Seasonal variation in the balance and strength of cooperative and competitive behavior in patches of blue mussels

Abstract

Aggregation into groups may affect performance of individuals through the balance and strength of facilitative versus competitive interactions. We studied in situ how seasonal variation in abiotic environment affects this balance for blue mussels, a semi-sessile species. We hypothesize that seasonal variation in stresses and resources affects the strength of the interaction. We expected that, in benign conditions (here: high food availability, medium temperatures, low hydrodynamic stress), performance is dominated by growth and is better at low densities, while at adverse conditions (here: low food availability, low or high temperatures, high hydrodynamic stress), performance is dominated by survival and higher at high densities. Mussels were kept in shallow subtidal exclosures at 10 different densities for a one-month period. This exact procedure was repeated seven times at the same location within a one-year period. We measured development in mussel patch shape, performance, and environmental parameters. Environmental conditions for mussels were most benign in summer and most adverse in winter. Patches developed into less complex shapes at lower densities, but also after stronger hydrodynamic disturbances. Towards summer, mussels became more active, aggregation behavior increased, and interactions became more pronounced. Towards winter, mussels became less active: aggregation behavior and growth rates declined and at the lowest temperatures survival started to decrease with mussel density. Survival and growth (by proxy of mussel condition) were both density-dependent; however, contrary to our expectations we found positive interactions between density and survival at the most benign conditions in summer and negative interactions at the most adverse conditions in winter. In between the two seasons, the strength of the interactions increased towards summer and decreased towards winter following a bell-shaped pattern. This pattern might be explained by the environmental mediated aggregation behavior of the mussels. The obvious seasonal pattern in balance and strength of density-dependent interactions demonstrates that strength and direction of intra-specific interactions are both strongly affected by environmental context.

Fig 1. Conceptual drawing of hypotheses and methodology.

A. Hypothetical interactions in subtidal mussel patches; in summer interactions are expected to be dominated by growth due to high level of resources and temperature (benign conditions). In winter interactions are expected to be dominated by survival or mortality, because of the high level of environmental stress, low resources and high turbidity (adverse conditions). Note that temperature is not considered a stressor under the subtidal conditions at the experimental area. We anticipate that in the balance between competition and facilitation, competition will play a more significant role when growth is dominant, while facilitation will be more important when survival is dominant and B. Method to map density dependent interactions; to map density-dependent interactions, mussels will be placed in cages at 10 different densities (in triplicate) from low to high and performance was monitored over approximately one month, at seven different time periods within a year (M1-7, indicated by the grey bars in A.). Photos illustrate this development (at the highest density of 19 kg/m²) and the binary image shows the mussel cover used to calculate the perimeter-to-area ratio.

Fig 2. Environmental parameters measured at the experimental location, with daily averages calculated from 10-minute intervals.

A. Water temperature (°C), B. Chlorophyll A (μg/l) and C. turbidity (FTU). Colors indicate the environmental parameters during the consecutive experimental runs, with in 2017: 1 = Mar 10^th–Apr 9^th, 2 = Apr 18^th–May 15^th, 3 = Jun 1^st-27^th, 4 = Jul 27^th–Aug 28^th, 5 = Sep 20^th–Oct 19^th, 6 = Nov 9^th–Dec 11^th and in 2018: 7 = Feb 21^st–Mar 19^th.

Fig 3. Slope coefficient (M_survival± SE, n = 30, a proxy for the strength of a density-dependent effect) of the log-log relation between end density and starting density for each consecutive experimental run: ln[D₁] = M_survival • ln[N₀]+b.

Each run consisted of 10 different densities in triplicate. When the slope is below 1, a negative interaction between the individuals is expected (indication of competition), because the chance of survival decreases with mussel density. When the slope is above 1, positive interactions are expected (indication of cooperation) because the chance of survival increases with mussel density.

Fig 4. Marginal and interaction effects of the temperature, chlorophyll-a, and turbidity on the odds of survival.

The color indicates the relative effect of the interaction terms on the linear predictions: more green means a higher survival chance (higher log-odds), more red means a lower survival chance (lower log-odds). More white means the relative effect is closer to 0 (neither increase nor decrease in log-odds).

Fig 5. Perimeter: Area ratio ± SE (n = 21) of the mussel patches at the end of the experimental run as a function starting density.

Fig 6. Slope coefficient M_CI ± SE (n = 30) between the log-log relation between mussel condition index (mg/cm³) and starting biomass for the different consecutive runs, as ln[CI] = M_CI • ln[D₀]+b., each run consisted of 10 different densities (ns = not significant, *p<0.05, **p<0.01, ***p<0.001).

Run 2 is not included due to a lack of data. At negative M_CI values mussel conditions decreases with mussel density (negative interactions) indicating competition for food, at positive M_CI values mussel condition increases with mussel density (positive interactions—not observed).