physical processes play a determinant role at deep sites. As example, systems located in the Atlantic sector of the Arctic (e.g. eastern North Water and Barents Sea, i.e. studies 9 and 11 in Table 1) are usually advantaged by the greater susceptibility of surface waters to vertical mixing relative to the strongly stratified Pacific-derived waters (e.g. Gulf of Amundsen, study 13 in Table 1) (see also Tremblay et al. 2008). Overall, the initial nitrate inventory shows a negative relationship (r2 = .36, p < .05) with the year-day of the first measurement (Fig. 5A) and a positive one (r2 = .41, p < .01) with the estimated duration of the ice-free season (Fig. 5B). This pattern implies that systems where surface waters are exposed to the atmosphere for a longer time experience a larger nitrogen load. The likely cause of this effect is that systems with a long ice-free period such as the North

Water or the Barents Sea are exposed to storms during late winter or autumn and convection that erode the weak vertical stratification and replenish nutrients (Tremblay et al. 2002a). Waters that open in late spring or summer (e.g. the Northeast Water) when winds are weak and the sun induced melt-water stratification is strong see little vertical replenishment of nutrients on an annual basis (Kattner and Budeus 1997). Parts of the Barents Sea that lay along storm tracks in summer also experience significant nutrient renewal where haline stratification is moderate (Sakshaug and Slagstad 1992; Wassman et al. 2006). In extreme cases, even the strong stratification that characterizes waters of the Canada Basin can be overturned, thus providing an ample supply of nutrients to an otherwise impoverished upper euphotic zone. This condition was observed in northwest Baffin Bay, but the underlying mixing process has yet to be confirmed (Tremblay et al. 2002b).

The initial inventory of nitrate in the upper 75 m proved to be a reasonably robust predictor of EP at 60 days (r2 = .53, p < .01; Fig. 6A) and at the end of each time course (r2 = .69, p < .001; Fig. 6B). These relationships explain why annual production can be predicted from cumulative values after 60 days, i.e. the overall level of primary production is largely pre-determined by the initial nutrient load although light may influence the initial rates of drawdown. The relationship between nitrate loading also held for annual P, but this time the North Water clearly stood out as a positive anomaly. Removing this outlier from the analysis yields a very robust correlation (r2 = .78, p < .0001; Fig. 6C). The anomalously high level of annual primary production in the North Water was caused by episodic mixing, a fall bloom and continued productivity through October (Garneau et al. 2007; Klein et al. 2002), which underscores how the initial inventories of nitrate provide only a limited view of nitrogen supply in some systems. Another example of this limitation is found in some regions of the Bering Sea, where persistent nutrient renewal leads to unusual productivity levels similar to and higher than those observed in the North Water (e.g. Sambrotto et al. 1984; Springer et al. 1996). Clearly, the inshore portions of shallow Arctic shelves do not experience a significant late-season, vertical re-supply of nitrate simply because the deep reservoir is absent. At the shelf break or in deeper regions, episodic mixing and upwelling may augment primary productivity far beyond the levels expected from initial conditions. It follows that, on a regional basis, the effects of climate change on primary productivity will greatly depend on bathymetry and distance from the shelf break, rivers and the adjacent Pacific and Atlantic oceans.

A multiple regression model that included EE in addition to the inventory of nitrate only slightly improved the prediction of EP, and the partial coefficient for the effect of irradiance was not significant (not shown). It is clear in this context that nitrogen loading is the primary control on the productivity of Artic waters. Considering the few studies were temperature data were available, we were unable to find a relationship between temperature and the initial increase in EP (see also Tremblay et al. 2006b).

Fig. 6. Relationships between the initial concentration of nitrate at the onset of the growth season and EP for (A) the first 60 days of each time course, (B) the total duration of each time course, and (C) 1 year. Symbols as per Fig. 2 except for panel C where crossesrepresent studies for which annual estimates but no time courses are available (ID 3, 4, 8a-d, and 12 in Table 1).

Fig. 6. Relationships between the initial concentration of nitrate at the onset of the growth season and EP for (A) the first 60 days of each time course, (B) the total duration of each time course, and (C) 1 year. Symbols as per Fig. 2 except for panel C where crossesrepresent studies for which annual estimates but no time courses are available (ID 3, 4, 8a-d, and 12 in Table 1).

