Patterns and ecological implications of historical marine phytoplankton change

Abstract

There is growing evidence that average global phytoplankton concentrations have been changing over the past century, yet published trajectories of change are highly divergent. Here, we review and analyze 115 published phytoplankton trend estimates originating from a wide variety of sampling instruments to explore the underlying patterns and ecological implications of phytoplankton change over the period of oceanographic measurement (1889 to 2010). We found that published estimates of phytoplankton change were much less variable when estimated over longer time series and consistent spatial scales and from the same sampling instruments. Average phytoplankton concentrations tended to increase over time in near-shore waters and over more recent time periods and declined in the open oceans and over longer time periods. Most published evidence suggests changes in temperature and nutrient supply rates as leading causes of these phytoplankton trends. In near-shore waters, altered coastal runoff and increased nutrient flux from land may primarily explain widespread increases in phytoplankton there. Conversely, in the open oceans, increasing surface temperatures are strengthening water column stratification, reducing nutrient flux from deeper waters and negatively influencing phytoplankton. Phytoplankton change is further affected by biological processes, such as changes in grazing regimes and nutrient cycling, but these effects are less well studied at large scales. The possible ecosystem consequences of observed phytoplankton changes include altered species composition and abundance across multiple trophic levels, effects on fisheries yield, and changing patterns of export production. We conclude that there is evidence for substantial changes in phytoplankton concentration over the past century, but the magnitude of these changes remains uncertain at a global scale; standardized long-term measurements of phytoplankton abundance over time can substantially reduce this uncertainty.

Full text

MARINE ECOLOGY PROGRESS SERIES
Mar Ecol Prog Ser
Vol. 534: 251–272, 2015
doi: 10.3354/meps11411 Published August 27
INTRODUCTION
Marine phytoplankton are a diverse group of pe -
lagic photosynthetic microbes that provide over 90%
of marine primary production (Charpy-Roubaud &
Sournia 1990). Individual cells range over 4 orders of
magnitude in size (~0.2 to 1000 µm; Fig. 1A; Sheldon
et al. 1972, Margalef 1978, Falkowski et al. 2004) and
are globally distributed. Although marine phyto-
plankton account for only 0.2% of global photo -
synthetic carbon biomass, they generate 46.2% of
the primary production (Field et al. 1998). To achieve
this, the global standing stock of phytoplankton turns
over every 2 to 6 d on average (Behrenfeld & Fal -
© Inter-Research 2015 · www.int-res.com*Corresponding author: dboyce@dal.ca
REVIEW
Patterns and ecological implications of historical
marine phytoplankton change
Daniel G. Boyce1,2,*, Boris Worm3
1Department of Biology, Queen’s University, Kingston, ON K7L 3N6, Canada
2Ocean Sciences Division, Bedford Institute of Oceanography, PO Box 1006, Dartmouth, NS B2Y 4A2, Canada
3Biology Department, Dalhousie University, Halifax, NS B3H 4R2, Canada
ABSTRACT: There is growing evidence that average global phytoplankton concentrations have
been changing over the past century, yet published trajectories of change are highly divergent.
Here, we review and analyze 115 published phytoplankton trend estimates originating from a
wide variety of sampling instruments to explore the underlying patterns and ecological implica-
tions of phytoplankton change over the period of oceanographic measurement (1889 to 2010). We
found that published estimates of phytoplankton change were much less variable when estimated
over longer time series and consistent spatial scales and from the same sampling instruments.
Average phytoplankton concentrations tended to increase over time in near-shore waters and
over more recent time periods and declined in the open oceans and over longer time periods. Most
published evidence suggests changes in temperature and nutrient supply rates as leading causes
of these phytoplankton trends. In near-shore waters, altered coastal runoff and increased nutrient
flux from land may primarily explain widespread increases in phytoplankton there. Conversely, in
the open oceans, increasing surface temperatures are strengthening water column stratification,
reducing nutrient flux from deeper waters and negatively influencing phytoplankton. Phytoplank-
ton change is further affected by biological processes, such as changes in grazing regimes and
nutrient cycling, but these effects are less well studied at large scales. The possible ecosystem
consequences of observed phytoplankton changes include altered species composition and abun-
dance across multiple trophic levels, effects on fisheries yield, and changing patterns of export
production. We conclude that there is evidence for substantial changes in phytoplankton concen-
tration over the past century, but the magnitude of these changes remains uncertain at a global
scale; standardized long-term measurements of phytoplankton abundance over time can substan-
tially reduce this uncertainty.
KEY WORDS: Phytoplankton · Marine · Trend · Drivers · Consequences · Global · Change ·
Ecological
Resale or republication not permitted without written consent of the publisher

Mar Ecol Prog Ser 534: 251–272, 2015
kowski 1997). Due to this rapid turnover
(Fig. 1A), phytoplankton growth often
depletes available nutrient resources.
Over a century of scientific research
has shown that marine phytoplankton
play an important role in determining
the structure and functioning of marine
ecosystems (Chavez et al. 2003, Richard-
son & Schoeman 2004) and can have
large effects on fisheries yields (Ryther
1969, Chavez et al. 2003, Ware & Thom-
son 2005, Chassot et al. 2007, 2010), bio-
geochemical cycles (Redfield 1958, Falkowski et
al. 1998), climate regulation(Charlson et al. 1987,
Murtu gudde et al. 2002), and weather patterns
(Gnanadesikan et al. 2010). Reflecting this scien-
tific interest, the proportion of peer-reviewed sci-
entific studies of marine phytoplankton has
increased markedly over time (Fig. 1B).
Despite these increased research ef forts, one of
the most fundamental questions in phytoplankton
re search remains poorly resolved: How are aver-
age marine phytoplankton biomass concentra-
tions chan ging over the long term? Answering
this seemingly simple question is complicated by
the fact that phytoplankton concentrations are
highly variable in space and time and are difficult
to distinguish from other marine microbes and
particles, making it difficult to obtain direct
measurements of their carbon biomass. As a con-
sequence, the total concentration of the light-har-
vesting pigment chlorophyll, which is present in
all phytoplankton cells, has been used as a first-
order proxy of abundance and biomass. Despite
documented variability in the phytoplankton
chlorophyll:carbon ratio (Geider 1987), chloro-
phyll continues to be the most practical and
extensively used proxy of phytoplankton carbon bio-
mass over large spatial scales (Huot et al. 2007, Hen-
son et al. 2010). This review deals with changes in
phytoplankton concentrations as measured via ocean
color and chlorophyll assessed over the era of
oceanographic measurement, 1889 to 2010, and at
regional to global scales. We did not attempt to
include the literature on phytoplankton cell counts or
species composition and make only limited infer-
ences on changes in primary production. Following
this, we review the physical and biological drivers of
long-term marine phytoplankton change. We con-
clude by summarizing some potential ecosystem con-
sequences of phytoplankton change both across eco-
systems and globally.
MATERIALS AND METHODS
Phytoplankton trends
We systematically searched scientific databases to
identify peer-reviewed studies of temporal marine
phytoplankton change. Our literature search covered
a minimum of ~22 million articles from over 16 500
peer-reviewed journals. We limited our search to
publications estimating phytoplankton change from
chlorophyll concentrations or ocean color collected
from the upper ocean at multi-year scales (>5 yr).
Studies conducted in fresh or brackish waters were
not included. We extracted 115 phytoplankton time
series and estimates of temporal phytoplankton
change from 25 publications (Table 1).
252
0
0.05
0.1
0.15
0.2 Fish (0.006)
Phytoplankton (0.004)
Mammals (0.002)
Zooplankton (0.001)
Birds (0.001)
1985 1995 2005 201020001990
104
102
100
10–2
10–4
10–10 10–8 10–6 10–4 10–2 100102104
Doubling time (d)
Organism size (m)
Bacteria
Phytoplankton
Viruses
Protozoa
Zooplankton
Whales
Seals
Seabirds
Fish
Minutes
Hours
Days
Months
Decades
Weeks
Years A
B
Proportion
Year
Fig. 1. Phytoplankton in the scientific literature. (A) Dominant
space and time scales at which major groups of marine organ-
isms operate. The average size range (x-axis) is plotted as a
function of the average doubling time (y-axis) for various marine
groups. Phytoplankton are represented in green. Figure was
adapted after Murphy et al. (1988). (B) Time trends in the scaled
proportion of peer-reviewed studies reporting on major marine
species groups (1985−2010; see Supplement at www.int-
res.com/articles/suppl/m534p251_supp.pdf for details). Taxo-
nomic groups are represented as colors, with the linear rates of
change over time reported in brackets

Boyce & Worm: Marine phytoplankton change
To standardize measurements that were reported
in different units, we extracted the estimated total
percentage change in phytoplankton over the
available time span as reported by the authors. In
some cases, data extraction software was used to ex -
tract and calculate these rates (www.getdata-graph-
digitizer. com). Where the time series were extracted
from the publication, we fitted linear time series
models to the observations and calculated the total
percentage change as the difference between the
average concentration at the start and end of the
fitted time series referenced to the initial value. The
percent change was then divided by the length of the
time series to yield the standardized percent change
per year, relative to the initial phytoplankton concen-
tration. To spatially standardize the rates of change,
we binned all estimates into 5° × 5° cells. This resolu-
tion was selected because the majority of published
phytoplankton time series were estimated over spa-
tial domains equal to or greater than 5°. The extrac -
ted trends were also referenced according to the
sampling instrumentation used to ge n erate the un -
der lying time series: (1) in situ, (2) contemporary
remote sensing, (3) Secchi disk, (4) continuous plank-
ton recorder (CPR), (5) Forel- Ule, and (6) multi-
sensor. Multi-sensor trends are those which were
generated by combining measure-
ments from 2 or more of these sam-
pling instruments.
Variability of phytoplankton trends
The direction and magnitude of
phytoplankton time trends reported
in the literature have been widely
conflicting (Venrick et al. 1987, Fal -
kowski & Wilson 1992, Antoine et al.
2005, Gregg et al. 2005, Behrenfeld et
al. 2006, Boyce et al. 2010, 2014, Wer-
nand et al. 2013). To better under-
stand the factors that may ex plain this
variability, we estimated the standard
deviation (σ) of the standardized
phytoplankton trends that were avail-
able within each 5° × 5° cell. We then
used statistical models to estimate
what combination of predictors would
best explain the variability in the
phytoplankton trend estimates (σ). To
account for the spatial dependence
between trend variances (σ) within
each 5° × 5° cell, we estimated the
trend variability as a function of several covariates
within a generalized least-squares model as:
log10[σi] = β0+ β1Predictori+ εi(1)
where σiis the standard deviation of the trends in
celli, which was log transformed to ensure normality;
β0is the model intercept; β1is the rate of response
change as a function of the predictor in question; and
εiis the model error, specified as:
ε~ N(0,δ) (2)
where 0 is the mean, and δis the error covariance
matrix. To account for spatial autocorrelation, the
covariance parameters of δwere assumed to follow a
spatially dependent process, whereby the correlation
between them decreases exponentially with increas-
ing spatial separation (Cressie 1993). Using this ap -
proach, we quantitatively estimated the influence of
several predictors on the variability of phytoplankton
time trends. Predictor variables tested include the
number and type of sampling instrument, range and
variance of the spatial and temporal extent of the
trends, average time series length, average baseline
year of the trend, distance of the cell from the nearest
coast, measurement units of the trend, and ocean
basin where the trend was estimated.
253
Reference Start-end year Span Instrument Driver
(yr)
Aebischer et al. 1990 1955−1987 32 CPR BU
Aksnes & Ohman 2009 1949−2007 58 Secchi BU
Antoine et al. 2005 1979−2002 23 Satellite –
Behrenfeld et al. 2006 1997−2006 9 Satellite BU
Boyce et al. 2010 1899−2008 109 Blended BU
Boyce et al. 2014 1890−2010 120 Blended BU
Chavez et al. 2011 1989−2009 20 In situ –
Falkowski & Wilson 1992 1900−1981 81 Secchi –
Frank et al. 2005 1962−2002 40 CPR TD
Goes et al. 2005 1997−2004 7 Satellite BU
Gregg & Conkright 2002 1979−2000 21 Blended –
Gregg et al. 2005 1998−2003 5 Satellite –
Head & Pepin 2010 1998−2006 8 CPR –
Karl et al. 2001 1969−1998 29 In situ BU
Saba et al. 2010 1990−2007 17 In situ B
McQuatters-Gollop et al. 2007 1948−2003 55 CPR –
McQuatters-Gollop et al. 2011 1948−2008 60 CPR –
Montes-Hugo et al. 2009 1978−2006 28 Satellite BU
Motoda et al. 1987 1949−1969 20 In situ B
Raitsos et al. 2005 1948−2002 54 CPR –
Shiomoto et al. 1997 1985−1994 9 In situ TD
Sugimoto & Tadokoro 1997 1972–1973 21 In situ B
Suikkanen et al. 2007 1979−2003 24 In situ BU
Venrick et al. 1987 1968−1985 17 In situ BU
Wernand et al. 2013 1889−1999 110 Forel-Ule –
Table 1. Published phytoplankton time series reviewed here. CPR: continuous
plankton recorder; BU: bottom-up; TD: top-down; B: both; – : not investigated

