Fig. 2: Scatter plots of the results of simulated data with and without alternative equilibria. | Nature Communications

Fig. 2: Scatter plots of the results of simulated data with and without alternative equilibria.

From: Bimodality and alternative equilibria do not help explain long-term patterns in shallow lake chlorophyll-a

Fig. 2

There is an increasing temporal scale: from single-year observations to 3-year means and then 5-year means for the simulated data in the two panels on the left. Scatter plots of the TN and TP relationship with chlorophyll-a from real-world data with the different temporal perspectives of single-year data, 3-year means and 5-year means. The data were sampled 1000 times iteratively (see methods) for each given time period and a model was derived from each of the selections where the GLM converged (see results for numbers), providing a distribution of models and residuals for each temporal perspective. The shading of each point in the plots reflects the number of times it was selected in the random sampling procedure. The results show that as the temporal perspective lengthens the relationship between nutrients and chlorophyll-a becomes more predictable and that the models based on the multiple-year mean data always explain more variation than their single-year counterpart.

Back to article page