example, bacterial chemotaxis to microscale DOM
gradients has been predicted to increase remineralization rates twofold (42). For phytoplankton, modeling predicts that productivity is several
times larger in heterogeneous than in homogeneous conditions (43). Similarly, observations of
phytoplankton growth in the face of nondetect-able levels of limiting nutrients have been attributed to intermittent nutrient pulses (44).
To identify potential effects of microscale
gradients on biogeochemical fluxes, it is instructive to consider how a small DOM patch affects
bacteria, for nonmotile bacteria and for chemotactic bacteria (Fig. 2A). Only a minuscule fraction of the bacteria will initially find themselves
by chance inside the patch, yet typically, most are
within 100 to 1000 mm of the next patch (45). As
the patch diffuses (Fig. 2, A and B), nonmotile
cells remain randomly distributed, whereas many
motile cells cluster inside the patch within tens of
seconds (5, 10, 24). Does this behavior affect the
total amount of DOM transformed into bacterial
biomass? We need to consider that the entire
DOM from the patch, which diffuses to uniformity within minutes, might ultimately be consumed
in both the motile and the nonmotile scenarios, in
which case DOM consumption may simply be
accelerated if perpetrated by motile bacteria (Fig.
2C). In other words, do responses to microscale
gradients purely change the time scale over which
DOM is remineralized or also the total amount of
DOM that is remineralized?
There are several mechanisms by which differential gradient utilization may affect total
amounts, not only time scales. A first mechanism
relates to the bacterial growth efficiency (BGE),
the fraction of carbon taken up that cells incorporate as biomass (the remainder, they respire).
BGE increases with growth rate and with
resource concentration when measured across
different marine provinces (6). Given the higher
concentrations within patches and the higher
maximum growth rates of copiotrophs, might
copiotrophs have larger BGEs than oligotrophs?
If metabolic studies were to verify this hypothesis, then DOM uptake by copiotrophs would
channel more carbon into the microbial loop
than uptake by oligotrophs (Fig. 2D).
A second mechanism concerns the feedback
between primary production and remineraliza-
tion. By clustering near phytoplankton (4), motile
bacteria may not only accelerate remineralization
of algal DOM but also enhance the productivity
of phytoplankton by supplying them with inor-
ganic nutrients. Evidence for the pervasiveness of
these associations has come from atomic force
microscopy measurements, which recently re-
vealed that heterotrophic bacteria and cyanobac-
teria are “conjoint” on average 30% of the time
(46). Calculations predict that motile bacteria have
orders-of-magnitude higher chances of ending up
conjoint than nonmotile bacteria and that this
propinquity markedly increases solute fluxes be-
tween heterotrophs and autotrophs. A further ex-
ample is the remineralization of organic matter
on sinking marine snow particles, which favors
the retention of limiting elements in the upper
water column and thereby stimulates primary pro-
duction and the formation of more marine snow.
Motility can enhance this process by increasing
particle colonization rates up to fivefold (9).
Fig. 3. Optimal foraging. The nutrient concentra-
tion encountered by marine bacteria varies consid-
erably over time scales as short as seconds to minutes,
owing to pervasive chemical and physical gradients
in their immediate environment. For motile bacte-
ria, which actively exploit nutrient gradients, this
variability is greater than for nonmotile bacteria.
Optimal foraging frameworks—where utilization of
nutrient patches is weighted against the cost of mo-
tility but also, e.g., against the increased risks of
predation and viral infection—promise to help de-
termine the dominant foraging strategies of marine
bacteria as a function of the environmental condi-
tions. These frameworks will require new informa-
tion on bacterial metabolism, including, for example,
the dependence of uptake kinetics and BGE on nu-
trient concentration, to determine to what extent
the behavioral responses of bacteria to microscale
gradients affect ocean ecosystem-level properties.
termined whether this degradation affects DOM
bioavailability on the time scale (~minutes) of
the consumption lag.
Behavioral responses to microenvironments
can also have indirect effects on biogeochemistry.
The attachment of heterotrophic bacteria to diatoms can favor diatom aggregation by stimulating
the production of sticky extracellular polymers
(48). Aggregation accelerates sinking and, thus,
the efficiency of the biological pump in transporting carbon from the surface ocean to depth.
Bacterial attachment to diatoms, in turn, could be
strongly favored by algal exudate gradients and
Outlook: Shrinking Our Fields of View While
Expanding Our Ecological Frameworks
Advances in microbial oceanography have been
repeatedly triggered by new tools, from the fluorescent staining of cells to flow cytometry to
metagenomics. As we begin to appreciate how
heterogeneous and diverse the world of marine
microbes is, there is now scope for techniques
that probe this world at the scale of single cells
and microenvironments. Bulk sampling techniques, where liters of water are collected and
homogenized, provide valuable information on
the mean microbial environment but cannot capture the local conditions experienced by microbes.
To do so, we must shrink our operational field
of view. Exciting opportunities are in sight on a
number of fronts: Genomics is reaching single-cell resolution (49), secondary ion mass spectrometry (nanoSIMS) is revealing the chemical signature
of individual cells (50), atomic force microscopy
is shedding light on the spatial organization of
marine microbes (46), and microfluidic technology is unveiling microbial behavior within realistic microenvironments (10, 24, 25, 51). Yet, it
remains difficult to interrogate microenvironments
in situ, owing to their small volumes and intermittent nature: There is “plenty of room at the
bottom” for measurements of microbial behavior
and the microscale chemical concentration gradients that shape it.
Tools, however, are not the sole limiting factor in our understanding of microbial ecology in
the context of a heterogeneous microlandscape.
We also lack quantitative ecological frameworks
to rationalize and scale up microenvironmental
processes. Unraveling the relation between gradients and motility; between patchiness and diversity; and between behavior, uptake kinetics, and
biogeochemical fluxes calls for theoretical ecologists to dive into microbial oceanography. Microbes’ fast generation times, vast numbers, disparate
interactions, and rich spatial organization make
microbial oceanography an intriguing, yet under-appreciated, model system for testing ecological
theory. Glimpses of this trend can be seen in
microbial biogeography, where predictions for
taxa-area relations and longitudinal gradients in
species abundance have been recently tested on
marine microorganisms (52).
In contrast, little ecological theory has been
applied at the scale of microbial microenvironments. Fitness-based models can provide unifying frameworks to evaluate the role of specific
adaptations, such as high swimming speeds, hybrid locomotion, and metabolic plasticity. For
example, the bacterial nutrient quest in a sea of
microscale patches is a quintessential optimal foraging problem (Fig. 3). Optimal foraging theory
predicts the movement behavior that maximizes
the fitness of an organism whose resources are
heterogeneous (53). Motile marine bacteria live in
a dynamic equilibrium between disparate micro-