The contribution of new and regenerated production

The allochthonous N assimilated by autotrophs (i.e. new production) is eventually channeled into waste or lost by transport or sinking. Some of the waste products that remain in the euphotic zone can be re-used directly or readily broken down into usable N forms by photochemistry or bacteria. The N thus recycled drives regenerated production, which contributes to total primary production but cannot sustain net increases in the yield of organic matter, ecosystem productivity and vertical carbon export (Dugdale and Goering 1967; Eppley and Peterson 1979). The relative importance of new and regenerated production must be resolved to understand what drives pan-Arctic differences in total production and what these changes signify for the yield of harvestable resources and air-sea fluxes of climate-active gases.

As the duration of the ice-free season increases, there is a distinct possibility that EP also does simply because N cycles more times in the euphotic zone. To investigate this possibility it is useful to consider published estimates of the /-ratio, i.e. the ratio of allochthonous N uptake to total N uptake. Ideally, the uptake of all sources of allochthonous and recycled N should be considered in the /-ratio, but only the net uptake of nitrate, ammonium and, sometimes, urea are routinely estimated.

Comparable, populated time series of the /-ratio that resolve a major portion of the growth season remain extremely rare. This data is available for the North Water (/-ratio = 0.58; Tremblay et al. 2006a) and the Northeast Water (/-ratio = 0.65; Smith et al. 1997), where repeated measurements were made over a similar period of ca. 3 months starting in spring. These /-ratios are remarkably similar despite the threefold difference in initial nitrate inventories and the fourfold difference in annual production between the two systems (252 g C m-2 for the North Water versus 66 g C m-2 for the Northeast Water). The comparison shows that regenerated production accounts for a substantial share of total primary production, but is merely proportional to new production and the supply of allochthonous N. Since the two systems were characterized by negligible inventories of ammonium and urea at the onset of the growth season, a transformation of the /-ratio, (1 - /)// can be used to estimate the number of times the N initially bound in nitrate was recycled during the 3-month period, which yields low values of 0.6-0.7 . In the North Water, a net amount of 920 mmol N m-2 was consumed between 23 April and July 1 (Tremblay et al. 2006a). By then, only 100 mmol N m-2 remained as PON and 22 mmol N m-2 as ammonium + urea in the euphotic zone (Tremblay et al. 2002a, 2006a). This rough budget implies that up to 87% of the N initially taken up as nitrate resided in the euphotic zone for less than 68 days, which is short relative to the duration of the ice-free period (ca. 6 months). In this view, a protracted ice-free season should promote regenerated production insofar as there is a continual or pulsed supply of external N to maintain it.

Implications, perspectives and future research

Assessments of the current and future productivity of the Arctic Ocean should distinguish between rate and yield-limiting processes, and between new and regenerated production. Light and the physiological state of phytoplankton influence the rate at which nutrients are consumed early (Tremblay et al. 2006b) and late (Garneau et al. 2007) in the year, whereas the supply of external nitrogen to the euphotic zone sets the upper yield of organic matter production (Tremblay et al. 2006b). Offshore, the ablation of multi-year ice will initially increase new production per unit area where residual nutrients linger, but the stimulation will lessen in subsequent years unless the new physical regime promotes recurrent nutrient renewal through vertical mixing, upwelling or eddy genesis. Over shallow shelves with a limited inventory of inorganic N and a secular, seasonal ice cover, it is doubtful that a protracted ice-free period will augment new production per unit area away from shelf breaks, rivers and adjacent oceans. In essence, the greater extent of seasonally-open water should increase basin-scale primary production but, beyond the initial transition from multi-year to seasonal ice, cannot in and off itself lead to a sustained, order-of-magnitude increase in EP per unit area (c.f. Rysgaard et al. 1999) in currently unproductive areas.

The environmental changes at work in the Artic may very well alter the upward supply of N to the euphotic zone. Whether these alterations result in a subsidy or an impoverishment in a given region will largely depend on how changes in the freshwater balance interact with atmospheric forcing of the upper ocean. While warming, precipitation and the increasing delivery of freshwater should increase vertical stratification and further curtail the upward flux of nutrients in the coastal zone, evaporation and the export of multiyear ice offshore are conducive to salinization (Polyakov et al. 2007), which by weakening stratification of the surface layer could make it more susceptible to mixing by storms. This susceptibility is also increased by the contraction of the ice-covered period, which submits surface waters to direct wind forcing during early spring or late fall (e.g. Tremblay et al. 2002a).