Mar Ecol Prog Ser 534: 251–272, 2015
Average phytoplankton trends
Based on the results of the analysis of trend vari-
ance in the previous paragraph (Eqs. 1 & 2), we
calculated the mean rate of phytoplankton change
from the extracted trend estimates for each indi-
vidual 5° × 5° cell while minimizing the major fac-
tors influencing trend variation (see ‘Results’ for
details). As an additional sensitivity check, we cal-
culated mean rates of phytoplankton change
weighted by the length of the time series used to
generate the trend (years), but this did not influ-
ence the results.
Patterns of phytoplankton change
To explore patterns of similarity and dissimilarity
among the phytoplankton trends, we identified re -
ported trends which were coincident in space and
time. Published estimates where the proportion of
spatial and temporal overlap of the trends was
greater than 50% were identified as coincident. We
then looked for patterns within these coincident
trends concerning the degree of agreement in the
direction and magnitude of the time trends.
RESULTS AND DISCUSSION
Summary of phytoplankton trends
The majority of the extracted phytoplankton
trends spanned less than 23 yr, were initiated after
1978, and extended over areas less than 73 × 105km2
(approximately half the size of the Arctic Ocean;
Fig. 2A). The majority of the trends were estimated
from time series derived from in situ (36%), satellite
remote sensing (32%), or multiple (31%; Fig 2A,
inset) sampling instruments. The remaining trends
were inferred from time series of water column
transparency measurements using the standardized
Secchi disk (15 %; Secchi 1886), CPR (13%), or
semi-quantitative assessments of ocean color using
the Forel-Ule color scale (3%; Forel 1890). The
extracted trend estimates were globally distributed,
but their availability was greatest in the Northern
Hemisphere and closer to the coasts, and they were
sparsely distributed at high latitudes and in the
Southern Hemisphere (Fig. 2B). The estimates of
phytoplankton change were observed to be larger
and more variable over shorter time intervals
(Fig. 2C).
Variability of phytoplankton trends
The variability between estimated phytoplankton
time trends within each 5° × 5° cell was found to dif-
fer spatially (Fig. 3A) and was well predicted by both
intrinsic and extrinsic factors (Table 2). Phytoplank-
ton trend variability was best predicted by the ocean
basin where the trend was recorded (r2= 0.66; p <
0.0001), trend variability being highest in the North
Indian and North Atlantic Oceans and lowest in the
Arctic and Southern Oceans (Fig. 3B). Phytoplankton
trends also become progressively more variable
when more sampling instruments were used (r2= 0.6;
p < 0.0001), when estimated over shorter (i.e. less
than ~55 yr; r2= 0.51; p < 0.0001) and more recent
(i.e. after ~1975; r2= 0.48; p < 0.0001) time periods,
and when the trends were estimated over different
spatial extents (r2= 0.25; p < 0.0001; Fig. 3C–E). The
type of sampling instrument used was also a signifi-
cant predictor in some cases. Trend variability within
a cell (σ) was significantly increased by the addition
of trends derived from remote sensing (r2= 0.62; p <
0.0001) or Forel-Ule (r2= 0.24; p < 0.0001) observa-
tions. Since trends derived from multiple sampling
instruments were available in all 5° × 5° cells globally,
it was not possible to explore what effect including or
removing these trends would have on the trend vari-
ance in a given cell. However, separating trends into
those which were estimated from single instruments
and those which used multiple instruments sugges -
ted that trends estimated by combining measure-
ments from multiple instruments were typically less
variable than those estimated from single instru-
ments (Fig. S1 in the Supplement at www.int-res.
com/ articles/ suppl/ m534p251_supp.pdf). This pat-
tern may partly be driven by the generally longer
time series length of trends estimated from combined
data sources.
Average phytoplankton trends
We calculated average rates of phytoplankton
change within each 5° × 5° cell using extracted trend
estimates which were approximately coincident in
time and estimated with the same sampling in -
struments. Following the se guidelines, we calculated
average time trends in phytoplankton over 4 inter-
vals:
(1) Oceanographic era: 1890− 1920 to 1980−2010,
derived from direct measurements of ocean color,
Secchi depth, and in situ chlorophyll concentrations
(4 studies; 3 instruments).
254

Boyce & Worm: Marine phytoplankton change
(2) Early satellite era: 1975 to 2000−2010,
derived using remote sensing measure-
ments (3 studies; 2 instruments);
(3) Contemporary satellite era: 1995− 2005
to 2005−2010, derived using remote sensing
measurements (3 studies; 1 instrument);
(4) CPR era: 1945−1955 to 1990−2010,
derived using CPR measurements (4 stud-
ies; 1 instrument).
Although 36% of all extracted trends
were derived from in situ sampling instru-
ments, most of these could not be incorpo-
rated into our analysis, as they tended to be
available over time periods which did not
coincide with any other studies. Further, the
average rates of change over the oceano-
graphic era were derived from 4 studies
which were similar in the spatio-tempo-
ral extent but estimated time trends
using 3 different sampling instruments.
Long-term rates of change suggested
declining trends over much of the
ocean, except for the North Atlantic,
where large increases were driven by
possibly unrealistic estimates (6.7%
yr−1) derived from semi-quantitative
Forel-Ule ocean color measurements
(Wernand et al. 2013; Fig. 4A). Most
estimates over this period suggested
declining trends across the North and
equatorial Pacific oceans. This con-
trasts greatly with satellite-derived
255
Fig. 2. Phytoplankton time series data. (A)
Standardized number of phytoplankton
trend estimates as a function of trend length
and areal extent. Colours denote the num-
ber of phytoplankton trend estimates. Histo-
grams in the outer margins depict the fre-
quency distribution of the trends as a
function of trend length and areal extent.
Inset depicts the number of phytoplankton
time trends estimated using measurements
collected from different sampling instru-
ments. (B) Spatial distribution of all phyto-
plankton trend estimates. Colors depict the
number of trends per 5° × 5° cell. (C) Stan-
dardized rate of phytoplankton change over
time as a function of trend length. Long-
term trends that transcend scales of natural
variability (35 yr; Henson et al. 2010,
Beaulieu et al. 2013) are shown as squares;
all others are shown as triangles. Colors
identify the source publication. The horizon-
tal dashed line denotes no change. CPR:
continuous plankton recorder
A
B
Trend length (years)
Phytoplankton change (% yr-1)
C
−80˚
−40˚
0˚
40˚
80˚
12
0 20406080100120
3456789
−10
0
10
20
30
40
Aksnes & Ohman 2009
Boyce et al. 2010
Boyce et al. 2014
Falkowski & Wilson 1992
Frank et al. 2005
Mcquatters-Gollop et al. 2007
Mcquatters-Gollop et al. 2011
Raitsos et al. 2005
Wernand et al. 2013
Aebischer et al. 1990
Antoine et al. 2005
Chavez et al. 2011
Goes et al. 2005
Gregg & Conkright 2002
Gregg et al. 2005
Head et al. 2010
Karl et al. 2001
Lomas et al. 2010
Montez-Hugo et al. 2009
Motoda et al. 1987
Shiomoto et al. 1997
Sugimoto & Tadokoro 1998
Suikkanen et al. 2007
Venrick et al. 1987
Number of trends
Latitude
1
2
3
4
5
6
7
8
9
10
11
20 60 80 100 12040
0
500
1000
1500
2000
2500
3000
3500
Areal extent (kms x 100000)
Trend length (years)
Forel−Ule
CPR
Secchi
Remote sensing
In situ
Number of trend estimates
0 102030405060

Mar Ecol Prog Ser 534: 251–272, 2015
estimates since the late 1970s, suggestive of large-
scale phyto plank ton increases, except in the South-
ern Ocean (Fig. 4B). Again, these trends were largely
driven by one study, which reported coherent in -
creases in phyto plankton biomass since 1979
(Antoine et al. 2005). Satellite estimates since 1997
suggest spatially variable rates of change, with
declines in open ocean regions and increases in near-
shore areas (Fig. 4C). Estimates derived from CPR
measurements indicate large increases across the
temperate North Atlantic Ocean (35 to 65°N) since
~1955 (Fig. 4D).
256
BC
Latitude
Avera
g
e trend span (years)
Number of sampling instruments used
D
A
Many
Few
Points per pixel
Ran
g
e of spatial extent (number of cells)
−1.0
−0.5
0.0
0.5
1.0
Arctic
S. Indian
N. Indian
S. Atlantic
E. Atlantic
N. Atlantic
N. Pacific
E. Pacific
S. Pacific
Southern
r2 = 0.25
p < 0.0001
r2 = 0.51
p < 0.0001
r2 = 0.66
p < 0.0001
r2 = 0.6
p < 0.0001
−1.0
−0.5
0.0
0.5
1.0 E
−0.75 1.251.000.75
12
40 60 80 100 0 100 200 300 400 500 600
34 5
0.500.250.00−0.25−0.50
Standard Deviation of phytoplankton time trends (log10[σ])
−80˚
−60˚
−40˚
−20˚
0˚
20˚
40˚
60˚
80˚
Standard deviation (log10[σ])
Fig. 3. Phytoplankton trend variability. (A) Standard deviation between extracted trends within each 5° × 5° cell. Colors depict
the log10-transformed standard deviation calculated within each cell using all available phytoplankton trend estimates and
spatial interpolation. (B−E) Strong univariate predictors of phytoplankton trend variability across all 5° × 5° cells. (B) Ocean
basin where the trend was reported, (C) number of sampling instruments present in the cell, (D) average length of time
spanned by all trends in a cell, and (E) differences (range) between the spatial extent of trends in a cell. For (B−E), colours
depict the density of the points, where blue represents few and red represents many points per pixel. Yellow points (B,C) and
trend lines (D,E) are generalized least squares model-predicted phytoplankton trend variability values (standard deviation).
Vertical lines (B,C) or shaded regions (D,E) are the 95% confidence limits about the model estimates. Broken lines in (D) and
(E) are estimates from generalized additive models (GAMs). All relationships are statistically significant (p < 0.0001)

Boyce & Worm: Marine phytoplankton change 257
Predictor Effect r2AIC
Ocean basin Ocean basin where trend was estimated 0.66 −1231
Data: remote sensing Inclusion of remote sensing-derived trends 0.62 −1645
Sampling instrument Number of sampling instruments used to derive estimates in the cell 0.60 −1314
Time series span Average trend length within cell 0.51 −1431
Trend start year Minimum baseline year of all trends 0.48 −1389
Spatial range Difference between spatial coverage of trends in cell 0.25 −750
Data: Forel-Ule Inclusion of CPR-derived trends 0.24 −588
Spatial variability Variability between spatial coverage of trends in cell 0.22 −677
Data: CPR Inclusion of CPR-derived trends 0.08 −369
Units Number of trend response units in cell 0.05 −353
Data: in situ Inclusion of in situ-derived trends 0.02 −342
Data: Secchi Inclusion of Secchi-derived trends 0.02 −322
Distance Distance of cell from the coastline 0.01 −321
Table 2. Summary of univariate generalized least-squares model estimation of the factors influencing phytoplankton time
trend variability. Akaike’s information criterion (AIC) indicated the information-theoretic quality of the selected model; lower
values denote higher quality. CPR: continuous plankton recorder
C
AB
Latitude
Latitude
n = 4 studies
n = 4 studies
n = 3 studies
n = 2 studies
Phytoplankton change (% yr–1)
Mean start/end year of trends
Phytoplankton change (% yr–1)
Mean start/end year of trends
D
−80°
−60°
−40°
−20°
0°
20°
40°
60°
80°
−0.50 −0.25 0.00 0.25 0.50
−80°
−60°
−40°
−20°
0°
20°
40°
60°
80°
−1.5 −1.0
1900 1920 1940 1960 1970 1980
1996 1998 2000 2002 1940 1950 1960 1970 1980 1990 2000 2010
2004
1990 2000 2010
1980 2000
−0.5
–15 –10 –5 0 5 10 15 −7.5 −5.0 −2.5 0.0 2.5 5.0 7.5
0.0 0.5 1.0 1.5
−80°
−60°
−40°
−20°
0°
20°
40°
60°
80°
20°
40°
60°
80°
Fig. 4. Average phytoplankton change over different time scales and sampling instruments. Average rate of phytoplankton
change from (A) direct oceanographic measurements of chlorophyll, ocean color, and transparency since 1890; (B) satellite ob -
servations since 1975; (C) contemporary satellite observations since 1995; and (D) continuous plankton recorder measurements
since 1945. Colors within the maps depict the average rate of phytoplankton change within each 5° × 5° cell and are spatial ly in-
terpolated; white depicts no data. The plots below each map are the distributions of the start (blue) and end (red) years for all
trends. The long vertical lines represent the averages, and the vertical ticks are the actual start and end values for each trend

