Respiration rates.
On average,
Trichodesmium colonies have lower carbon and nitrogen fixation rates than free trichomes (
12). With all else being equal, this implies that
Trichodesmium colonies are always outcompeted by free trichomes.
Figure 1a highlights this possibility for a range of fixed respiration rates. Respiration rates of >0.009 mol C h
−1 mol C
−1 lead to a net loss in carbon storage for colonies over a period of 30 days, and similarly, rates of >00.0135 mol C h
−1 mol C
−1 lead to a net loss of carbon for free trichomes. This means that colonies hit net negative in total carbon storage before free trichomes. In such cases where respiration rates are invariant,
Trichodesmium colonies can only outcompete free trichomes in some circumstances: respiration rates of colonies must be significantly lower than that of free trichomes.
As respiration rates in phytoplankton can vary with time and space (
23,
24), both
Trichodesmium colonies and free trichomes can have different respiration rates driven by physiological or ecological factors. The respiration rate is likely to be affected by the level of oxygen management because respiratory protection (extra respiration to decrease intracellular oxygen) has been predicted to account for a large part of carbon loss (
19,
25). Regardless of the contributing factor, we attempted to quantify the decrease needed for colonies to compete with free trichomes. In the absence of other variables that might influence carbon fluxes, respiration rates in colonies must be a fraction of the rates in free trichomes. Interestingly, higher possible respiration rates for
Trichodesmium increase the maximum limit of colonies to metabolically compete with free trichomes. Our model, although coarse in resolution, suggests that
Trichodesmium colonies have better opportunities under conditions that maximize carbon loss, regardless of the mechanism involved.
We also chose to test the scenario where metabolic costs were relative to the carbon fixation rate of both colonies and free trichomes. In this scenario, both
Trichodesmium colonies and free trichomes can modulate their respiration rates based on the available carbon that is fixed. Colonies performed better than free trichomes for respiration rates that are greater than 88.8% of carbon fixation rates (
Fig. 2). Colony formation might be a successful strategy when environmental or ecological circumstances lead to high metabolic costs. In marine phytoplankton, respiration becomes an increasingly higher proportion of carbon fixation as temperatures are increased (
26). When respiration costs change proportionately with carbon fixation rates, colony formation in
Trichodesmium might be a mechanism of reducing net carbon loss over longer periods of time. This would be because lower carbon fixation rates also lead to lower respiration rates, which minimize the loss of stored carbon (
Fig. 2). Our results suggest that colony formation might have evolved in environments that are prone to periods of high metabolic stress. In such situations, lowering carbon fixation rates, and by association, respiration rates, would allow
Trichodesmium colonies to persist for longer than free trichomes.
External drivers.
Colony formation in marine phytoplankton can be subject to various environmental and ecological drivers, such as temperature constraints, turbulence, predator cues, or nutrient availability (
27–29). Colony adaptations can often be species-specific, with additive effects when phytoplankton are exposed to more than one environmental change (
30). We added a parameter to our model to test the influence of external drivers on colony performance in
Trichodesmium. By incorporating a general “interaction” term, our goal was to quantify the rates at which an increase or decrease in net carbon and nitrogen fixation could potentially provide advantages in forming colonies over free trichomes.
In general, we found that a large negative effect on carbon and nitrogen fixation tends to favor colonies (
Fig. 3). Negative influence on carbon and nitrogen fixation rates can stand for environmental stress or micronutrient limitations. Similar to the case with varying respiration rates, the lower starting carbon and nitrogen fixation rates of colonies allow for colonies to outperform free trichomes by the end of model simulations. Cultures of
Trichodesmium typically produce colonies after the exponential growth phase and under nutrient limitations (
31,
32). By lowering their investment in carbon and nitrogen fixation rates, colonies could be employing a trade-off that increases their survival under some conditions at the cost of slower growth.
Conversely, conditions that would lead to an increase in carbon and nitrogen fixation rates appear to favor free trichomes (
Fig. 3;
ip > 0). A positive interaction strength could stand for the ready availability of nutrients or optimal light and temperature conditions. Such conditions could allow for quicker increases in biomass for free trichomes relative to colonies. Interestingly, the relative advantage of free trichomes is significantly reduced in simulations where respiration rates are high relative to the carbon fixation rates (
Fig. 3d). In this scenario, the increase in carbon fixation rate is muted by a similar, proportionate increase in metabolic costs. This means that if free trichomes were to dominate in an environment, any increases in carbon and nitrogen fixation would have to happen with minimal metabolic costs of enacting those increases. For example, an increase in nitrogen fixation rates due to ocean warming (
33) might not alter the relative success of free trichomes, as thermal compensation would also increase the metabolic costs of the cells (
26). On the other hand, elevated CO
2 concentrations have also been shown to increase nitrogen fixation rates in
Trichodesmium (
34), and it is possible that such increases benefit the formation of free trichomes over colonies. Earth system models would need to resolve potential changes in
Trichodesmium trichome and colony abundance (and, by association, bulk C fixation and N fixation) to reduce uncertainty and improve predictions (
35–37).