The role of advection and rivers as direct, horizontal sources of nutrients remains difficult to assess. Historically, rivers did not deliver large amounts of inorganic N and phosphorous to the Arctic Ocean, but rising discharge and contributions of melted permafrost to affluent waters may increase nutrient load in coastal waters (Frey et al. 2007; McClelland et al. 2007). This subsidy, however, is likely to be local in nature since inorganic nutrients are readily consumed inshore. The scenario may differ for dissolved organic nitrogen (DON) that escapes early consumption and potentially sustains a portion of primary production in coastal waters.

Recent data from the southeast Beaufort Sea shows that the net depletion of inorganic phosphorus and dissolved inorganic carbon (DIC) continues after nitrate is exhausted (Simpson et al. 2008; Tremblay et al. 2008), which suggests that autotrophs have access to an alternate source of allochthonous N. Some of the DON supplied by rivers can be used directly by some phytoplankton or made available by photochemical or bacterial attack of non-readily labile compounds (e.g. Vahatalo and Zepp 2005). New evidence implies that the input and biological availability of dissolved organic matter during the spring freshet is higher than previously thought on the basis of summer data (Holmes et al. 2008). The excess phosphorus and low availability of nitrate in the high Arctic also makes it a fertile ground for N2 fixation. Yamamoto-Kawai et al. (2006) suggested that the excess phosphorus sustains N2 fixation downstream in the North Atlantic, but it might also be so within the Arctic proper, especially when and where conditions (e.g. temperature) are favourable for diazotrophs. This hypothesis remains to be confirmed experimentally, but it is plausible that regional N2 fixation will gain in importance as the upper Arctic warms and becomes more stratified.

Much of the literature dealing with river discharge understandably focuses on dissolved organic carbon (DOC) inputs and their impact on ocean biogeochemistry, including the heterotrophic and photochemical release of DIC (e.g. Bélanger et al. 2006). The effect of this release on air-sea fluxes of CO2, however, could be mitigated by the incidental increase of net primary production resulting from a relaxation of N limitation. It is possible, however, that declining water transparency could trigger a shift from nutrient limitation to light limitation over shallow shelves. An integrated study of these aspects in addition to N2 fixation, nitrification and sediment denitrification is warranted given their potential impact on the air-sea exchange of CO2 and N2O (e.g. Duce et al. 2008).

Ongoing changes in the physical environment will probably affect taxonomic dominance during production pulses (e.g. Arrigo et al. 1999; Walsh et al. 2004), but this topic remains severely understudied in the Arctic. Apart from peripheral seas (i.e. Bering, Barents, Labrador and Greenland), where Emiliana huxleii or colonies of Phaeocystis pouchetii can reach high biomasses, documented phytoplankton blooms in the interior of the Arctic Ocean have been dominated by diatoms so far. No satisfactory explanation has been proposed for this geographical segregation but, as in the Southern Ocean, distinct adaptations to light regime (Arrigo et al. 2007) possibly play a part. More research is needed on this question since changes in taxonomic dominance have profound consequences for food webs, biogeochemical fluxes and climate feedbacks (e.g. Tozzi et al. 2004).

Another unknown is how recurrent reductions in the duration of the seasonal ice cover and the production of ice algae will affect sympagic-pelagic-benthic interactions. Algae from the water column can either be trapped in new ice or colonize the bottom of thicker first-year ice during winter and spring, but the converse is much less obvious. On the one hand, there is evidence that ice algae, when released from the ice, sink promptly to the bottom and provide food to benthic communities, especially over shallow continental shelves (Legendre et al. 1992). On the other hand, retention of sloughed ice algae in the upper water column has been observed (Michel et al. 1993) and inferred to provide significant energy to pelagic grazers on the basis of carbon budgets (Michel et al. 1996). The extent to which ice-grown algae currently seed and shape pelagic blooms is an open question. Although it may not affect the overall yield of organic matter during the growth season, a change in the timing and species composition of pelagic blooms is likely to alter food webs and the success of herbivores with less flexible life histories.