Mar Ecol Prog Ser 534: 251–272, 2015
Patterns of phytoplankton change
Although the average phytoplankton trends were
generally variable (Fig. 4), examining only those re -
ported trends which were coincident in space and
time enabled us to identify instances where the
direction and magnitude of change were in agree-
ment. Long-term phytoplankton trends in the North
and equatorial Pacific Oceans all indicated declining
trends (Fig. 5A). Phytoplankton trends in the North
and equatorial Pacific Oceans were estimated from
measurements of Secchi depth, water colour, or in
situ chlorophyll between ~1911 and ~2003 and sug-
gested that phytoplankton had declined at rates of
between −0.48 (Wernand et al. 2013) and −0.05%
yr−1 (Boyce et al. 2014). Three phytoplankton trends
in the Northeast Atlantic Ocean also showed good
agreement and suggested an increase between
~1918 and ~2009 at a rate of change between 0.5
(Raitsos et al. 2005) and 2.4% yr−1 (McQuatters-
Gollop et al. 2011). Long-term trends in the Arctic
and Southern Oceans also agreed, but this is perhaps
unsurprising, since they were estimated using similar
data sources and methods (Boyce et al. 2010, 2014).
Phytoplankton trends in the North Atlantic Ocean
appeared to be particularly variable, and both long-
and short-term phytoplankton trends there dis-
agreed widely in terms of sign and magnitude of
change (Fig. 5). While limited data availability may
contribute variability and disagreement among
trends in the Indian, South Atlantic, and South
Pacific Oceans, this is not the case in the North Atlan -
tic (Boyce et al. 2012). To some extent, the high trend
variability there is driven by unrealistic rates of
258
North Atlantic (1901−2002)
Eq. Atlantic (1904−2001)
South Atlantic (1911−2003)
North Indian (1918−1997)
Indian (1911−2002)
North Pacific (1912−1999)
Eq. Pacific (1916−2005)
South Pacific (1956−2006)
Southern (1906−2006)
Arctic (1899−2004)
NE Atlantic (1918−2009)
NE Atlantic (1948−2005)
(3)
(3)
(2)
(2)
(3)
(4)
(3)
(2)
(2)
(2)
(2)
(2)
−6 −4 −2 0 2468
Indian (1979−2001)
Central Indian (1979−2001)
North Pacific (1979−2001)
South Pacific (1979−2001)
Eq. Pacific (1979−2001)
North Atlantic (1979−2001)
South Atlantic (1979−2001)
Eq. Atlantic (1979−2001)
North Sea (1952−1995)
NE Pacific (1956−2002)
(2)
(2)
(3)
(2)
(2)
(3)
(2)
(2)
(2)
(2)
Similar
DissimilarSimilar
Dissimilar
Rate of phytoplankton chan
g
e (% yr–1)
A
B
Antoine et al. 2005
Gregg & Conkright 2002
Boyce et al. 2010
Boyce et al. 2014
Wernand et al. 2013
Falkowski & Wilson 1992
Mcquatters et al. 2011
Aebischer et al. 1990
Mcquatters-Gollop et al. 2007
Raitsos et al. 2005
Aksnes & Ohman 2009
Fig. 5. Patterns in the di -
rec tion of space- and time-
coincident phytoplankton
trends. Coincident rates
of phytoplankton change
over (A) long (>40 yr time
series span) and (B) short
(<40 yr time series span)
periods. Points depict the
study from which each
trend was obtained. Geo-
graphic location and aver-
age time series start and
end points are given on
the y-axis. Numbers in
the right margin are the
number of coincident esti-
mates. The dashed vertical
line denotes 0 change; the
horizontal line delineates
trends which agree or dis-
agree in their reported
direction of change. Eq.:
Equatorial

Boyce & Worm: Marine phytoplankton change
change (6.7% yr−1) estimated from semi-quantitative
ocean color measurements (Wernand et al. 2013).
However, even after removing this outlier, variability
in the North Atlantic Ocean remained high. Interest-
ingly, similar vari abil ity has been predicted for future
estimates of phytoplankton change derived from
ocean circulation models, which are also highly di -
vergent in the North Atlantic Ocean (Henson et al.
2010). As the North Atlantic is subject to strong vari-
ability on seasonal, decadal, and multi-decadal time
scales (Martinez et al. 2009, Boyce et al. 2010), it is
likely that high natural variability masks smaller
inter-annual changes that are occurring (Henson et
al. 2010).
Again, we observed that coincident trends avail-
able over shorter time periods tended to be more
variable and less similar in the direction of change
(Fig. 5B). This likely reflects quasi-periodic climate
variability, which may strongly influence shorter-
term (less than ~27 to ~40 yr) trends (Behrenfeld et al.
2006, Martinez et al. 2009, Boyce et al. 2010, Henson
et al. 2010, Chavez et al. 2011, Beaulieu et al. 2013).
As such, some of the trends reported here, particu-
larly those estimated from contemporary remote
sensing estimates of ocean color (Fig. 4B,C), may re -
flect climate-driven variability rather than sustained
long-term changes. Phytoplankton trend estimates
were observed to switch from negative to positive
through time and with proximity to the nearest coast-
line (Fig. S2 in the Supplement), similar to the find-
ings of other long-term studies (Boyce et al. 2010,
2014). Phytoplankton declines in the open oceans
have also been observed previously (Gregg &
Conkright 2002, McClain & Signorini 2004, Polovina
et al. 2008) and are predicted to continue into the
future (Polovina et al. 2011). Increases in nearshore
waters are well documented in many regions and are
likely related to increasing coastal eutrophication
there (see ‘Environmental conditioning’ for further
details).
Phytoplankton trends in the 21st century
Similar to our results from observational measure-
ments, predicted patterns of future phytoplankton
change from process-based ocean models are vari-
able (Table 3). Despite this variability, 15 of 18 stud-
ies (83%) predict a global phytoplankton decline
over the next century. Most predictions suggested
phytoplankton increases at high latitudes and de -
clines at low and middle latitudes (Schmittner et al.
2008, Henson et al. 2010, Steinacher et al. 2010, Hof-
mann et al. 2011, Mora et al. 2013). Some of the
largest and most variable declines are predicted to
occur in the North Atlantic Ocean (Henson et al.
2010, Steinacher et al. 2010, Mora et al. 2013), where
published empirical estimates are also highly vari-
able. This suggests that temporal phytoplankton dy -
na mics in the North Atlantic are particularly difficult
to constrain from both empirical estimates (Boyce et
al. 2014) and process-based models (Table 3).
259
Reference Simulation Span (yr) Response Change Unit Forcing
range (yr AD)
Hofmann et al. 2011 2000−2200 200 Chl −50 % CO2
Schmittner et al. 2008 2000−4000 2000 Chl + 5 % CO2
Henson et al. 2010 2001−2100 99 Chl −0.0002 mg m−3 yr−1 Temperature
Boyd & Doney 2002 2000−2080 80 Chl −8.5 % CO2
Beaulieu et al. 2013 2001−2100 99 Chl −1.53 ×10–4 mg m−3 yr−1 CO2
Olonscheck et al. 2013 2000−2100 100 Chl −50 % CO2
Mora et al. 2013 2014−2100 86 C −4 % CO2
Mora et al. 2013 2014−2100 86 C −10 % CO2
Henson et al. 2010 2001−2100 99 PP −0.15 mg C m−2 d–1/yr Temperature
Sarmiento et al. 2004 2040−2060 20 PP +4.4 % Temperature
Taucher & Oschlies 2011 2000−2100 100 PP −5.3 % Temperature
Bopp et al. 2001 2000−2080 80 PP −8.9 % CO2
Bopp et al. 2001 2000−2080 80 PP −8.5 % CO2
Boyd & Doney 2002 2060−2070 10 PP −5.5 % CO2
Steinacher et al. 2010 1860−2099 239 PP −11 % CO2
Schmittner et al. 2008 2000−4000 2000 PP +100 % CO2
Cermeño et al. 2008 2000−2100 100 PP −14 % CO2
Cox et al. 2000 2000−2100 100 PP −5 % CO2
Table 3. Published projections of future changes in phytoplankton chlorophyll (Chl), carbon biomass (C), or primary produc-
tion (PP). CO2: carbon dioxide

Mar Ecol Prog Ser 534: 251–272, 2015
Drivers of phytoplankton change
To the first order, phytoplankton cell growth is de -
termined by the availability of sunlight and macro -
nutrients (bottom-up processes) as well as grazing,
viral infection, auto-catalyzed programmed cell death
(PCD; Agusti et al. 1998, Bidle & Falkowski 2004),
pathogenic bacteria, and fungi (top-down processes).
Based on this, we discuss drivers of plankton change
in the context of changes in (1) physical forcing and
(2) biological forcing, which may alter the strength of
bottom-up and top-down processes on marine phyto-
plankton.
Physical forcing
Particularly in the open oceans, which account for
90% of the ocean surface, studies have observed
phytoplankton growth and productivity to be strong -
ly driven by physical processes, such as mixing and
upwelling, which control nutrient flux (Oschlies &
Garcon 1998, McGillicuddy et al. 2007). Passive dif-
fusion across the thermocline (Chavez & Toggweiler
1995), biological nitrogen fixation (Capone et al.
1997), and the atmospheric deposition of iron are also
of regional importance (Behrenfeld et al. 1996).
Hence, it is likely that the observed chlorophyll de -
clines in open ocean regions (Figs. 4C & 5) are driven
by factors affecting these processes. Primary among
these is increasing sea surface temperature, which
generally leads to reduced mixing depth, enhanced
stratification, and reduced nutrient flux from deeper
waters. Studies using observational measurements
have reported strong temperature-driven stratifica-
tion (TDS) effects on phytoplankton concentration at
seasonal (Lozier et al. 2011), inter-annual (Behren-
feld et al. 2006), multi-decadal (Martinez et al. 2009),
and geological time scales (Romero et al. 2011, Ver-
meij 2011). Analyses of satellite observations suggest
that TDS may also be leading to an expansion of the
low-chlorophyll gyres of the open oceans (McClain &
Signorini 2004, Polovina et al. 2008); bio-physical
models also predict this expansion to continue over
the coming century (Polovina et al. 2011). Studies
using empirical observations (Behrenfeld et al. 2006,
Boyce et al. 2010, Boyce 2013) and process-based
models (Henson et al. 2010) have provided strong
empirical evidence that TDS effects on phytoplank-
ton also vary by latitude, with strong negative effects
at low and middle latitudes but positive effects at
high latitudes. This pattern of change is partly at
odds with observations, which suggest declining
trends at high latitudes (>70°N or S; Fig. 4A,B). In
well-mixed high-latitude oceans, increasing TDS
may positively influence phytoplankton growth by
retaining phytoplankton cells above the critical
depth (Sverdrup 1953, Jacobs et al. 2002, Montes-
Hugo et al. 2009, Arrigo et al. 2012) or by modifying
phytoplankton−grazer interactions (Behrenfeld 2010).
Process-based models also predict that over the com-
ing century, rising temperatures may lead to reduced
ice cover and increased light availability; combined
with a longer growing season, this may lead to in -
creased phytoplankton biomass and productivity at
high latitudes (Schmittner et al. 2008, Henson et al.
2010, Steinacher et al. 2010, Mora et al. 2013). Large-
scale phytoplankton trend estimates are generally
less available at these high latitudes (Fig. 2B), likely
contributing to the variability of the empirical esti-
mates of change there.
Experimental, field, and modeling studies suggest
that TDS may also lead to declines in the concentra-
tion of larger phytoplankton species such as diatoms
and increases in smaller species such as small flagel-
lates and cyanobacteria (Li et al. 2009, Morán et al.
2010, Barnes et al. 2011, Boyce et al. 2015a). These
effects may be related to different nutrient uptake
strategies between large and small phytoplankton
species (Bopp et al. 2005, Cermeño et al. 2008, Li et
al. 2009), the temperature-size rule (Atkinson 1994,
Atkinson et al. 2003, Morán et al. 2010), or increased
sinking rates of larger phytoplankton species (Ro drí -
guez et al. 2001).
Changes in a range of additional physical variables
such as wind intensity or salinity may modify the in -
fluence of temperature on stratification and nutrient
flux in some locations. For instance, observations of
changing wind intensity over the past century will
have large effects on upwelling intensity, including
highly productive Eastern Boundary Current systems
(Bakun 1990, Vecchi et al. 2006). In the Indian
Ocean, warming of the Eurasian land mass has been
linked to intensifying monsoon winds and upwelling,
leading to reported phytoplankton increases of 300 to
350% (Goes et al. 2005; Fig. 4B,C; Table 1). Wind-
driven atmospheric deposition of iron is of regional
importance to phytoplankton growth (Behrenfeld et
al. 1996). In polar oceans, melting sea ice has been
linked to increased upper-ocean irradiance and re -
duced surface salinity, which may have stronger
effects on phytoplankton than TDS (Lee et al. 2012,
Post et al. 2013). Increasing ocean acidification may
also alter phytoplankton community structure, bene-
fiting smaller species and possibly hindering calci -
fying ones (Orr et al. 2005, Iglesias-Rodríguez et al.
260

