When life is not sweet for biodiversity

Dr Ivana Gudelj and Dr Rob Beardmore (Department of Mathematics) along with colleagues from University of California, Santa Cruz (UCSC) and University of Bath have recently published a paper in Nature

Dr Ivana Gudelj and Dr Rob Beardmore (Department of Mathematics) along with colleagues from University of California, Santa Cruz (UCSC) and University of Bath have recently published a paper in Nature which increases our understanding of the relationship between nutrients and biodiversity in experimental and real life ecosystems. The results show that nutrient availability has an impact on biodiversity, but that the precise relationship depends on the genetic detail of the species involved. 

Studying evolutionary and ecological processes in wild ecosystems presents obvious difficulties and often in vitro laboratory based models are used as an alternative. However, the applicability of these systems into real world situations can be limited. This research outlines the limits to generalising experimental evolutionary systems using mathematical models.

The study determines how host-parasite co-evolution might affect diversity at different nutrient levels. To define this relationship the group combined a test tube in vitro system with a series of mathematical models to determine how these models can be applied to real life situations.

The mathematical models are based on experimental ecosystems consisting of Escherichia coli bacteria and the lytic bacterial virus or ‘bacteriophage’ T7. In the lab of Professor Samantha Forde at UCSC, mini experimental ecosystems were set up, consisting of cultures of E.coli and T7 grown in solutions with varying levels of glucose for 17 days, equivalent to 150 generations. Following incubation, the density of bacteria and variation of bacteria and bacteriophage was measured. Results showed that when levels of nutrients were high there were low levels of diversity. In this situation infected bacteria grew densely, producing many virus particles and less resistant strains were wiped out. In systems when nutrient levels were low, diversity increased as there were less viruses present and more strains of bacteria could survive.

The results of these experiments were interpreted into a mathematic model, using the genetic detail of each organism. This model was then used to determine how diversity as a function of resource input varies for alternative co-evolving partners. In this case the T7 phage was substituted with the genetic details of another phage, the bacteriophage lambda. In contrast to the E.coli-T7 ecosystem, when lambda was substituted for T7 in the model, maximal diversity was found at high glucose levels.

These results highlight the limits of applying artificial ecosystem results to those in the field and show that in order to determine the relationship between nutrient density and phenotypic variation, the detailed genetics of the species involved needs to be studied. This new mathematical model can be used to predict the impact of varying nutrient levels on biodiversity in numerous ecosystems and can explain why some ecosystems are teeming with biodiversity whereas others are limited. 

The study provides the first real evidence that the theory known as geographic mosaic co-evolution hypothesis can be observed in real ecosystems. This hypothesis suggested that nutrient availability will only have an impact on diversity if the organism is involved in a co-evolutionary arms race with a pathogen or competitor. Evidence for this was shown in the mini-ecosystems of E. coli and T7 and the mathematical model suggests that this also exists in wild ecosystems.
 
Paper:Understanding the limits to generalizability of experimental evolutionary models. The paper was also discussed in New Scientist.