In conclusion, this review showed that the positive correlation between the duration of the ice-free period and annual primary production observed by Rysgaard et al. (1999) does not have predictive power per se. This correlation was essentially driven by bathymetry and the differential susceptibility of distinct oceanic provinces to vertical mixing. The productivity of shallow or strongly stratified Arctic waters will not catch up with that of weakly stratified or upwelling-prone regions simply because the duration of the growth season increases. Such an increase will of course modulate annual productivity within a given body of water, notably by increasing production in the subsurface chlorophyll maximum, but the change will not measure up to the order-of-magnitude range in EP observed at the pan-arctic scale. Almost any short-term response is possible within this envelope, as recently shown by Arrigo et al. (2008) for the 2007 growth season. However, sustained increases in primary production per unit area will only occur where the greater exposure of surface waters to solar radiation is matched by annually recurring nutrient subsidies. Whether this occurs or not essentially depends on the net result of concomitant changes in the freshwater balance, horizontal nutrient loading and atmospheric forcing of the upper Arctic Ocean.

Acknowledgments

We thank Danny Dumont for his timely help with the NCEP reanalysis of solar radiation data. This work was supported by grants to JET from the Natural Sciences and Engineering Research Council of Canada and the ArcticNet Network Center of Excellence, and is a contribution to the programs of Québec Océan and the Canada Research Chair on the Response of Arctic Marine Ecosystems to Climate Change.

References

ACIA (2005) Scientific Report: Arctic Climate Impact Assessment. Cambridge University Press, Cambridge

Alexander V (1974) Primary productivity regimes of the nearshore Beaufort Sea, with reference to potential roles of ice biota. In: Reed JC, Sater JE (eds.) The Coast and Shelf of the Beaufort Sea, The Arctic Institute of North America, Calgary Andersen OGN (1977) Primary production, illumination and hydrography in Jorgen Bronlund

Fjord, North Greenland. Medd om Granland 205:1-27 Andersen OGN (1981) The annual cycle of phytoplankton primary production and hydrography in the Disko Bugt area, West Greenland. Medd om Granland 6:3-65 Apollonio S (1980) Primary production in Dumbell Bay in the Arctic Ocean. Mar Biol 61:41-51 Arrigo KR (2007) Physical control of primary production in Arctic and Antarctic polynyas. In: Barber DG, Smith WOJ (eds.) Polynyas: Windows into the World, Elsevier, Oceanography Series, Amsterdam

Arrigo KR, Robinson DH, Worthen DL, Dunbar RB, DiTullio GR, VanWoert M, Lizotte MP (1999) Phytoplankton community structure and the drawdown of nutrients and CO2 in the Southern Ocean. Science 283:365-367 Bélanger S, Xie HX, Krotkov N, Larouche P, Vincent WF, Babin M (2006) Photomineralization of terrigenous dissolved organic matter in arctic coastal waters from 1979 to 2003: Interannual variability and implications of climate change. Global Biogeochem Cycles. doi: 10.1029/2006GB002708 Carey AG, Jr. (1978). Marine biota. In: Environmental Assessment Alaskan Continental Shelf Interim Synthesis, Beaufort/Chuckchi, NOAA, Boulder, CT

Carmack E, Chapman DC (2003) Wind-driven shelf/basin exchange on an arctic shelf: The joint roles of ice cover extent and shelf-break bathymetry. Geophys Res Lett. doi: 10.1029/ 2003GL017526

Carmack E, Wassmann P (2006) Food webs and physical-biological coupling on pan-Arctic shelves: Unifying concepts and comprehensice perspectives. Prog Oceanogr 71:446-477 Duce RA, LaRoche J, Altieri K, Arrigo KR, Baker AR, Capone DG, Cornell S, Dentener F, Galloway J, Ganeshram RS, Geider RJ, Jickells T, Kuypers MM, Langlois R, Liss PS, Liu SM, Middelburg JJ, Moore CM, Nickovic S, Oschlies A, Pedersen T, Prospero J, Schlitzer R, Seitzinger S, Sorensen LL, Uematsu M, Ulloa O, Voss M, Ward B, Zamora L (2008) Impacts of atmospheric anthropogenic nitrogen on the open ocean. Science 320:893-897 Dugdale RC, Goering JJ (1967) Uptake of new and regenerated forms of nitrogen in primary productivity. Limnol Oceanogr 12:196-206 Emmerton CA, Lesack LFW, Vincent WF (2008) Mackenzie River nutrient delivery to the Arctic Ocean and effects of the Mackenzie Delta during open water conditions. Global Biogeochem Cycles doi: 10.1029/2006GB002856 Eppley RW, Peterson BJ (1979) Particulate organic matter flux and planktonic new production in the deep ocean. Nature 282:677-680 Forest A, Sampei M, Makabe R, Sasaki H, Barber D, Gratton Y, Wassmann P, Fortier L (2008) The annual cycle of particulate organic carbon export in Franklin Bay (Canadian Arctic): Environmental control and food web implications. J Geophys Res. doi: 10.1029/2007 JC004262