Boyce & Worm: Marine phytoplankton change
2008, Beaufort et al. 2011, Riebesell et al. 2013).
Acidification-driven reductions in the bioavailability
of iron could also lead to phytoplankton declines in
expansive high-nutrient, low-chlorophyll regions of
the ocean (Shi et al. 2010).
Biological forcing
Trophic control. Consumers may drive changes in
phytoplankton biomass and species composition
through their trophic (feeding) behaviour. These ef -
fects may be caused by modified grazing pressure
(direct) or by changes to other consumers which
may propagate across multiple trophic links, ulti-
mately modifying grazing pressure (indirect). For
instance, the removal of a top predator from the
Northwest Atlantic ecosystem led to cascading
trophic effects which may have contributed to a
long-term (~40 yr) increase in phytoplankton there
(Frank et al. 2005, 2011). Such trophic cascades
have been observed across diverse ecosystems but
often weaken at the plankton level (McQueen et al.
1986, Micheli 1999, Shurin et al. 2002, Baum &
Worm 2009, Boyce et al. 2015b). While it is unclear
what factors determine the occurrence and strength
of such cascades, the intensity of fisheries exploita-
tion is likely a contributing factor (Frank et al. 2005,
2006, Myers et al. 2007, Baum & Worm 2009). It is
also possible that short food chains with fewer
trophic transfers between predators and producers
may be more susceptible to cascading effects, with
reduced diversity and lower functional redundancy
rendering the systems generally less stable (Frank
et al. 2006, Worm et al. 2006, Casini et al. 2008).
Grazing pressure (Loeb et al. 1997, Sommer et al.
2007), heterotrophic bacterial activity (Llewellyn et
al. 2008), and viral infection (Suttle 1994) can all be
influential in controlling phytoplankton concentra-
tions. Experimental and modeling studies also show
that ocean warming in duces a more rapid metabolic
response in heterotrophs as compared to autotrophs,
which leads to increased grazer control and reduced
standing biomass of phytoplankton (O’Connor et al.
2009, Lewan dowska et al. 2014), although this effect
also appears to be context-specific (Lewandowska
et al. 2014).
Environmental conditioning. Marine organisms
mo di fy their environment through a range of non-
trophic activities, thereby promoting or inhibiting
phytoplankton growth in a process termed environ-
mental conditioning (Smetacek 2008). For instance,
whales and seals forage at depth and excrete fecal
plumes in surface waters. In this manner, essential
macronutrients such as nitrogen and iron are trans-
ported from deeper to surface waters, promoting
phyto plankton growth. Changes in this so-called
whale pump have been suggested as a possible dri -
ver of phytoplankton change in some regions (Sme -
ta cek 2008, Lavery et al. 2010, 2014, Roman &
McCarthy 2010). Particularly, long-term reductions
in whale biomass in the Northwest Atlantic (Roman &
Palumbi 2003) and Southern (Smetacek 2008)
Oceans may have led to reduced efficiency of the
whale pump and could contribute to observed long-
term phytoplankton declines there (Boyce et al. 2010,
2014).
The activities of biological organisms may also in -
fluence phytoplankton through their effects on ocean
mixing (Munk 1966). Kinetic energy generated by
swimming organisms could account for 33% of
global ocean mixing; this is comparable to wind- or
tidal-driven mixing (Dewar et al. 2006). Observa-
tional studies have also reported that the swimming
activities of krill may induce 4 orders of magnitude
increases in turbulence in nearshore waters (Kunze
et al. 2006). Given the global distribution and large
biomass of vertically migrating marine organisms
(Gjosaeter & Kawaguchi 1980, Irigoien et al. 2014),
biologically generated turbulence may have larger
impacts on the global flux of nutrients to phytoplank-
ton in surface waters than previously recognized.
The harvesting of large-bodied consumers (Estes et
al. 2011) may have disproportionately reduced nutri-
ent cycling and physical mixing, with possible effects
on phytoplankton (Behrenfeld et al. 2006, Polovina et
al. 2008, Boyce et al. 2010). This mechanism could
have contributed to part of the observed phytoplank-
ton declines in the open oceans (Fig. S2 in the Sup-
plement), where vertical mixing is a particularly
strong driver of phytoplankton change.
By accounting for 20% of all marine microorganism
mortality, viruses may have large effects on nutrient
fluxes in the oceans, with consequences for phyto-
plankton (Suttle 2007). Viruses negatively influence
phytoplankton directly via cell lysis, or their presence
may trigger phytoplankton PCD, likely as an anti-
viral defence mechanism (Bidle & Falkowski 2004).
Viruses may also infect consumers ranging from bac-
teria to whales, thereby increasing the amount of dis-
solved and particulate organic matter available for
phytoplankton growth (Suttle 2007).
The activities of humans provide perhaps the clear-
est examples of environmental conditioning. Some
examples concern strong effects on coastal nutrient
inputs stemming from soil erosion, agricultural prac-
261

Mar Ecol Prog Ser 534: 251–272, 2015
tises, and industrial activities. For instance, anthro-
pogenic activity has led to global increases in the
river-borne deposition of nitrate and phosphate to
nearshore waters by up to 300% (Duce et al. 1991) or
more in some regions (Howarth et al. 1996), while
atmospheric deposition of nitrate has increased by up
to 50% in some regions (Brimblecombe & Pitman
1980). Such large-scale environmental conditioning
by humans in nearshore oceans is almost certainly
contributing to the large phytoplankton increases ob -
served there (Figs. 4 & 5).
Synergistic and context-dependent forcing
Individual physical and biological drivers of phyto-
plankton change might reinforce or counteract each
other. For instance, ocean warming generally in -
creases phytoplankton growth rates (Sarmiento et al.
2004) and microbial metabolism (Taucher & Oschlies
2011), which could counteract negative TDS effects
on phytoplankton. However, the metabolic theory of
ecology (MTE; Brown et al. 2004) and experimental
results (Sommer & Lengfellner 2008, O’Connor et al.
2009) suggest that rising temperature increases
grazer metabolic rates faster than phytoplankton
metabolic rates, leading to reduced phytoplankton
via increased grazing pressure. Hence, it is important
to distinguish clearly between the physically me -
diated temperature effects on phytoplankton via
changes in stratification and nutrient delivery and
the biologically mediated temperature effects on
phytoplankton via altered phytoplankton and con-
sumer metabolism. One experimental study com-
pared the relative importance of these processes and
found that this varied depending on average nutrient
availability in the ecosystem (Lewandowska et al.
2014). In nutrient-limited systems, the effect of rising
temperature on nutrient delivery was dominant,
while in nutrient-replete systems, the effect of rising
temperature on grazing pressure was stronger. How-
ever, under both nutrient regimes, the net effect of
increasing temperature on phytoplankton was nega-
tive. Such context-dependent forcing has also been
revealed in a recent synthesis of published studies
which found that trophic control in marine eco -
systems scaled unimodally with temperature: strong
resource control occurred between 5 and 15°C, and
consumer control occurred at the cold and warm
extremes of this range (Boyce et al. 2015b). Such con-
text-dependent physical−biological effects on phyto-
plankton are an important frontier for further
research.
Case study 1: Global patterns of phytoplankton,
nutrients, and grazers
To quantitatively explore primary controls on
phyto plankton biomass across the seascape, we ex -
am ined spatial gradients in chlorophyll (mg m−3) in
conjunction with spatial data for nitrate concentra-
tion (µmol l–1) and total zooplankton carbon biomass
(mg m−3) at global scales (Fig. 6A–C). This approach
has been used to show the strong positive relation-
ship be tween phytoplankton and zooplankton con-
centration across the Atlantic Ocean (Irigoien et al.
2004) but to our knowledge had not yet been applied
globally. All data were extracted from publicly avail-
able sources (see Supplement at www.int-res.com/
articles/ suppl/ m534p251_supp.pdf). Based on this
simple approach, global patterns in chlorophyll
appeared similar to those of nitrate and zooplankton
(Fig. 5A−C). Elevated levels in nearshore, high-
latitude, and up welling regions were well delin-
eated, as are the oligo trophic gyres of the major
ocean basins, where lower nitrate and phytoplankton
concentrations prevail. Ordinary least-squares (OLS)
regressions of log-transformed mean nitrate or zoo-
plankton on phytoplankton measurements for each
1° × 1° cell statistically confirmed this relationship, a
result suggestive of bottom-up control of both phyto-
plankton and zooplankton concentrations by nitrate
(Fig. 5D,F). The relationship between nitrate and
chlorophyll was strongly positive (r = 0.51, p <
0.0001) and was best approximated by a polynomial
regression (r2= 0.39; p < 0.0001; Fig. 6D). The non-
linearity of the relationship likely relates to the
phytoplankton re quirement for additional resources
such as phosphate, silicate, carbon, and iron but may
also be driven by a nutrient saturation. For example,
despite high available nitrate concentrations in some
re gions, phytoplankton biomass is limited by, and
responds strongly to, the addition of iron across 20 to
40% of ocean surface waters (Behrenfeld et al. 1996,
Boyd et al. 2000, Moore et al. 2009). It is therefore
possible that changes in physically or biologically
driven iron deposition may have influenced the
observed phytoplankton trends, particularly in the
Pacific, Atlantic, and Southern oceans (Fig. 4).
Ecological consequences of phytoplankton change
Consequences of phytoplankton change globally
Globally, spatial variation in phytoplankton con-
centration is strongly and positively related to varia-
262

Boyce & Worm: Marine phytoplankton change
tion in zooplankton (r = 0.63; p < 0.0001) (Fig. 6E),
suggesting that phytoplankton biomass strongly
influences zooplankton via resource control (Fig.
6B,C). The well-established positive relationship
between zoo plank ton and fish for both the larval and
adult stages (Lasker 1975, Cushing 1990, Beaugrand
et al. 2003) suggests that these observed relation-
ships (Fig. 6B,C) likely propagate to higher trophic
levels. Additionally, bottom-up linkages between
phytoplankton primary production, zooplankton,
mesopelagic fishes biomass, and total fisheries land-
ings have been observed at regional scales (Ware &
Thom son 2005, Chassot et al. 2007) and globally
(Chassot et al. 2010, Irigoien et al. 2014). These cor-
relations between spatial gradients of primary and
secondary productivity do not necessarily imply cau-
sation but support the hypothesis that phytoplankton
productivity sets the carrying capacity of marine
ecosystems through resource control (Fig. 6B,C). It
needs to be observed, however, that such variation
across ecosystems only captures order of magnitude
changes in the abundance of different trophic groups
on a log−log scale. Within individual ecosystems,
there is ample evidence for top-down effects of graz-
ers on phytoplankton that may interact with the
bottom-up forcing discussed in the previous subsec-
tion (Verity & Smeta cek 1996, Micheli 1999, Frank et
al. 2006, Baum & Worm 2009, Estes et al. 2011).
263
Fig. 6. Global climatology of phytoplankton, zooplankton, and nutrient concentrations. Averaged (A) nitrate, (B) chlorophyll,
and (C) zooplankton concentration per 1° × 1° cell depicted as colors. White represents areas with no data. (D) Chlorophyll as
a function of nitrate, and (E) zooplankton as a function of chlorophyll. All variables are log10-transformed average concentra-
tions per 1° × 1° cell. Colors depict the number of measurements per pixel. Relationship in (D) was best approximated by a poly-
nomial function, and relationship in (E) was best approximated by a quadratic function. Shading represents the 95 % confi-
dence limits around the fitted curve

