Science - USA (2021-10-29)

are thought to be important determinants of
plankton community composition ( 25 ), we
also quantified local biotic conditions by using
the relative read counts of major eukaryotic
groups (excluding the focal group; see materials
and methods). Biotic conditions, like abiotic
ones, partly have a latitudinal structure, and
we refer here to them collectively as“envi-
ronmental conditions”(figs. S20 and S21). The
resulting environmental maps can be inter-
preted as the hypothetical geographic patterns
that are expected for organisms with a strong
linear coupling to local environmental condi-
tions. Therefore, the relative match of bio-
geography to environmental maps versus
Moran maps indicates whether local environ-
mental factors are able to linearly explain spatial
variations in community composition or

whether community composition results from

a more complex combination of the physical

and ecological processes that occur alongside

transport: mixing, stirring, ecological drift

(i.e., stochastic variations in abundances), and

plankton interactions and responses to var-

iations in temperature and nutrient supplies

along currents.

We found that the total variance in surface

community composition that can be statis-

tically explained by Moran maps and local en-

vironmental conditions (abiotic and biotic)

averages 27% across major eukaryotic groups

(from 0% for Porifera to 58% for Ciliophora;

see materials and methods). Expectedly, this

total explained variance is tightly correlated

to the strength of biogeographic structure (i.e.,

group position on first axis;rP= 0.88; Fig. 4A).

The variance that is statistically explained by

Moran maps averages 21% across groups at

the surface (fig. S22A) and is primarily con-

tributed by maps corresponding to large-scale

patterns between basins (figs. S19 and S23).

This supports a key role of transport by the

main current systems in shaping eukaryotic

plankton biogeography, in particular by ex-

tending the distribution of taxa within basins

( 9 , 13 , 26 , 27 ) and constraining long-distance

dispersal between basins ( 22 ). As expected

from our previous results, the ratio of the frac-

tions of variance that are statistically explained

by Moran maps and local environmental con-

ditions increases with group position on the

second axis of variation (rS= 0.51,P= 10−^5 ; Fig.

4B) and with group mean body size (rS= 0.46,

P= 10−^4 ; Fig. 4C) and depends on ecological

function (ANOVAFtest:P= 0.009; Fig.

4D). Local environmental factors can thus

explain spatial variations in community

composition better in phototrophs than in

metazoans, the community composition of

which appears to result from a more complex

combination of the processes that occur along-

side transport. Last, within the variance statis-

tically explained by the local environment, an

approximately equal share can be attributed

to biotic and abiotic conditions for most groups

(fig. S22B), irrespective of their mean body

size, ecological function, diversity, or biogeog-

raphy (fig. S24). The results are similar at the

DCM but are far less pronounced (fig. S25).

Although we cannot exclude the possibility

that local biotic conditions reflect the indirect

effect of local abiotic factors that are not ac-

counted for in our study—such as fluxes of

nutrients, which are often more relevant to

planktonic organisms than instantaneous nu-

trient concentrations ( 27 )—these results indicate

an additional role for interspecific interactions

in shaping community composition ( 25 , 28 ).

In the early days of the exploration of micro-

eukaryote biogeography, the dominant view

was that all eukaryotes up to a size of about

1 mm were globally dispersed and primarily

constrained by abiotic conditions ( 11 ). That

plankton of all sizes can be dispersal-limited

is now well recognized, but the relationship

between body size and biogeography remains

unclear ( 12 , 13 , 29 ). It has been proposed that

larger organisms tend to be more dispersal

limited because of their lower population sizes

and reduced ability to form stress-resistant

propagules ( 29 ). However, compositional turn-

over along currents is slower, rather than faster,

with increasing body size ( 13 ). Our results re-

concile these views; at the global scale, organisms

of larger sizes indeed have more-restricted

ranges, whereas they have wider spatial dis-

tributions at the regional scale that are pre-

sumably linked to their specific ecologies, longer

life spans, and reduced sensitivity to local en-

vironmental variations. Body size influences

598 29 OCTOBER 2021•VOL 374 ISSUE 6567 science.orgSCIENCE

P=0.88

0.1 P=2e–23

0.2

0.3

0.4

0.5

0.6

–0.2 –0.1 0.0 0.1 0.2

Biogeographic axis 1

Total variance explained

by Moran maps and environment S

=0.51

0.5 P=1e–05

1.0

3.0

–0.2 –0.1 0.0 0.1 0.2

Biogeographic axis 2

Ratio of variance explained by

Moran maps over environment (log scale)

S=0.46

0.5 P=1e–04

1.0

3.0

30 100 300

Body size (μm), log scale

Ratio of variance explained by

Moran maps over environment (log scale)

ANOVA:

0.5 P=0.009

1.0

3.0

PhototrophsPhagotrophsMeta

zoansParasites

B

C D

RAD-C

RAD-C

RAD-C

A

Fig. 4. Drivers of surface biogeography across major eukaryotic plankton groups.(A) The total
variance in surface biogeography that can be explained by the combination of Moran maps and (abiotic
and biotic) local environmental conditions tightly correlates with group position on the first axis of
biogeographic variation (rP, two-sidedttest) (see Fig. 2A). (BtoD) The ratio of the variance that is explained
by Moran maps over the variance explained by local environmental conditions (B) increases with group
position on the second axis (rS, two-sidedttest), (C) increases with group mean body size, and (D) varies
across broad ecological categories (ANOVA,Ftest). The ratio is greater than 1 (dashed line) for most groups,
which supports an overall stronger influence of processes occurring alongside transport than of local
environmental conditions on plankton biogeography at the surface. The gray dot represents radiolarian group
C (RAD-C), an outlier group that we excluded from statistical tests. We did not find any significant
explanatory variable for Porifera and therefore excluded this group from these analyses.

RESEARCH | REPORTS