Frey KE, McClelland JW, Holmes RM, Smith LC (2007) Impacts of climate warming and permafrost thaw on the riverine transport of nitrogen and phosphorus to the Kara Sea. J Geophys Res. doi: 10.1029/2006JG000369 Garneau ME, Gosselin M, Klein B, Tremblay JE, Fouilland E (2007) New and regenerated production during a late summer bloom in an arctic polynya. Mar Ecol-Prog Ser 345:13-26 Grainger EH (1980) Primary production in Frobisher Bay, Arctic Canada. In: Dunbar MJ (ed.)

Marine Production Mechanisms, Cambridge University Press, Cambridge Holmes RM, McClelland JW, Raymond PA, Frazer BB, Peterson BJ, Stieglitz M (2008) Lability of DOC transported by Alaskan rivers to the Arctic Ocean. Geophys Res Lett. doi: 10.1029/2007GL032837 Horner R, Schrader GC (1982) Relative Contributions of Ice Algae, Phytoplankton, and benthic microalgae to primary production in nearshore regions of the Beaufort Sea. Arctic 35:485-503 Kangas P, Alasaarela E, Lax HG, Jokela S, Storgaard-Envall C (1993) Seasonal variation of primary production and nutrient concentrations in the coastal waters of the Bothnian Bay and the Quark. Aqua Fenn 23:165-176 Kattner G, Budeus G (1997) Nutrient status of the Northeast Water Polynya. J Mar Syst 10:185197

Klein B, LeBlanc B, Mei ZP, Béret R, Michaud J, Mundy CJ, von Quillfeldt CH, Garneau ME, Roy S, Gratton Y, Cochran JK, Belanger S, Larouche P, Pakulski JD, et al. (2002) Phytoplankton biomass, production and potential export in the North Water. Deep-Sea Res II 49:4983-5002

Kwok R, Cunningham GF, Zwally HJ, Yi D (2007) Ice, Cloud, and land Elevation Satellite (ICESat) over Arctic Sea ice: Retrieval of freeboard. J Geophys Res 112:doi: 10.1029/ 2006JC003978

Legendre L, Ackley SF, Dieckmann GS, Gulliksen B, Horner R, Hoshiai T, Melnikov IA, Reeburgh WS, Spindler M, Sullivan CW (1992) Ecology of sea ice biota. 2. Global significance. Polar Biol 12:429-444 Lemke P, Ren J, Alley RB, Allison I, Carrasco J, Flato G, Fujii Y, Kaser G, Mote P, Thomas RH, Zhang T (2007). Observations: Changes in snow, ice and frozen ground. In: Solomon S et al. (eds.) Climate Change 2007: The Physical Science Basis Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK/New York

Maslanik J, Fowler C, Stroeve J, Drobot S, Zwally HJ, Yi D, Emery WJ (2007) A younger, thinner arctic ice cover: Increased potential for rapid extensive sea ice loss. Geophys Res Lett. doi:10.1029/2007/GL032043 Mathis JT, Pickart RS, Hansell DA, Kadko D, Bates NR (2007) Eddy transport of organic carbon and nutrients from the Chukchi Shelf: Impact on the upper halocline of the western Arctic Ocean. J Geophys Res. doi: 10.1029/2006JC003899 McClelland JW, Stieglitz M, Pan F, Holmes RM, Peterson BJ (2007) Recent changes in nitrate and dissolved organic carbon export from the upper Kuparuk River, North Slope, Alaska. J Geophys Res-Biogeosci. doi: 10.1029/2006JG000371 McRoy CP, Goering JJ (1976) Annual budget of primary production in the Bering Sea. Mar Sci Commun 2:255-267

Meskus E (1976) The primary production in the northeastern Bothnian Bay. Acta Univ Oulu A 42:55-62

Michel C, Legendre L, Therriault JC, Demers S, Vandevelde T (1993) Springtime coupling between ice algal and phytoplankton assemblages in southeastern Hudson Bay, Canadian Arctic. Polar Biol 13:441-449 Michel C, Legendre L, Ingram RG, Gosselin M, Levasseur M (1996) Carbon budget of sea-ice algae in spring: Evidence of a significant transfer to zooplankton grazers. J Geophys Res 101:18345-18360

