Technical Comments

Comment on “Patterns of Diversity in Marine Phytoplankton”

See allHide authors and affiliations

Science  30 Jul 2010:
Vol. 329, Issue 5991, pp. 512
DOI: 10.1126/science.1189880


Barton et al. (Reports, 19 March 2010, p. 1509) argued that stable conditions enable neutral coexistence of many phytoplankton species in the tropical oceans, whereas seasonal variation causes low biodiversity in subpolar oceans. However, their model prediction is not robust. A minor deviation from the neutrality assumption favors coexistence in fluctuating rather than stable environments.

Terrestrial ecosystems show a spectacular increase in plant biodiversity from the poles to the tropics. Phytoplankton play a similar role in the oceans as grass and trees in the terrestrial world, converting sunlight and mineral nutrition to biomass production at the basis of the food web. However, whether marine phytoplankton show a similar latitudinal biodiversity gradient as their terrestrial counterparts is largely unknown. Barton et al. (1) developed a large-scale ocean model to investigate phytoplankton diversity across the global ocean. The model links phytoplankton biodiversity with ocean circulation patterns and biogeochemical cycles and is an exciting novel attempt to predict marine biodiversity. The model predicts an increase in phytoplankton diversity from the poles to the tropics, similar to the latitudinal diversity gradients in terrestrial ecosystems. In addition, the model predicts diversity hotspots in areas with highly energetic ocean circulation patterns. What explains these predicted biodiversity patterns?

During the past years, theory and experiments have revealed several subtle forms of niche differentiation promoting plankton diversity, including differential use of nutrients (2, 3), partitioning of the light spectrum (4), predator-mediated coexistence (3), and nonequilibrium dynamics generated by species interactions (5, 6). However, Barton et al. (1) consider the special case that all phytoplankton species are equivalent competitors, as in the neutral theory of biodiversity (7). The competitive ability of a species can be summarized by its R* value, indicating the lowest environmental nutrient concentration at which growth and mortality are in balance. Resource competition theory predicts that, in a constant environment, the species with the lowest R* value will be the superior competitor (2, 3). Hence, species with identical R* values will be equivalent competitors in a constant environment.

For instance, in Fig. 1A, two species have different maximum growth rates and half-saturation constants, yet they have identical R* values of 0.25 mmol m−3. In this case, simulations with a resource competition model (see Supporting Online Material) predict that stable environments enable neutral coexistence of phytoplankton species for thousands of years (Fig. 1C). In fluctuating environments, however, the model predicts that the species with the highest maximum specific growth rate (species 1) competitively excludes its slower-growing competitors. Competitive exclusion occurs more rapidly when the period and amplitude of the nutrient fluctuations increase (Fig. 1, C and E). Based on these model predictions, Barton et al. argue that high diversity in tropical and subtropical oceans is due to prolonged coexistence of equivalent competitors in stable environments, whereas low diversity in polar regions is caused by rapid competitive exclusion in fluctuating environments (1).

Fig. 1

A slight deviation from neutrality yields very different predictions on the time scale of competitive exclusion. In panels on the left, the two species are equivalent competitors with exactly the same R* values (0.25 mmol m−3), as in Barton et al. (1). In panels on the right, the R* value of species 2 (0.23 mmol m−3) is slightly lower than the R* value of species 1 (0.25 mmol m−3). (A and B) Specific growth rates (solid lines) of the two species as a function of the external nutrient concentration. The dashed horizontal line shows the mortality rate of the two species. (C to F) The time scale of competitive exclusion as a function of period [(C) and (D)] and amplitude [(E) and (F)] of the nutrient fluctuations. Model equations and parameter values are given in the Supporting Online Material.

However, these model predictions are not robust. Figure 1B uses the same parameter values as Fig. 1A, except for the half-saturation constant of species 2, which is reduced from 0.15 to 0.14 mmol m−3. This yields a slightly lower R* value for species 2, which thus becomes a slightly stronger competitor in a constant environment than species 1. The change in parameter values is so small that one can barely see the difference (compare Fig. 1A and Fig. 1B). Yet this minor deviation from the neutrality assumption leads to completely different results. Competitive exclusion is now relatively fast in stable environments and in seasonal environments with low-frequency but high-amplitude fluctuations (Fig. 1, D and F). However, competitive exclusion is much slower in environments fluctuating at intermediate periodicity (Fig. 1D). Fluctuations at intermediate amplitude even allow species coexistence (Fig. 1F). These model predictions differ from Barton et al. (1) but are in line with the intermediate-disturbance hypothesis (8), which states that biodiversity will be highest at intermediate frequencies and intensities of environmental fluctuations. Furthermore, contrary to Barton et al., species 2 is now the superior competitor over a large region of parameter space despite its lower maximum growth rate than species 1 (Fig. 1, D and F). Thus, slight deviations from neutrality are sufficient to generate fundamentally different model predictions.

Interestingly, the underlying mechanisms for species coexistence are different as well. In Fig. 1E, species are equivalent competitors in a constant environment, enabling neutral coexistence of the species [by an equalizing mechanism (9)]. In contrast, in Fig. 1F, species coexist in a fluctuating environment because they respond differently to environmental variation [stabilizing mechanism (9)].

Laboratory competition experiments provide ideal systems to test these contrasting model predictions. Competition experiments find that most phytoplankton species are not equivalent competitors. Differences in the R* values of phytoplankton species generally result in rapid competitive exclusion when species compete for a single limiting resource in a constant environment (1012). I know of only two experimental examples of competitive equivalence among phytoplankton species [Asterionella and Fragilaria (10) and Chlorella and Synechocystis (12)], and even in these two cases it seems that the species did not have exactly the same R* values (as in Fig. 1A) but just very similar R* values (as in Fig. 1B). Moreover, competition experiments have also shown that environmental fluctuations allow coexistence of phytoplankton species that would not coexist in a constant environment (13, 14). Highest plankton diversity is often found at intermediate frequencies of disturbance (15), in accordance with the intermediate-disturbance hypothesis (8).

In conclusion, the explanation for latitudinal diversity gradients offered by Barton et al. (1) is intriguing but is mathematically not robust and inconsistent with the available experimental data. Temporal variation may enhance rather than reduce biodiversity, especially if fluctuations occur at intermediate frequency. A better understanding of the mechanisms underlying marine diversity patterns will require further refinement of models and experiments, in combination with large-scale investigations of marine biodiversity across the global ocean.

Supporting Online Material

SOM Text


View Abstract

Navigate This Article