Trichodesmium has an associated epibiont of microbes that significantly extends its metabolic potential (
38). Some studies have shown that
Trichodesmium colonies can interact with bacteria to preferentially acquire iron and phosphorus from dust particles (
39,
40). As there exists significant diversity in the
Trichodesmium epibiont (
41), there might be cases where different epibiont communities in colonies and free trichomes lead to mixed responses to the same environmental change.
In model simulations in which external influence directly affects the total carbon stored (
it) and not the metabolic rates, our models still showed free trichomes outperforming colonies in all cases in which respiration rates are comparatively low (
Fig. 4). For high respiration rates (90% of C fixation), colonies outperformed free trichomes in all cases, regardless of the direction of influence. As environmental influence in the natural environment can be a chaotic and dynamic process, there probably exist environmental drivers that affect both metabolic rates and total carbon. Our results suggest that even under cases of high carbon loss due to mortality, respiration rates are a larger factor contributing to the success of colony formation in
Trichodesmium.
Colony formation in marine phytoplankton provides benefit against grazers (
42), and it is possible that
Trichodesmium colony formation can reduce the rates of predation by particular classes of zooplankton, possibly due to biotoxin accumulation (
43,
44). The reduction of grazing could boost colony success against free trichomes, but such a reduction should compensate for the lowered C fixation and N fixation rates. Future experimental studies could test this hypothesis by measuring the growth rates of
Trichodesmium trichomes and colonies in the presence and absence of predator cues.
Relative metabolic rates.
Phytoplankton can change their rates of carbon and nitrogen fixation, and global change stressors, such as changes in temperature or nutrient availability, are expected to alter these rates (
45).
Trichodesmium has been shown to alter its metabolic rates in response to various stressors: high light (
46), elevated carbon dioxide concentrations (
34), and anthropogenic nutrient pollution (
47). C and N fixation rates in
Trichodesmium can also vary over the course of a single day (
48). We tested the effects of different C and N fixation ratios (τ) on the competitiveness of colonies and free trichomes to broadly evaluate their performance in a dynamic environment.
Colony formation reduces the ability of the aggregate to fix nitrogen more than it reduces their ability to fix carbon (
12). This means that colonies have higher τ ratios than free trichomes, and a further increase would continue to benefit colonies in their competitiveness against free trichomes. In contrast, circumstances that increase the τ of free trichomes exponentially decrease colony success (
Fig. 5, moving right on the
x axis). Interestingly, if the τ of free trichomes reaches 1.5 times the τ of colonies, colonies get outcompeted, given that everything else is unchanged. However, other physiological, ecological, or environmental benefits of colony formation could be oriented toward escaping competitive exclusion from free trichomes.
Conclusions.
Colony formation in marine phytoplankton can be subject to various ecological and environmental pressures. Because Trichodesmium can persist as both colonies and free trichomes, identifying the trade-offs of colony formation is imperative in understanding the ecological role of Trichodesmium in a changing world. We simulated multiple scenarios of metabolic demand, external influence, and relative carbon and nitrogen fixation rates to compare the performance of colonies and free trichomes in a model system.
In the case where respiration rates are equal and unchanging, colonies always get outcompeted by free trichomes. When respiration rates are unequal and variable, colonies should necessarily have only a fraction of the metabolic costs of free trichomes to stand a chance of competing. If respiration rates are proportional to carbon fixation rates, colonies perform better and outcompete free trichomes when costs exceed 88.8% of the carbon fixed in the cell. Negative external influence on metabolic rates, such as micronutrient limitations, allows colonies to perform better than free trichomes, whereas positive external influence, such as mutualistic interactions or optimal environmental conditions, would benefit free trichomes more than colonies. For both colonies and free trichomes, increasing τ would increase their competitiveness in the environment; however, if the τ of free trichomes is greater than 1.5 times the τ of colonies, there are no scenarios where colonies can compete with free trichomes.
Despite lower carbon and nitrogen fixation rates, Trichodesmium colonies can outperform free trichomes under many circumstances. Our results lay out a framework and many testable hypotheses that could motivate further study into the ecological and environmental drivers of colony formation in marine phytoplankton. As global environmental change is expected to alter the marine environment, changes in the ability of phytoplankton taxa to form and subsist as colonies could have globally significant ecological and biogeochemical effects.