What is going on with invasive knapweed? OR How I spent my PhD

Over the course of my PhD work (published here, and most recently here), I have found evidence for evolved differences in phenotype (in other words, in their morphology, development, phenology, stress responses) between native and invasive populations of diffuse knapweed. Why is that interesting? Well the invasive populations didn’t even exist until barely 100 years ago. And something about them has let them succeed and spread over vast areas of their new habitat. Perhaps what has made them so successful can be revealed by comparing the invasive diffuse knapweed populations to their closest relatives, the native diffuse knapweed populations.

Knapweed in the greenhouse, 2009.

There could be several reasons for differences between native and invasive populations of a species, and some of them are hard to test for without genetic data. For example, no one really knows which populations in eastern Europe and western Asia are the source of the populations now invading North America (though we have some ideas), so it’s hard to judge whether we are comparing apples to oranges. Without genetic data (I’m working on it!), the best we can do is try to compare as many populations as possible. Native and invasive populations might be different because natural selection may vary between the ranges. For example, in the native range, the plant may be attacked by herbivores which have specialized on feeding on this plant. In the invasive range, these specialist herbivores may not exist. Without the stress of herbivores, this could relax selection for being tolerant to herbivores, and allow the plant to shift resources to producing more/more competitive babies (Joshi & Vrieling 2005) .

Measuring phenotypes can be complicated. The environment in which the plants are growing can have a huge impact on what we see. Measuring phenotype from ‘natural’ populations means you can’t control for environmental variation, so two populations could be genetically identical, but have different phenotypes because one site got more water than the other, for example. Measuring phenotypes in a greenhouse setting is far more controlled. This allows you to judge evolved differences between populations. However, a greenhouse, with abundant resources, is far from the realities of the field, where the invasion is occurring. Phenotypic differences you see in one environment, may not matter in the other. So we used many populations and we looked at many environments, including the greenhouse, and an experimental field, but also several stress treatments, like drought, flooding, herbivory, and nutrient deficiency. And when we looked at how diffuse knapweed does in many environments, we saw an interesting pattern.

Adapted from Turner et al. 2014. An example of a size trait measured across several different treatments. Red asterisk indicates a significant difference between native and invasive populations; red dot is marginally significant (alpha = 0.05).

Invasive populations did as well or better than natives across nearly all of the environments we tested them in, both under stress and in different experimental conditions. In fact, they seemed to just do well across the board. This suggests that not only are they not trading stress tolerance for growth, but that they can maintain their fitness across environmental variation, and they can do this to a greater extent than native populations. This ability is a type of phenotypic plasticity known by a confusing number of names in the literature, including fitness homeostasis/jack-of-all-trades plasticity/adaptive plasticity. What this means for invasive diffuse knapweed is that it may be able to live in many different environments. In fact, when we take a coarse look at what kind of climatic environments we find both native and invasive diffuse knapweed in, it does seem that invasive populations occur in a broader range of environments.

From Turner et al. 2015. Approximately 700 global occurrence record of diffuse knapweed. PC1 and PC2  represent the largest axes of climatic variation amongst record locations.

As far as we know, this is rare; at least among Holarctic invasive plants, evidence of invasive populations (potentially enabled by adaptive plasticity) spreading to environments outside of those experienced in the native range doesn’t happen often (Petitpierre et al. 2012). While we now have some evidence that North America has been invaded by populations of C. diffusa with enhanced adaptive plasticity, the source of this increased plasticity, and whether it arose before or after introduction, are still open questions.

For way more details, and all the appropriate caveats, see the full papers here and here.

Joshi, J., & Vrieling, K. (2005). The enemy release and EICA hypothesis revisited: incorporating the fundamental difference between specialist and generalist herbivores Ecology Letters, 8 (7), 704-714 DOI: 10.1111/j.1461-0248.2005.00769.x

Petitpierre B, Kueffer C, Broennimann O, Randin C, Daehler C, & Guisan A (2012). Climatic niche shifts are rare among terrestrial plant invaders. Science (New York, N.Y.), 335 (6074), 1344-8 PMID: 22422981

Turner, K., Fréville, H., & Rieseberg, L. (2015). Adaptive plasticity and niche expansion in an invasive thistle Ecology and Evolution DOI: 10.1002/ece3.1599

Turner, K., Hufbauer, R., & Rieseberg, L. (2014). Rapid evolution of an invasive weed New Phytologist, 202 (1), 309-321 DOI: 10.1111/nph.12634