Adaptation to the urban heat island in weedy plant species across US cities

Background/Question/Methods

The urban heat island (UHI) is a powerful environmental filter and has potential to be a strong ecoevolutionary force on urban plants. Weedy plant species in particular have been successful in the urban environment, with adaptations in response to temperature and habitat fragmentation. The question remains: to what extent are these changes in plant traits the result of adaptive evolution or phenotypically plastic responses to UHI? In addition, it is unknown whether or not plant species are responding similarly across cities (i.e. parallel evolution). To assess this, we compared an urban weed (Digitaria sanguinalis, hairy crabgrass) in a common garden greenhouse from 10 sites each in two cities: Baltimore, MD and Minneapolis-St. Paul, MN. Broadly, Baltimore is warmer than Minneapolis-St. Paul, and both experience strong UHI effects of appx. 8º-10º C higher air temperatures compared to non-urban areas. Seeds were collected along a gradient of UHI strength in summer 2018 and selfed for one generation (F1) in order to eliminate maternal effects (n=189). Land surface temperature (LST) was calculated to infer geographic air temperatures from Landsat-8 satellite imagery. Traits were measured and include specific leaf area (SLA), leaf dry matter content (LDMC), number of ramets, and aboveground biomass.

Results/Conclusions

Overall, LST from all sites ranged from 8.90 ºC to 76.20 ºC, with a mean temperature of 38.70 ºC in Baltimore and 38.15 ºC in Minneapolis-St. Paul. Interestingly, increasing LST is positively correlated with an increase in the number of ramets in Baltimore lines (tLM=0.0216, p=0.0216). In addition, D. sanguinalis from Baltimore has significantly greater LDMC (FANOVA=9.325, p=0.00259) and number of ramets (FANOVA =6.029, p=0.015) than those from Minneapolis-St. Paul,. This may suggest that populations of these weeds in Baltimore may be prioritizing recovery from tissue damage (i.e. by herbivory or trampling) via growth of new ramets. Alternatively, higher LST may result in higher photosynthetic rates, and the subsequent production of new ramets may be a result of higher carbon gains. Conversely, in Minneapolis-St. Paul lines, increasing LST is correlated with an increase of SLA (tLM=-2.165, p=0.033), and lines from Minneapolis-St. Paul have greater SLA (FANOVA =23.59, p< 0.05) and aboveground biomass (FANOVA =6.808, p=0.00981) than Baltimore lines. This may indicate that plants from Minneapolis-St. Paul are prioritizing maximizing photosynthetic activity. Based on these results, these data do not suggest parallel evolution in D. sanguinalis traits, though it is possible they are both opting to increase primary productivity.