Mar Ecol Prog Ser 534: 251–272, 2015
In addition to the spatial approaches discus sed
above, temporal approaches have also re vea led
resource regulation of the grazer food web by phyto-
plankton affecting such taxonomically distant organ-
isms as leatherback turtles (Saba et al. 2008), octo-
puses (Otero et al. 2008), seabirds (Frederiksen et al.
2006), and fishes (Richardson & Schoeman 2004).
Phyto plankton concentrations also influence higher
trophic levels via changes in the timing and magnitu -
de of phenological cycles (Hjort 1914, Cushing 1990).
Observational studies have demonstrated that the
amount, species composition, and timing of phyto-
plankton blooms can strongly influence the survival of
larvae and the subsequent population size of fish (Las -
ker 1975, Platt et al. 2003). Such pheno logical chan -
ges in the concentration and quality of phytoplankton
may be manifest as temporal changes in overall bio-
mass and can also affect ecosystem structure from the
bottom up (Edwards & Richardson 2004).
Apart from the effects operating within pelagic
waters, the observed changes in plankton abun-
dance may also affect deep-sea ecosystems, which
are almost entirely sustained by the rain of particu-
late organic matter (POM) from surface waters, the
majority of which is produced by phytoplankton
(Ruhl et al. 2008). The downward flux of particulate
organic carbon (POC) accounts for up to 67% of
deep-sea benthic biomass in some regions (Johnson
et al. 2007). Studies have also documented positive
relationships between spatial gradients of surface
chlorophyll, POC flux, and deep-sea macro-faunal
abundance (Johnson et al. 2007, Ruhl et al. 2008).
Phytoplankton-derived POC flux may also in fluence
inter-specific body size distributions of deep-sea
macrofauna (Ruhl et al. 2008) and diversity of deep-
sea ecosystems. There is broad consensus among
physically based models, which predict de clining
export production over the coming century to be
driven in part by rising temperature and changes in
phytoplankton biomass and community composition
(Steinacher et al. 2010). The strong dependence of
food-stressed deep-water ecosystems on export pro-
duction would likely render them particularly sensi-
tive to changes in phytoplankton concentration and
community composition.
Consequences of phytoplankton change
across ecosystems
The evidence reviewed thus far suggests that
phytoplankton biomass and productivity place first-
order constraints on the carrying capacity of pelagic
and deep-sea ecosystems. Additional factors such as
the structure of the ecosystem, the degree to which
productivity is affected, altered phenology, and
changes in species composition and size structure
will likely further modify the ecological response to
phytoplankton biomass changes across ecosystems.
In the open ocean oligotrophic gyres, phytoplank-
ton biomass is low and comprised mainly of pico- and
nanophytoplankton (<0.2 to 20 µm diameter). Due to
the small primary producer cell size and the con-
straints of size-based predation (Barnes et al. 2010,
Wirtz 2012, 2013, Boyce et al. 2015a), primary pro-
duction in the open ocean is inefficiently channelled
to higher trophic levels through a microbial food
chain, or microbial loop, consisting of picophyto-
plankton, viruses, bacteria, and small heterotrophic
protists (Ryther 1969, Azam et al. 1983, Azam &
Worden 2004, Azam & Malfatti 2007). The preva-
lence of the microbial loop in open ocean ecosystems
(Pomeroy et al. 2007) results in long, complex flows of
primary production from producers to grazers and
highly efficient recycling of organic matter (Ryther
1969). Ultimately, the microbial loop increases the
recycling efficiency of phytoplankton and other dis-
solved organic matter but reduces the amount of pri-
mary production available to both the grazer and
deep-sea ecosystems (Iverson 1990). This, in combi-
nation with the low phytoplankton biomass, con-
tributes to the low fishery landings per unit area and
export production of open ocean ecosystems (Ryther
1969); hence, they are sometimes referred to as bio-
logical deserts (Polovina et al. 2008).
Since the open oceans are already food stressed,
ecosystems there may be particularly sensitive to any
reductions in phytoplankton biomass. Process-based
models and field and experimental studies suggest
that continued warming will lead to increases in the
abundance of picophytoplankton (Cermeño et al.
2008, Polovina & Woodworth 2012), expansions of the
oligotrophic oceans (Polovina et al. 2008, 2011), and
increased microbial metabolism (Taucher & Oschlies
2011). Such changes may increase the relative impor-
tance and turnover rate of the microbial loop both in
the oligotrophic gyres and elsewhere, thereby in-
creasing primary production, but may re duce the
channeling of primary production to the grazer and
deep-sea food chains. These changes in the oligotro-
phic open oceans may be exacerbated by predicted
temperature-driven reductions in phytoplankton di-
versity over the coming century (Thomas et al. 2012).
Such diversity losses may alter the structure (Hooper
et al. 2012) and stability (Worm et al. 2006) of open
ocean ecosystems and may further reduce primary
264

Boyce & Worm: Marine phytoplankton change
productivity in these ecosystems through the loss of
productive species (Tilman et al. 1996), reduced com-
plementarity (Reich et al. 2012), or increased grazer
pressure (Hillebrand & Cardinale 2004).
In contrast to the open oceans, nearshore eco -
systems are supported by an abundance of large
microphytoplankton species (~20 to 1000 µm in dia -
meter; Cermeño et al. 2008). These ecosystems often
have shorter food chains and are thought to be more
efficient, with fewer trophic transfers between phyto-
plankton and predators. Large blooms of rapidly
sinking diatoms, slower turnover, and sloppy grazing
by large zooplankton result in large fluxes of POC to
benthic ecosystems (Ryther 1969, Cermeño et al.
2008, Guidi et al. 2009, Chavez et al. 2011, Norris et
al. 2013). There is evidence for phytoplankton in -
creases in most nearshore waters (Fig. S2 in the Sup-
plement), likely due to human-derived nutrient input
(Jickells 1998). Increasing phytoplankton in these
nearshore systems is hypothesized to have a positive
effect on global fishery landings, ~50% of which de -
rive from nearshore and shelf systems (FAO 2010),
but may also trigger negative effects in some regions.
For instance, large phytoplankton blooms are known
to increase heterotrophic bacterial activity and can
lead to large subsurface anoxic regions known as
dead zones (Grantham et al. 2004). Such effects have
been linked with decreased secondary biomass and
fishery yield (Diaz & Rosenberg 2008), particularly in
nearshore waters. Additionally, some phytoplankton
species can form harmful algal blooms, which nega-
tively affect secondary production and fisheries
(Nixon & Pilson 1983).
Upwelling ecosystems occur in both nearshore and
oceanic waters and contain characteristics of both
systems. These ecosystems are predominantly influ-
enced by the wind-driven upwelling of nutrient-rich
waters, resulting in large blooms of microphyto-
plankton, which support large fisheries and export
large amounts of POM to the deep sea. Contrary to
nearshore systems, phytoplankton trends in up -
welling systems are mostly related by changes in up -
welling intensity as driven by changes in wind, tem-
perature, and stratification. Any increases in TDS
here would reduce total phytoplankton biomass but
may have disproportionate negative effects on larger
phytoplankton, which are outcompeted by pico -
phyto plankton under conditions of warming, stratifi-
cation, or prolonged nutrient limitation (Atkinson et
al. 2003, Cermeño et al. 2008, Li et al. 2009, Morán et
al. 2010). Since large grazers in these systems are
often incapable of consuming picophytoplankton
(Hansen et al. 1994, Sommer & Stibor 2002, Sommer
& Sommer 2006), a shift towards smaller phytoplank-
ton may decrease the transfer efficiency of primary
production through the grazer food chain (Ryther
1969, Barnes et al. 2010, Chavez et al. 2011). These
size-selective negative effects are predicted to be
strongest in the North Atlantic and tropical up -
welling systems, possibly due to the proportionally
larger contribution of microphytoplankton to phyto-
plankton standing stock (Cermeño et al. 2008).
Studies of the relationship between phytoplankton
changes and fisheries landings confirm these obser-
vations, with the average effect of changing chloro-
phyll on fish yield being strongest in upwelling, tem-
perate, and nearshore marine ecosystems (Ware &
Thomson 2005, Chassot et al. 2007, 2010). Although
model predictions for upwelling systems are variable
and uncertain (Wang et al. 2010), many predict tem-
perature-driven future declines in phytoplankton
biomass and size (i.e. Henson et al. 2010, Steinacher
et al. 2010). Such changes are hypothesized to have
strong and negative effects on productivity.
Case study 2: Ecological effects of climate-driven
phytoplankton variability
Some of the clearest examples of the drivers and
ecological consequences of marine phytoplankton
change derive from studies of the effects of quasi-
periodic climate fluctuations, for example from the El
Nino Southern Oscillation or North Atlantic Oscilla-
tion (NAO; Barber & Chavez 1986, Chavez et al.
1999, Behrenfeld et al. 2006, Martinez et al. 2009).
Such climate fluctuations represent natural experi-
ments which can shed light on the drivers and conse-
quences of longer-term trends in phytoplankton
change.
A well-known example of the effects of climate
variability propagating up the food web comes from
the North Sea (Aebischer et al. 1990). Here, the posi-
tive correspondence between standardized long-
term (1955 to 1987) time series of westerly weather,
phytoplankton, zooplankton, herring Clupea haren-
gus abundance, and breeding success of kittiwakes
Rissa tridactyla suggests that environmental effects
on phytoplankton abundance are transmitted up the
food chain. Although the study accounted for the
influence of weather patterns, the potential effects of
periodic climate variability were not realized at the
time. The NAO is a major mode of climate variability
in the region and is negatively related to the average
concentration of phytoplankton (Boyce et al. 2010)
and zooplankton (Fromentin & Planque 1996) in the
265

Mar Ecol Prog Ser 534: 251–272, 2015
North Atlantic. On longer time scales, the Atlantic
Multidecadal Oscillation (AMO) may be the domi-
nant mode of climate variability (Martinez et al. 2009,
Chavez et al. 2011) and is positively related to marine
phytoplankton concentration in the region (Martinez
et al. 2009). To examine the interplay between
decadal climate fluctuations, plankton abundance,
and ecosystem structure, we extracted time series of
westerly weather, phytoplankton, zooplankton, her-
ring, and kittiwake chicks (Aebischer et al. 1990) as
well as time series for the NAO and AMO. All series
were filtered to remove any high-frequency variabil-
ity and re-scaled such that they ranged over the same
interval (see the Supplement for data sources and full
methods). Westerly weather had a low explanatory
power and was thus removed from the analysis. All
series were positively related (Fig. 7), yet the AMO
emerged as the dominant climate driver of observed
ecological change, showing a much stronger correla-
tion than the NAO or westerly weather. However, it
is unclear if the AMO alters ecosystem dynamics
directly through physical processes or indirectly by
modifying the trophic state of the environment.
To more quantitatively address this issue, we ex -
amined the linear correlation between all series. If
climate is driving consumer abundance via changes
in phytoplankton, the correlation between adjacent,
trophically coupled trophic levels should be stron -
ger than the correlation between individual trophic
levels and climate. Using this simple approach, we
ob served strong evidence of bottom-up effects
mediated by the influence of climate on phyto-
plankton (Fig. S3 in the Supplement). Climate in -
dices were the strongest predictors of phytoplank-
ton concentration. For example, the AMO shows
almost perfect positive correlation (NAO: r = −0.596;
AMO: r = 0.998; Fig. S3). Zooplankton and herring
were best predicted by the concentration of phyto-
plankton on which they graze (zooplankton: r =
0.961; herring: r = 0.781). Also, the number of kitti-
wake chicks was most strongly predic ted by their
primary food source, herring (r = 0.989). While
these correlations do not imply causation, they do
provide observational support for the hypo thesis of
climate-induced control of the eco system and sug-
gest that long-term changes in phytoplankton
could cascade up the food web, ultimately influ-
encing apex predators and humans. The strong
influence of the AMO particularly highlights the
importance of de cadal- scale temperature variation
in determining phytoplankton concentration in the
upper ocean.
SUMMARY AND OUTLOOK
Our analysis suggests that the high variability
among estimated changes in marine phytoplankton
over the past century likely relates to a larger num-
ber of local and regional factors that cannot be easily
identified in a global overview. However, when we
compiled published trend estimates from throughout
the global ocean, we tended to observe declining
phytoplankton concentrations more commonly in
studies conducted over longer time
scales and in the open oceans. Con-
versely, phytoplankton increases
were observed more frequently over
recent time periods and closer to
shore. Regionally, our analysis sug-
gests that phytoplankton concentra-
tions have de clined across the North
and equatorial Pacific Oceans and at
high latitudes and increased in the
South Indian Ocean and in the North-
east Atlantic. Estimates of change in
the North Atlantic Ocean and in the
Southern Hemisphere appeared par-
ticularly variable. Continued monitor-
ing of phytoplankton levels using
standardized methods, such as in situ,
remote sensing, Secchi disk, and
CPR, will lead to improved inter-cali-
bration and more accurate estimates
of long-term phytoplankton changes.
266
1960 1965 1970 1975 1980 1985
−1.0
−0.5
0.0
0.5
1.0
1.5 Inverse NAO
AMO
Phytoplankton
Zooplankton
Herring
Kittiwake
Year
Scaled index
Fig. 7. Bottom-up cascade driven by low-frequency climate effects on phyto-
plankton. Time series of climate and abundance across multiple trophic levels
in the North Sea. Dashed lines represent climate indices, and colors depict
different trophic levels within the food web. Biological time series were
extracted from Aebischer et al. (1990). Series were smoothed with a moving
average and normalized between −1 and 1 (see Supplement for full details).
NAO: North Atlantic Oscillation; AMO: Atlantic Multidecadal Oscillation