Nelson DM, Smith WO (1991) Sverdrup revisited - Critical depths, maximum chlorophyll levels, and the control of Southern-Ocean productivity by the irradiance-mixing regime. Limnol Oceanogr 36:1650-1661 Nielsen TG, Hansen BW (1999) Plankton community structure and carbon cycling on the western coast of Greenland during the stratified summer situation. 1. Hydrography, phytoplankton and bacterioplankton. Aquat Microb Ecol 16:205-221 Petersen GH (1964) The hydrography, primary production, bathymetry, and "Tagsaq" of Disko

Bugt, West Greenland. Medd om Granland 159:1-45 Peterson BJ, McClelland J, Curry R, Holmes RM, Walsh JE, Aagaard K (2006) Trajectory shifts in the Arctic and Subarctic freshwater cycle. Science 313:1061-1066 Polyakov IV, Alexeev VA, Belchansky GI, Dmitrenko IA, Ivanov VV, Kirillov SA, Korablev AA, Steele M, Timokhow LA, Yashayaev I (2007) Arctic Ocean freshwater changes over the past 100 years and their causes. J Climate 21:364-384 Reigstad M, Wassmann P, Riser CW, Oeygarden S, Rey F (2002) Variations in hydrography, nutrients and chlorophyll a in the marginal ice-zone and the Central Barents Sea. J Mar Syst 38:1-2

Rey F, Skjoldal HR, Slagstad D (1987) Primary production in relation to climatic changes in the Barents Sea. In: Loeng H (ed.) The Effect of Oceanographic Conditions on the Distribution and Population Dynamics of Commercial Fish Stocks in the Barents Sea, Proceedings of the 2nd Soviet-Norwegian Symposium, Institute of Marine Research, Bergen Rysgaard S, Nielsen TG, Hansen BW (1999) Seasonal variation in nutrients, pelagic primary production and grazing in a high-arctic coastal marine ecosystem, Young Sound, northeast Greenland. Mar Ecol Prog Ser 179:13-25 Rysgaard S, Glud RN, Risgaard-Petersen N, Dalsgaard T (2004) Denitrification and anammox activity in arctic marine sediments. Limnol Oceanogr 49:1493-1502 Sakshaug E (2004) Primary and secondary production in the Arctic Seas. In: Stein R, Macdonald

RW (eds.) The Organic Carbon Cycle in the Arctic Ocean, Springer, Berlin Sakshaug E, Slagstad D (1991) Light and productivity of phytoplankton in polar marine ecosystems - A physiological view. Polar Res 10:69-85 Sakshaug E, Slagstad D (1992) Sea ice and wind: Effects on primary productivity in the Barents

Sea. Atmos Ocean 30:579-591 Sambrotto RN, Goering JJ, McRoy CP (1984) Large yearly production of phytoplankton in the Western Bering Strait. Science 225:1147-1150

Simpson K, Tremblay J-E, Gratton Y, Price NM (2008) An annual study of nutrient distribution in the southeastern Beaufort Sea, Mackenzie Shelf and Amundsen Gulf. J Geophys Res. doi: 10.1029/2007 JC004462 Slagstad D, St0le-Hansen K (1991) Dynamics of plankton growth in the Barents Sea: Model studies. Polar Res 10:173-186. Smetacek V, Passow U (1990) Spring bloom initiation and Sverdrup's critical-depth model.

Limnol Oceanogr 35:228-234 Smith W, Jr., Gosselin M, Legendre L, Wallace D, Daly K, Kattner G (1997) New production in the Northeast Water Polynya: 1993. J Mar Syst 10:199-209 Springer AM, McRoy CP, Flint MV (1996) The Bering Sea Green Belt: Shelf-edge processes and ecosystem production. Fish Oceanogr 5:205-223 Stroeve J, Sereeze M, Drobot S, Gearheard S, Holland MM, Maslanik J, Meier T, Scambos TA (2008) Arctic Sea ice extent plummets in 2007. EOS Transactions. Am Geophys Union 89:13-14

Taguchi S (1972) Mathematical analysis of primary production in the Bering Sea in summer. In: Takenouchi A (ed.) Biological Oceanography of the Northern North Pacific Ocean, Idemitsu Shoten, Tokyo

