Why individuals of the same species are not identical
At the most basic level, within-species variation can come from genetic differences, environmental variation, or measurement error [2]. A tree's drought tolerance, for instance, depends partly on physiological adjustments like osmotic adjustment—changes that help cells hold water—and on carbon isotope discrimination, a signal of how efficiently the plant uses water [1]. These traits vary because individual trees experience different conditions and may also carry different genes.
The relative importance of these sources is not fixed. In a study of trembling aspen and ponderosa pine, almost all measured traits varied more from one year to the next than across an elevation gradient [1]. That means a single tree can change its physiology substantially depending on whether the year is wet or dry, and two trees at the same elevation can differ just as much as trees from different elevations. Variation is therefore not just a fixed property of a population; it is partly a dynamic response to current conditions.
How local climate and weather leave their mark
Environmental gradients can produce predictable differences among populations. In two dune-engineering grass species, seed germination responses to temperature followed opposite trends depending on the climate of origin: populations from severe-winter sites germinated better when cued by warmth, while populations from warm, dry-summer sites germinated better when cued by cold [5]. Seed mass was also smaller in sites with low precipitation, and rainfall continentality explained more of that variation than summer aridity did [5].
But the environment does not act alone, and its effects can be surprisingly local. Within-population variability in seed germination accounted for 5 to 54 percent of the total variation, while seed mass variability was lower than 40 percent [5]. Even across a broad latitudinal gradient, a large share of the differences occurred among individuals growing side by side, and the variability only partially correlated with the measured climatic variables [5]. This suggests that local microconditions, genetic differences, or developmental noise can matter as much as the broad climate gradient.
Variation in behavior can outweigh species differences
The same principle applies to behavior. In a systematic comparison of bonobos and chimpanzees, the two species showed similar overall tendencies to console distressed group members, but within-species analyses revealed substantial variation [4]. Bonobo consolation was most often directed toward and received by younger individuals, while chimpanzee consolation was most often directed toward close social partners [4]. In both species, males and females consoled less as they aged, with some evidence that chimpanzee males consoled more than young females [4].
The authors concluded that within-species variation in these socio-emotional abilities was greater than the average difference between the two species [4]. In other words, knowing whether an individual is a bonobo or a chimpanzee tells you less about its consoling behavior than knowing its age, sex, or social relationships. This is a concrete example of how individual and social factors can generate more diversity inside a species than exists between closely related species.
Why within-species variation matters for predicting change
Variation is not just noise; it can determine whether a population persists under stress. In trembling aspen, trees at low elevation were pushed past their drought tolerance limit during a severe drought year, and higher specific leaf area during drought was correlated with more canopy dieback the following year [1]. Ponderosa pine showed less within-species variation in functional traits, though osmotic adjustment was still a key mechanism for drought tolerance [1]. The scope of variation therefore affects how differently two species may respond to the same climate stress.
For scientists, ignoring within-species variation can distort conclusions. When phenotypic correlations within species differ from evolutionary correlations between species, estimates of evolutionary coefficients are pulled toward the phenotypic values, and evolutionary rates can be under- or overestimated depending on the mismatch [2]. In a plant genus (Polemonium), leaflet size showed a strong negative correlation with elevation across species, even though the correlation within species was positive [2]. That reversal is a reminder that patterns measured within a species do not automatically scale up to explain differences among species.
Sources used in this answer
Quantifying within‐species trait variation in space and time reveals limits to trait‐mediated drought response
In trembling aspen and ponderosa pine, drought-related traits varied more year-to-year than across elevation, and osmotic adjustment and carbon isotope discrimination were key determinants of drought tolerance, with aspen at low elevation suffering mortality in a severe drought year.
Accounting for Within-Species Variation in Continuous Trait Evolution on a Phylogenetic Network
Within-species trait variation can arise from genetic variation, environmental variation, or measurement error, and failing to account for it can bias estimates of evolutionary rates and correlations.
Same, same but different! Exploring children’s understandings of within-species variation
Primary school children's explanations for within-species variation shift with age toward more scientifically acceptable thinking, but misconceptions are common and can serve as stepping-stones in a learning progression.
Within-species variation eclipses between-species differences in Pan consolation
In bonobos and chimpanzees, consolation behavior varied more within each species than between the two species, with age, sex, and social partner preferences shaping who consoles whom.
Within-species variation of seed traits of dune engineering species across a European climatic gradient
Seed germination responses and seed mass in two dune grass species varied substantially within and among populations across a climatic gradient, but the variability only partially correlated with climatic variables.