Biological Fitness: A Discussion of Definintions and Metrics

The concept of biological fitness is foundational for our understanding of both ecology and evolution. Fitness is often described vaguely as an organism’s contribution to the next generation. The reason this is vague is because researchers define and measure fitness differently across fields. I suggest that the myriad definitions and ways to measure fitness commonly employed have led to debates and, seemingly contradictory results. In order to investigate the use of the concept of fitness, I performed a literature review and asked, (1) How is biological fitness defined and used by researchers? (2) How is fitness actually measured by researchers? To address these questions, I surveyed 478 papers published between 2012 and 2016, that included the word ‘fitness’ in the title, and were in the Web of Science categories of ‘ecology’ and ‘evolutionary biology’. In my analysis of the journal articles fitness was only defined 33% of the time. Among studies that did explicitly define fitness, I categorized 18 different definitions, though only 7 were found in more than 5% of papers. I also found differences in how fitness was measured. I found 87 measurements that I grouped into 13 categories. In addition to my survey of the literature, I performed an experiment to explore the relationship between different measures of fitness. Vegetative biomass and reproductive biomass are often both used as metrics of fitness by plant ecologists. In this experiment I determined the relationship between two popular measures of plant fitness vegetative biomass and reproductive yield. I found that these two proxies for plant fitness, vegetative biomass and reproduction, were unimodally related, meaning: 1) intermediate sized plants have the greatest reproductive output, and; 2) for any unique amount of reproduction there is both a small and a large plant with identical reproductive output. Two things emerge from the literature review and the experiment: first, given the many definitions that exist, researchers should be clear about which one they are using. Second, one must be clear about the expected relationship between proxy measurements and fitness, as it may be complex, or non-existent.