Tozzi S, Schofield O, Falkowski PG (2004) Historical climate change and ocean turbulence as selective agents for two key phytoplankton functional groups. Mar Ecol Prog Ser 274:123132

Tremblay JE, Smith WOJ (2007). Primary production and nutrient dynamics in polynyas. In: Barber DG, Smith WOJ (eds.) Polynyas: Windows to the World, Elsevier, Oceanography Series, Amsterdam

Tremblay JE, Gratton Y, Fauchot J, Price NM (2002a) Climatic and oceanic forcing of new, net and diatom production in the North Water Polynya. Deep-Sea Res II 49:4927-4946 Tremblay JE, Gratton Y, Carmack EC, Payne CD, Price NM (2002b) Impact of the large-scale arctic circulation and the North Water Polynya on nutrient inventories in Baffin Bay. J Geophys Res. doi: 10.1029/2000JC000595 Tremblay JE, Hattori H, Michel C, Ringuette M, Mei Z-P, Lovejoy C, Fortier L, Hobson KA, Amiel D, Cochran JK (2006a) Trophic structure and pathways of biogenic carbon flow in the eastern North Water Polynya. Prog Oceanogr 71:402-425 Tremblay JE, Michel C, Hobson KA, Gosselin MG, Price NM (2006b) Bloom dynamics in early-opening waters of the Arctic Ocean. Limnol Oceanogr 51:900-912 Tremblay JE, Simpson K, Martin J, Miller L, Gratton Y, Barber D, Price NM (2008) Vertical stability and the annual dynamics of nutrients and chlorophyll fluorescence in the coastal, southeast Beaufort Sea. J Geophys Res. doi: 10.1029/2007JC004547 Vahatalo AV, Zepp R (2005) Photochemical mineralization of dissolved organic nitrogen to ammonium in the Baltic Sea. Environ Sci Technol 39:6985-6992 Vetrov AA, Romankevich EA (2004) Biological production of the Arctic Seas of Russia. In: Vetrov AA, Romankevich EA (eds.) Carbon Cycle in the Russian Arctic Seas, Springer, New York.

Walsh JJ, McRoy CP, Coachman LK, Goering JJ, Nihoul JJ, Whitledge TE, Blackburn TH, Parker PL, Wirick CD, Shuert PG, Grebmeier JM, Springer AM, Tripp RD, Hansell DA, Djenidi S, Deleersnijder E, Henriksen K, Lund BA, Andersen P, Mullerkarger FE, Dean K (1989) Carbon and nitrogen cycling within the Bering Chukchi Seas - source regions for organic-mjatter effecting AOU demands of the Arctic Ocean. Prog Oceanogr 22:277-359 Walsh JJ, Dieterle DA, Maslowski W, Whitledge TE (2004) Decadal shifts in biophysical forcing of arctic marine food webs: Numerical consequences. J Geophys Res. doi: 10.1029/ 2003JC001945

Wassman P, Slagstad D, Wexels Riser C, Reigstad M (2006) Modelling the ecosystem dynamics of the Barents Sea including the marginal ice zone II. Carbon flux and interannual variability. J Mar Syst 59:1-24

Welch HE, Kalff J (1975) Marine metabolism at Resolute Bay, Northwest Territories. Proceedings of the Circumpolar Conference on Northern Ecology. Part II, NRC, Ottawa Yamamoto-Kawai M, Carmack E, McLaughlin F (2006) Nitrogen balance and arctic throughflow. Nautre 443:43

Yang J, Comiso J, Walsh D, Krishfield R, Honjo S (2004) Storm-driven mixing and potential impact on the Arctic Ocean. J Geophys Res. doi: 10.1029/2001JC001248 Zhang X, Walsh JE, Zhang J, Bhatt US, Ikeda M (2004) Interannual variability of arctic cyclone activity, 1948-2002. J Climate 17:2300-2317

Bibliography

Simpson KG, Tremblay JE, Price NM (2009) Nutrient dynamics in the Amundsen Gulf and Cape Bathurst Polynya: 1. New production in spring inferred from nutrient draw-down. Mar Ecol Prog Ser (in press)

Smith WO, Jr. (1995) Primary productivity and new production in the Northeast Water (Greenland) Polynya during summer 1992. J Geophys Res 100:4357-4370

Was this article helpful?

0 0

Post a comment