Boyce & Worm: Marine phytoplankton change
While empirical estimates vary widely, most pre-
dictive models suggest that globally averaged phyto-
plankton concentrations will gradually decline over
the coming century (Table 3). Although increases are
predicted at high latitudes and in nearshore waters,
global trends will likely be dominated by phyto-
plankton declines across the low- and mid-latitude
oceans and in the open oceans, where ~82% of an -
nual global ocean primary production occurs (Ryther
1969).
Multiple lines of evidence point towards changes
in temperature as an important (but certainly not
exclusive) driver of observed phytoplankton trends,
particularly in the open oceans. Increasing tempera-
tures are predicted to induce shifts in phytoplankton
biomass concentration (Behrenfeld et al. 2006), diver-
sity (Thomas et al. 2012), phenology (D’Ortenzio et
al. 2012), species composition (Cermeño et al. 2008,
Li et al. 2009), size structure (Polovina & Woodworth
2012), and zooplankton grazing pressure (O’Connor
et al. 2009, Sommer et al. 2012, Boyce et al. 2015b).
The pathways by which temperature changes influ-
ence phytoplankton are multifarious, but TDS has
emerged as an important mechanism over geological
(Schmittner 2005), historical (Boyce et al. 2010, 2014,
Boyce 2013), contemporary (Behrenfeld et al. 2006),
and future (Henson et al. 2010, Hofmann et al. 2011,
Olonscheck et al. 2013) time horizons. Experiments
and process models further suggest that warming is
shifting the balance of autotrophic to heterotrophic
metabolism, which may exacerbate (O’Connor et al.
2009, Sommer et al. 2012, Olonscheck et al. 2013) or
counterbalance (Taucher & Oschlies 2011) any de -
clines in biomass driven by TDS.
Such changes in plankton abundance, composi-
tion, and diversity, variable as they may be, will like -
ly have effects on the wider ocean food web. For
example, progressive declines in phytoplankton bio-
mass would likely reduce the carrying capacity of
marine ecosystems if not counterbalanced by in crea -
ses in biomass-specific productivity. A robust exami-
nation of future phytoplankton change and its eco-
logical consequences will depend on better resolving
critical uncertainties, such as the influence of con-
sumers on marine phytoplankton, the net effect of
changing metabolic rates on productivity, and the
effects of size-restructured phytoplankton communi-
ties on ecosystem functioning. Further, the scarcity of
consistent, long-term measurements of consumer
abundance across trophic levels limits any rigorous
analysis of their relevance as drivers of long-term
phytoplankton change. Such challenges and limita-
tions may explain why relatively few studies have
considered the importance of top-down effects at
global scales. Further investigation may be facili-
tated by combining process-based models with ex -
perimentation and field observation and through the
formation of coordinated working groups (i.e. ICES
and Scientific Committee on Oceanic Research wor -
king groups and the International Group for Marine
Ecological Time Series) aimed at integrating, shar-
ing, and validating phytoplankton time series.
Acknowledgements. We are very grateful to all data pro -
viders and to Kenneth Frank for providing helpful comments
and suggestions and Michael Dowd for reviewing the paper
and providing statistical expertise and editorial comments.
Funding was provided by the Natural Sciences and Engi-
neering Research Council of Canada.
LITERATURE CITED
Aebischer NJ, Coulson JC, Colebrook JM (1990) Parallel
long-term trends across four marine trophic levels and
weather. Nature 347: 753−755
Agusti S, Satta MP, Mura MP, Benavent E (1998) Dissolved
esterase activity as a tracer of phytoplankton lysis: evi-
dence of high phytoplankton lysis rates in the northwest-
ern Mediterranean. Limnol Oceanogr 43: 1836−1849
Aksnes DL, Ohman MD (2009) Multi-decadal shoaling of
the euphotic zone in the southern sector of the California
Current System. Limnol Oceanogr 54: 1272−1281
Antoine D, Morel A, Gordon HR, Banzon VF, Evans RH
(2005) Bridging ocean color observations of the 1980s
and 2000s in search of long-term trends. J Geophys Res
110: 1−22
Arrigo KR, Perovich DK, Pickart RS, Brown ZW and others
(2012) Massive phytoplankton blooms under Arctic Sea
ice. Science 336: 1408
Atkinson D (1994) Temperature and organism size — a bio-
logical law for ectotherms? Adv Ecol Res 25: 1−58
Atkinson D, Ciotti BJ, Montagnes DJS (2003) Protists
decrease in size linearly with temperature: ca. 2.5% °C−1.
Proc R Soc B 270: 2605−2611
Azam F, Malfatti F (2007) Microbial structuring of marine
ecosystems. Nat Rev Microbiol 5: 782−791
Azam F, Worden AZ (2004) Microbes, molecules, and mar-
ine ecosystems. Science 303: 1622−1624
Azam F, Fenchel T, Field JG, Gray JS, Meyer-Reil LA,
Thingstad F (1983) The ecological role of water-column
microbes in the sea. Mar Ecol Prog Ser 10: 257−263
Bakun A (1990) Global climate change and intensification of
coastal ocean upwelling. Science 247: 198−201
Barber RT, Chavez FR (1986) Ocean variability in relation to
living resources during the 1982−83 El Nino. Nature 319:
279−285
Barnes C, Maxwell D, Reuman DC, Jennings S (2010)
Global patterns in predator−prey size relationships re -
veal size dependency of trophic transfer efficiency.
Ecology 91: 222−232
Barnes C, Irigoien X, De Oliveira JAA, Maxwell D, Jennings
S (2011) Predicting marine phytoplankton community
size structure from empirical relationships with remotely
sensed variables. J Plankton Res 33: 13−24
267
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤

Mar Ecol Prog Ser 534: 251–272, 2015
Baum JK, Worm B (2009) Cascading top-down effects of
changing oceanic predator abundances. J Anim Ecol 78:
699−714
Beaufort L, Probert I, de Garidel-Thoron T, Bendif EM and
others (2011) Sensitivity of coccolithophores to carbonate
chemistry and ocean acidification. Nature 476: 80−83
Beaugrand G, Brander KM, Alistair Lindley J, Souissi S,
Reid PC (2003) Plankton effect on cod recruitment in the
North Sea. Nature 426: 661−664
Beaulieu C, Henson SA, Sarmiento JL, Dunne JP, Doney SC,
Rykaczewski RR, Bopp L (2013) Factors challenging our
ability to detect long-term trends in ocean chlorophyll.
Biogeosciences 10: 2711−2724
Behrenfeld MJ (2010) Abandoning Sverdrup’s critical depth
hypothesis on phytoplankton blooms. Ecology 91:
977−989
Behrenfeld MJ, Falkowski PG (1997) Photosynthetic rates
derived from satellite-based chlorophyll concentration.
Limnol Oceanogr 42: 1−20
Behrenfeld M, Bale AJ, Kolber ZS, Aiken J, Falkowski PG
(1996) Confirmation of iron limitation of phytoplankton
photosynthesis in the equatorial Pacific Ocean. Nature
383: 508−511
Behrenfeld MJ, O’Malley RT, Siegel DA, McClain CR and
others (2006) Climate-driven trends in contemporary
ocean productivity. Nature 444: 752−755
Bidle KD, Falkowski PG (2004) Cell death in planktonic,
photosynthetic microorganisms. Nat Rev Microbiol 2:
643−655
Bopp L, Monfray P, Aumont O, Dufresne JLL and others
(2001) Potential impact of climate change on marine
export production. Global Biogeochem Cycles 15: 81−99
Bopp L, Aumont O, Cadule P, Alvain S, Gehlen M (2005)
Response of diatoms distribution to global warming and
potential implications: a global model study. Geophys
Res Lett 32: L19606, doi: 10.1029/2005GL023653
Boyce DG (2013) Patterns and drivers of marine phytoplank-
ton changes over the past century. PhD dissertation, Dal-
housie University, Halifax, NS
Boyce DG, Lewis MR, Worm B (2010) Global phytoplankton
decline over the past century. Nature 466: 591−596
Boyce DG, Lewis M, Worm B (2012) Integrating global
chlorophyll data from 1890 to 2010. Limnol Oceanogr
Methods 10: 840−852
Boyce DG, Dowd M, Lewis MR, Worm B (2014) Estimating
global chlorophyll changes over the past century. Prog
Oceanogr 122: 163−173
Boyce DG, Frank KT, Leggett WC (2015a) From mice to ele-
phants: overturning the ‘one size fits all’ paradigm in
marine plankton food chains. Ecol Lett 18: 504−515
Boyce DG, Frank KT, Worm B, Leggett WC (2015b) Spatial
patterns and predictors of trophic control across marine
ecosystems. Ecol Lett, doi:10.1111/ele.12481
Boyd PW, Doney SC (2002) Modelling regional responses by
marine pelagic ecosystems to global climate change.
Geophys Res Lett 29: 1806−1810
Boyd PW, Watson AJ, Law CS, Abraham ER and others
(2000) A mesoscale phytoplankton bloom in the polar
Southern Ocean stimulated by iron fertilization. Nature
407: 695−702
Brimblecombe P, Pitman J (1980) Long-term deposit at
Rothamsted, southern England. Tellus 32: 261−267
Brown JH, Gillooly JF, Allen AP, Savage VM, West GB
(2004) Toward a metabolic theory of ecology. Ecology 85:
1771−1789
Capone DG, Zehr JP, Paerl HW, Bergman B, Carpenter EJ
(1997) Trichodesmium, a globally significant marine
cyanobacterium. Science 276: 1221−1229
Casini M, Lovgren J, Hjelm J, Cardinale M, Molinero JCC,
Kornilovs G, Lövgren J (2008) Multi-level trophic cas-
cades in a heavily exploited open marine ecosystem.
Proc R Soc B 275: 1793−1801
Cermeño P, Dutkiewicz S, Harris RP, Follows M, Schofield
O, Falkowski PG (2008) The role of nutricline depth in
regulating the ocean carbon cycle. Proc Natl Acad Sci
USA 105: 20344−20349
Charlson RJ, Lovelock JE, Andreae MO, Warren SG (1987)
Oceanic phytoplankton, atmospheric sulphur, cloud al -
bedo and climate. Nature 326: 655−661
Charpy-Roubaud C, Sournia A (1990) The comparative
estima tion of phytoplanktonic, microphytobenthic and
macro phytobenthic primary production in the oceans.
Mar Microb Food Webs 4: 31−57
Chassot E, Mélin F, Le Pape O, Gascuel D (2007) Bottom-up
control regulates fisheries production at the scale of eco-
regions in European seas. Mar Ecol Prog Ser 343: 45−55
Chassot E, Bonhommeau S, Dulvy NK, Mélin F, Watson R,
Gascuel D, Le Pape O (2010) Global marine primary pro-
duction constrains fisheries catches. Ecol Lett 13: 495−505
Chavez FP, Toggweiler JR (1995) Physical estimates of
global new production: the upwelling contribution. In:
Summerhayes CP, Emeis KC, Angel MV, Smith RL,
Zeitzschel B (eds) Upwelling in the ocean: modern pro-
cesses and ancient records. John Wiley & Sons, New
York, NY, p 313−320
Chavez FP, Strutton PG, Friederich CE, Feely RA, Feldman
GC, Foley DC, McPhaden MJ (1999) Biological and
chemical response of the equatorial Pacific Ocean to the
1997-98 El Nino. Science 286: 2126−2131
Chavez FP, Ryan J, Lluch-Cota SE, Niquen C (2003) From
anchovies to sardines and back: multidecadal change in
the Pacific Ocean. Science 299: 217−221
Chavez FP, Messie M, Pennington JT (2011) Marine primary
production in relation to climate variability and change.
Annu Rev Mar Sci 3: 227−260
Cox PM, Betts RA, Jones CD, Spall SA, Totterdell IJ (2000)
Acceleration of global warming due to carbon-cycle
feedbacks in a coupled climate model. Nature 408:
184−187
Cressie NAC (1993) Statistics for spatial data. John Wiley &
Sons, New York, NY
Cushing DH (1990) Plankton production and year-class
strength in fish populations: an update of the match/
mismatch hypothesis. Adv Mar Biol 26: 249−293
D’Ortenzio F, Antoine D, Martinez E, Ribera d’Alcalà M
(2012) Phenological changes of oceanic phytoplankton in
the 1980s and 2000s as revealed by remotely sensed
ocean-color observations. Global Biogeochem Cycles 26:
GB4003, doi: 10.1029/2011GB004269
Dewar WK, Bingham RJ, Iverson RL, Nowacek DP, St. Lau-
rent LC, Wiebe PH (2006) Does the marine biosphere mix
the ocean? J Mar Res 64: 541−561
Diaz RJ, Rosenberg R (2008) Spreading dead zones and con-
sequences for marine ecosystems. Science 321: 926−929
Duce RA, Liss PS, Merrill JT, Buat-Menard P and others
(1991) The atmospheric input of trace species to the
world ocean. Global Biogeochem Cycles 5: 193−259
Edwards M, Richardson AJ (2004) Impact of climate change
on marine pelagic phenology and trophic mismatch.
Nature 430: 881−884
268
➤
➤
➤
➤
➤
➤
➤
➤
➤➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤

Boyce & Worm: Marine phytoplankton change
Estes JA, Terborgh J, Brashares JS, Power ME and others
(2011) Trophic downgrading of planet Earth. Science
333: 301−306
Falkowski PG, Wilson C (1992) Phytoplankton productivity
in the North Pacific Ocean since 1900 and implications
for absorption of anthropogenic CO2. Nature 358:
741−743
Falkowski PG, Barber RT, Smetacek V (1998) Biogeochemi-
cal controls and feedbacks on ocean primary production.
Science 281: 200−206
Falkowski PG, Katz ME, Knoll AH, Quigg A, Raven JA,
Schofield O, Taylor FJR (2004) The evolution of modern
eukaryotic phytoplankton. Science 305: 354−360
FAO (Food and Agriculture Organization of the United
Nations) (2010) The state of world fisheries and aquacul-
ture 2010. FAO Fisheries and Aquaculture Department,
FAO, Rome
Field CB, Behrenfeld MJ, Randerson JT (1998) Primary pro-
duction of the biosphere: integrating terrestrial and
oceanic components. Science 281: 237−240
Forel FA (1890) Une nouvelle forme de la gamme de couleur
pour l’etude de l’eau des lacs. Bull Soc Vaud Sci Nat 25: 6
Frank KT, Petrie B, Choi JS, Leggett WC (2005) Trophic cas-
cades in a formerly cod-dominated ecosystem. Science
308: 1621−1623
Frank KT, Petrie B, Shackell NL, Choi JS (2006) Reconciling
differences in trophic control in mid-latitude marine eco-
systems. Ecol Lett 9: 1096−1105
Frank KT, Petrie B, Fisher JA, Leggett WC (2011) Transient
dynamics of an altered large marine ecosystem. Nature
477: 86−89
Frederiksen M, Edwards M, Richardson AJ, Halliday NC,
Wanless S (2006) From plankton to top predators: bot-
tom-up control of a marine food web across four trophic
levels. J Anim Ecol 75: 1259−1268
Fromentin JM, Planque B (1996) Calanus and environment
in the eastern North Atlantic. II. Influence of the North
Atlantic Oscillation on C. finmarchicus and C. helgo -
landicus. Mar Ecol Prog Ser 134: 111−118
Geider RJ (1987) Light and temperature-dependence of the
carbon to chlorophyll aratio in microalgae and cyano-
bacteria: implications for physiology and growth of
phytoplankton. New Phytol 106: 1−34
Gjosaeter J, Kawaguchi K (1980) A review of the world
resources of mesopelagic fish. FAO Fish Tech Pap 193:
151
Gnanadesikan A, Emmanuel K, Vecchi GA, Anderson WG,
Hallberg R (2010) How ocean color can steer Pacific trop-
ical cyclones. Geophys Res Lett 37: L18802
Goes JI, Thoppil PG, Gomes HR (2005) Warming of the
Eurasian landmass is making the Arabian Sea more pro-
ductive. Science 308: 545−547
Grantham BA, Chan F, Nielsen KJ, Fox DS and others (2004)
Upwelling-driven nearshore hypoxia signals ecosystem
and oceanographic changes in the northeast Pacific.
Nature 429: 749−754
Gregg WW, Conkright ME (2002) Decadal changes in global
ocean chlorophyll. Geophys Res Lett 29: 1730−1734
Gregg WW, Casey NW, McClain CR (2005) Recent trends in
global ocean chlorophyll. Geophys Res Lett 32: 1−5
Guidi L, Stemmann L, Jackson GA, Ibanez FF and others
(2009) Effects of phytoplankton community on produc-
tion, size and export of large aggregates: a world-ocean
analysis. Limnol Oceanogr 54: 1951−1963
Hansen B, Bjornsen PK, Hansen PJ (1994) The size ratio
between planktonic predators and their prey. Limnol
Oceanogr 39: 395−403
Head EJH, Pepin P (2010) Monitoring changes in phyto-
plankton abundance and composition in the Northwest
Atlantic: a comparison of results obtained by continuous
plankton recorder sampling and colour satellite imagery.
J Plankton Res 32: 1649−1660
Henson SA, Sarmiento JL, Dunne JP, Bopp L and others
(2010) Detection of anthropogenic climate change in
satellite records of ocean chlorophyll and productivity.
Biogeosciences 7: 621−640
Hillebrand H, Cardinale BJ (2004) Consumer effects decline
with prey diversity. Ecol Lett 7: 192−201
Hjort J (1914) Fluctuations in the great fisheries of northern
Europe viewed in the light of biological research. Rapp
P-v Réun Cons Perm Int Explor Mer 20: 1−288
Hofmann M, Worm B, Rahmstorf S, Schellnhuber HJ (2011)
Declining ocean chlorophyll under unabated anthro-
pogenic CO2emissions. Environ Res Lett 6: 034035
Hooper DU, Adair EC, Cardinale BJ, Byrnes JE and others
(2012) A global synthesis reveals biodiversity loss as a
major driver of ecosystem change. Nature 486: 105−108
Howarth RW, Billen G, Townsend A, Jaworski N and others
(1996) Regional nitrogen budgets and riverine N & P
fluxes for the drainages to the North Atlantic Ocean: nat-
ural and human influences. Biogeochemistry 35: 75−139
Huot Y, Babin M, Bruyant F, Grob C, Twardowski MS,
Claustre H (2007) Does chlorophyll aprovide the best
index of phytoplankton biomass for primary productivity
studies? Biogeosciences Discuss 4: 707−745
Iglesias-Rodriguez MD, Halloran PR, Rickaby REM, Hall IR
and others (2008) Phytoplankton calcification in a high-
CO2world. Science 320: 336−340
Irigoien X, Huisman J, Harris RP (2004) Global biodiversity
patterns of marine phytoplankton and zooplankton.
Nature 429: 863−867
Irigoien X, Klevjer TA, Røstad A, Martinez U and others
(2014) Large mesopelagic fishes biomass and trophic
efficiency in the open ocean. Nat Commun 5: 3271
Iverson RL (1990) Control of marine fish production. Limnol
Oceanogr 35: 1593−1604
Jacobs SS, Giulivi CF, Mele PA (2002) Freshening of the
Ross Sea during the late 20th century. Science 297:
386−389
Jickells TD (1998) Nutrient biogeochemistry of the coastal
zone. Science 281: 217−221
Johnson NA, Campbell JW, Moore TS, Rex MA, Etter RJ,
McClain CR, Dowell MD (2007) The relationship be -
tween the standing stock of deep-sea macrobenthos and
surface production in the western North Atlantic. Deep-
Sea Res Part I 54: 1350−1360
Karl DM, Bidigare RR, Letelier RM (2001) Long-term
changes in plankton community structure and productiv-
ity in the North Pacific subtropical gyre: the domain shift
hypothesis. Deep-Sea Res II 48: 1149−1470
Kunze E, Dower JF, Beveridge I, Dewey R, Bartlett KP (2006)
Observations of biologically generated turbulence in a
coastal inlet. Science 313: 1768−1770
Lasker R (1975) Field criteria for survival of anchovy larvae:
the relation between inshore chlorophyll maximum lay-
ers and successful first feeding. Fish Bull 73: 453−462
Lavery TJ, Roudnew B, Gill P, Seymour J and others (2010)
Iron defecation by sperm whales stimulates carbon
export in the Southern Ocean. Proc R Soc B 277:
3527−3531
269
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤

