Skip to main content

Human mediated land use change drives intraspecific plant trait variation

Cite this dataset

Schroeder, Hayley; Grab, Heather; Kessler, Andre; Poveda, Katja (2020). Human mediated land use change drives intraspecific plant trait variation [Dataset]. Dryad.


In the Anthropocene, more than three quarters of ice-free land has experienced some form of human driven habitat modification, with agriculture dominating 40% of the Earth’s surface. This land use change alters the quality, availability, and configuration of habitat resources, affecting the community composition of plants and insects, as well as their interactions with each other. Landscapes dominated by agriculture are known to support a lower abundance and diversity of pollinators and frequently larger populations of key herbivore pests. In turn, insect communities subsidized by agriculture may spill into remaining natural habitats with consequences for wild plants persisting in (semi) natural habitats. Adaptive responses by wild plants may allow them to persist in highly modified landscapes; yet how landscape-mediated variation in insect communities affects wild plant traits related to reproduction and defense remains largely unknown. We synthesize the evidence for plant trait changes across land use gradients and propose potential mechanisms by which landscape-mediated changes in insect communities may be driving these trait changes. Further, we present results from a common garden experiment on three wild Brassica species demonstrating variation in both defensive and reproductive traits along an agricultural land use gradient. Our framework illustrates the potential for plant adaptation under land use change and predicts how defense and reproduction trait expression may shift in low diversity landscapes. We highlight areas of future research into plant population and community effects of land use change.


Two wild brassica species (Barbarea vulgaris and Thlaspi arvense) were collected in July, 2017 and on species (Capsella bursa-pastoris) in July, 2019 from the field margins of farms (hereafter called sites) in the Finger Lakes Region of Upstate New York. We collected B. vulgaris from 10 sites, T. arvense from 9 sites and C. bursa-pastoris from 15 sites. Seeds from each site were germinated in petri dishes and then individually planted in 4-inch pots in a greenhouse common garden in the fall of 2017 for B. vulgaris and T. arvense and the fall of 2019 for C. bursa pastoris. These experiments provide preliminary data intended to spur more rigorous examination of the hypotheses presented in this review. We do not intend for the data provided to stand alone.

Defense traits: In the fall of 2017, we conducted a caterpillar bioassay on B. vulgaris. Plants were sourced from 10 sites, with 8-10 plants grown from each site (apart from one site in which there were only two replicates due to low germination). Two leaves were collected from each plant and placed in individual floral water tubes. Each tube was then inserted into a plastic lid and covered with a plastic cup to create a rearing chamber. Two Trichoplusia ni neonate caterpillars (sourced from Benzon Research) were placed on each leaf and allowed to feed for three days. At the conclusion of the bioassay, we recorded the mass of each surviving caterpillar. Caterpillars experienced 14% mortality across the experiment. Interestingly, of the two caterpillars placed on an individual leaf, across all replicates one caterpillar developed normally while the other’s growth was stunted. In our analysis of caterpillar growth, we used only the caterpillar that was successful on each leaf. 

Floral traits: In December 2017, T. arvense plants were vernalized in a refrigerator at 4ºC and returned to the greenhouse early in 2018. Upon blooming, we measured petal width using handheld calipers as a proxy of flower size. Flowers were measured each week during the blooming period so that 1-3 flowers were measured per plant depending on the length of the bloom. Overall, 158 T. arvense flowers were measured representing 2-15 plants per site. 

Self Compatibility: C. bursa-pastoris seeds planted in August 2019 bloomed without vernalization in October and November. Each plant was allowed to self pollinate with no hand pollination. After the plants finished blooming, we collected and weighed the seeds from the first 15 non-aborted pods.


Cornell University

National Institute of Food and Agriculture, Award: 2017-07100