Mar Ecol Prog Ser 534: 251–272, 2015
Lavery TJ, Roudnew B, Seymour J, Mitchell JG, Smetacek
V, Nicol S (2014) Whales sustain fisheries: blue whales
stimulate primary production in the Southern Ocean.
Mar Mamm Sci 30: 888−904
Lee SH, Joo HM, Liu Z, Chen J, He J (2012) Phytoplankton
productivity in newly opened waters of the western Arc-
tic Ocean. Deep-Sea Res II 81−84: 18−27
Lewandowska AM, Boyce DDGD, Hofmann M, Matthiessen
B, Sommer U, Worm B (2014) Effects of sea surface
warming on marine plankton. Ecol Lett 17: 614−623
Li WKW, McLaughlin FA, Lovejoy C, Carmack EC (2009)
Smallest algae thrive as the Arctic Ocean freshens. Sci-
ence 326: 539
Llewellyn CA, Tarran GA, Galliene CP, Cummings DG and
others (2008) Microbial dynamics during the decline of a
spring diatom bloom in the northeast Atlantic. J Plankton
Res 30: 261−273
Loeb V, Siegel V, Holm-Hansen O, Hewitt R, Fraser W, Triv-
elpiece W, Trivelpiece S (1997) Effects of sea-ice extent
and krill or salp dominance on the Antarctic food web.
Nature 387: 897−900
Lozier MS, Dave AC, Palter JB, Gerber LM, Barber RT
(2011) On the relationship between stratification and pri-
mary productivity in the North Atlantic. Geophys Res
Lett 38: L18609
Margalef R (1978) Life-forms of phytoplankton as survival
alternatives in an unstable environment. Oceanol Acta 1:
493−509
Martinez E, Antoine D, D’Ortenzio F, Gentili B (2009) Cli-
mate-driven basin-scale decadal oscillations of oceanic
phytoplankton. Science 326: 1253−1256
McClain CR, Signorini SR (2004) Subtropical gyre variability
observed by ocean-color satellites. Deep-Sea Res II 51:
281−301
McGillicuddy DJ Jr, Anderson LA, Bates NR, Bibby T and
others (2007) Eddy/wind interactions stimulate extra -
ordinary mid-ocean plankton blooms. Science 316:
1021−1026
McQuatters-Gollop A, Raitsos DE, Edwards M, Pradhan Y,
Mee LD, Lavender SJ, Attrill MJ (2007) A long-term
chlorophyll data set reveals regime shift in North Sea
phytoplankton biomass unconnected to nutrient trends.
Limnol Oceanogr 52: 635−648
McQuatters-Gollop A, Reid PC, Edwards M, Burkhill PH
and others (2011) Is there a decline in marine phyto-
plankton? Nature 472: E6−E7
McQueen DJ, Post JR, Mills EL, Post R (1986) Trophic rela-
tionships in freshwater pelagic ecosystems. Can J Fish
Aquat Sci 43: 1571−1581
Micheli F (1999) Eutrophication, fisheries, and consumer-
resource dynamics in marine pelagic ecosystems. Sci-
ence 285: 1396−1398
Montes-Hugo M, Doney SC, Ducklow HW, Fraser W, Mar-
tinson D, Stammerjohn SE, Schofield O (2009) Recent
changes in phytoplankton communities associated with
rapid regional climate change along the western Ant -
arctic Peninsula. Science 323: 1470−1473
Moore CM, Mills MM, Achterberg EP, Geider RJ and oth-
ers (2009) Large-scale distribution of Atlantic nitrogen
fixation controlled by iron availability. Nat Geosci 2:
867−871
Mora C, Wei CL, Rollo A, Amaro T and others (2013) Biotic
and human vulnerability to projected changes in ocean
biogeochemistry over the 21st century. PLoS Biol 11:
e1001682
Morán XAG, López-Urrutia Á, Calvo-Díaz A, Li WKW
(2010) Increasing importance of small phytoplankton in a
warmer ocean. Glob Change Biol 16: 1137−1144
Motoda NS, Kotori M, Tahara H (1987) Long-term phyto-
plankton changes in Oshoro Bay, Hokkaido, and Matoya
Bay, central Honshu, Japan. Bull Mar Sci 41: 523−530
Munk WH (1966) Abyssal recipes. Deep-Sea Res 13:
707−730
Murphy E, Morris D, Watkins J, Priddle J (1988) Scales of
interaction between Antarctic krill and the environment.
In: Sahrhage D (ed) Antarctic ocean and resources vari-
ability. Springer-Verlag, Berlin, p 120−303
Murtugudde R, Beauchamp RJ, McClain CR, Lewis MR,
Busalacchi A (2002) Effects of penetrative radiation on
the upper tropical ocean circulation. J Clim 15: 470−486
Myers RA, Baum JK, Shepherd TD, Powers SP, Peterson CH
(2007) Cascading effects of the loss of apex predatory
sharks from a coastal ocean. Science 315: 1846−1850
Nixon SW, Pilson MEQ (1983) Nitrogen in estuarine and
coastal marine systems. In: Carpenter EJ, Capone DG
(eds) Nitrogen in the marine environment. Academic
Press, New York, NY
Norris RD, Turner SK, Hull PM, Ridgwell A (2013) Marine
ecosystem responses to Cenozoic global change. Science
341: 492−498
O’Connor MI, Piehler MF, Leech DM, Anton A, Bruno JF
(2009) Warming and resource availability shift food web
structure and metabolism. PLoS Biol 7: e1000178
Olonscheck D, Hofmann M, Worm B, Schellnhuber HJ
(2013) Decomposing the effects of ocean warming on
chlorophyll aconcentrations into physically and biologi-
cally driven contributions. Environ Res Lett 8: 014043
Orr JC, Fabry VJ, Aumont O, Bopp L and others (2005)
Anthropogenic ocean acidification over the twenty-first
century and its impact on calcifying organisms. Nature
437: 681−686
Oschlies A, Garcon V (1998) Eddy-induced enhancement of
primary production in a model of the North Atlantic
Ocean. Nature 394: 266−269
Otero J, Alvarez-Salgado XA, Gonzalez AF, Miranda A and
others (2008) Bottom-up control of common octopus
Octopus vulgaris in the Galician upwelling system,
north east Atlantic Ocean. Mar Ecol Prog Ser 362:
181−192
Platt T, Fuentes-Yaco C, Frank T (2003) Spring algal bloom
and larval fish survival. Nature 423: 398−399
Polovina JJ, Woodworth PA (2012) Declines in phytoplank-
ton cell size in the subtropical oceans estimated from
satellite remotely-sensed temperature and chlorophyll,
1998−2007. Deep-Sea Res II 77-80: 82−88
Polovina JJ, Howell EA, Abecassis M (2008) Ocean’s least
productive waters are expanding. Geophys Res Lett 35:
L03618
Polovina JJ, Dunne JP, Woodworth PA, Howell EA (2011)
Projected expansion of the subtropical biome and con-
traction of the temperate and equatorial upwelling bio-
mes in the North Pacific under global warming. ICES J
Mar Sci 68: 986−995
Pomeroy LR, Williams PJI, Azam F, Hobbie JE (2007) The
microbial loop. Oceanography 20: 28−33
Post E, Bhatt US, Bitz CM, Brodie JF and others (2013) Eco-
logical consequences of sea-ice decline. Science 341:
519−524
Raitsos DE, Reid PC, Lavender SJ, Edwards M, Richardson
AJ (2005) Extending the SeaWiFS chlorophyll data set
270
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤

Boyce & Worm: Marine phytoplankton change
back 50 years in the northeast Atlantic. Geophys Res Lett
32: 1−4
Redfield AC (1958) The biological control of chemical factors
in the environment. Am Sci 46: 205−221
Reich PB, Tilman D, Isbell F, Mueller K, Hobbie SE, Flynn
DFB, Eisenhauer N (2012) Impacts of biodiversity loss
escalate through time as redundancy fades. Science 336:
589−592
Richardson AJ, Schoeman DS (2004) Climate impact on
plankton ecosystems in the northeast Atlantic. Science
305: 1609−1612
Riebesell U, Gattuso JP, Thingstad TF, Middelburg JJ (2013)
Arctic ocean acidification: pelagic ecosystem and bio-
geochemical dynamics responses during a mesocosm
study. Biogeosciences 10: 5619−5626
Rodríguez J, Tintore J, Allen JT, Blanco JM and others
(2001) Mesoscale vertical motion and the size structure
of phytoplankton in the ocean. Nature 410: 360−363
Roman J, McCarthy JJ (2010) The whale pump: marine
mammals enhance primary productivity in a coastal
basin. PLoS ONE 5: e13255
Roman J, Palumbi SR (2003) Whales before whaling in the
North Atlantic. Science 301: 508−510
Romero OE, Leduc G, Vidal L, Fischer G (2011) Millennial
variability and long-term changes of the diatom produc-
tion in the eastern equatorial Pacific during the last gla-
cial cycle. Paleoceanography 26: PA2212, doi: 10.1029/
2010PA002099
Ruhl HA, Ellena JA, Smith KL (2008) Connections between
climate, food limitation, and carbon cycling in abyssal
sediment communities. Proc Natl Acad Sci USA 105:
17006−17011
Ryther JH (1969) Photosynthesis and fish production in the
sea. Science 166: 72−76
Saba VS, Spotila JR, Chavez FP, Musick JA (2008) Bottom-
up and climatic forcing on the worldwide population of
leatherback turtles. Ecology 89: 1414−1427
Saba VS, Freidrichs MAM, Carr ME, Antoine D and others
(2010) Challenges of modeling depth-integrated marine
primary productivity over multiple decades: a case study
at BATS and HOT. Global Biogeochem Cycles 24:
GB3020, doi: 10.1029/2009GB003655
Sarmiento JL, Slater R, Barber R, Bopp L and others (2004)
Response of ocean ecosystems to climate warming. Glo -
bal Biogeochem Cycles 18: GB3003, doi: 10.1029/ 2003
GB002134
Schmittner A (2005) Decline of the marine ecosystem caused
by a reduction in the Atlantic overturning circulation.
Nature 434: 628−633
Schmittner A, Oschlies A, Matthews HD, Galbraith ED
(2008) Future changes in climate, ocean circulation, eco-
systems, and biogeochemical cycling simulated for a
business-as-usual CO2emission scenario until year 4000
AD. Global Biogeochem Cycles 23: GB1013
Secchi PA (1886) Relazione delle esperienze fatte a bordo
della pontificia pirocorvetta l‘Imacolata concezione per
determinare la trasparenza del mare. In: Cialdi A (ed)
Sul moto ondoso del mare e su le correnti di esso special-
mente su quelle littorali, 2nd edn. Tipografia delle Belle
Arti, Rome, p 258−288
Sheldon RW, Sutcliff WH, Prakash A (1972) Size distribution
of particles in the ocean. Limnol Oceanogr 17: 327−340
Shi D, Xu Y, Hopkinson BM, Morel FMM (2010) Effect of
ocean acidification on iron availability to marine phyto-
plankton. Science 327: 676−679
Shiomoto A, Tadokoro K, Nagasawa K, Ishida Y (1997)
Trophic relations in the subarctic North Pacific ecosys-
tem: possible feeding effect from pink salmon. Mar Ecol
Prog Ser 150: 75−85
Shurin JB, Borer ET, Seabloom EW, Anderson K and others
(2002) A cross-ecosystem comparison of the strength of
trophic cascades. Ecol Lett 5: 785−791
Smetacek V (2008) Are declining antarctic krill stocks a
result of global warming or of the decimation of the
whales? In: Duarte CM (ed) Impacts of global warming
on polar ecosystems. Fundación BBVA Press, Madrid,
p 45−83
Sommer U, Lengfellner K (2008) Climate change and the
timing, magnitude, and composition of the phytoplank-
ton spring bloom. Glob Change Biol 14: 1199−1208
Sommer U, Sommer F (2006) Cladocerans versus copepods:
the cause of contrasting top-down controls on freshwater
and marine phytoplankton. Oecologia 147: 183−194
Sommer U, Stibor H (2002) Copepoda−Cladocera−Tunicata:
the role of three major mesozooplankton groups in pela -
gic food webs. Ecol Res 17: 161−174
Sommer U, Aberle N, Engel A, Hansen T and others (2007)
An indoor mesocosm system to study the effect of climate
change on the late winter and spring succession of Baltic
Sea phyto- and zooplankton. Oecologia 150: 655−667
Sommer U, Aberle N, Lengfellner K, Lewandowska A (2012)
The Baltic Sea spring phytoplankton bloom in a chang-
ing climate: an experimental approach. Mar Biol 159:
2479−2490
Steinacher M, Joos F, Frolicher TL, Bopp L and others (2010)
Projected 21st century decrease in marine productivity: a
multi-model analysis. Biogeosciences 7: 979−1005
Sugimoto T, Tadokoro K (1997) Interannual and inter-
decadal variations in zooplankton biomass, chlorophyll
concentration and physical environment in the subarctic
Pacific and Bering Sea. Fish Oceanogr 6: 74−93
Suikkanen S, Laamanen M, Huttunen M (2007) Long-term
changes in summer phytoplankton communities of the
open northern Baltic Sea. Estuar Coast Shelf Sci 71:
580−592
Suttle CA (1994) The significance of viruses to mortality in
aquatic microbial communities. Microb Ecol 28: 237−243
Suttle CA (2007) Marine viruses — major players in the
global ecosystem. Nat Rev Microbiol 5: 801−812
Sverdrup HU (1953) On conditions for the vernal blooming
of phytoplankton. J Cons Int Explor Mer 18: 287−295
Taucher J, Oschlies A (2011) Can we predict the direction of
marine primary production change under global warm-
ing? Geophys Res Lett 38: L02603
Thomas MK, Kremer CT, Klausmeier CA, Litchman E (2012)
A global pattern of thermal adaptation in marine phyto-
plankton. Science 338: 1085−1088
Tilman D, Wedin D, Knops J (1996) Productivity and sus -
tainability influenced by biodiversity in grassland eco -
systems. Nature 379: 718−721
Vecchi GA, Soden BJ, Wittenberg AT, Held IM, Leetmaa A,
Harrison MJ (2006) Weakening of tropical Pacific atmo -
spheric circulation due to anthropogenic forcing. Nature
441: 73−76
Venrick EL, McGowan JA, Cayan DR, Hayward TL (1987)
Climate and chlorophyll a: long-term trends in the cen-
tral north Pacific Ocean. Science 238: 70−72
Verity PG, Smetacek V (1996) Organism life cycles, preda-
tion, and the structure of marine pelagic ecosystems. Mar
Ecol Prog Ser 130: 277−293
271
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤

Mar Ecol Prog Ser 534: 251–272, 2015
Vermeij GJ (2011) Shifting sources of productivity in the
coastal marine tropics during the Cenozoic era. Proc R
Soc B 278: 2362−2368
Wang M, Overland JE, Bond NA (2010) Climate projections
for selected large marine ecosystems. J Mar Syst 79:
258−266
Ware DM, Thomson RE (2005) Bottom-up ecosystem trophic
dynamics determine fish production in the northeast
Pacific. Science 308: 1280−1284
Wernand MR, van der Woerd HJ, Gieskes WW (2013)
Trends in ocean colour and chlorophyll concentration
from 1889 to 2000, worldwide. PLoS ONE 8: e63766
Wirtz KW (2012) Who is eating whom? Morphology and
feeding type determine the size relation between plank-
tonic predators and their ideal prey. Mar Ecol Prog Ser
445: 1−12
Wirtz KW (2013) Mechanistic origins of variability in phyto-
plankton dynamics: Part I: niche formation revealed by a
size-based model. Mar Biol 160: 2319−2335
Worm B, Barbier EB, Beaumont N, Duffy JE and others
(2006) Impacts of biodiversity loss on ocean ecosystem
services. Science 314: 787−790
272
Editorial responsibility: Katherine Richardson,
Copenhagen, Denmark
Submitted: June 16, 2014; Accepted: July 6, 2015
Proofs received from author(s): August 18, 2015
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