Reeb, RA and Kuebbing, S.E. Phenology mediates direct and indirect interactions amount co-occurring invasive plant species


Methods:
      We conducted an experimental greenhouse study at the University of Pittsburgh, Pittsburgh, Pennsylvania, USA starting in January 2021 to test the effect of neighboring species interactions on the performance of three invasive plant species common to the region. 
Seed and plant collection: We sourced plants from three natural populations per species and randomly assigned individual plants across the experimental mesocosms. All collection sites were located in Pittsburgh, Pennsylvania, USA (44.4406 N, 79.9959 W) and spaced to ensure that populations of the same species were separated by at least 1.5 km. Seeds of the annual M. vimineum were collected in the fall of 2019 and refrigerated until use. To experimentally manipulate M. vimineum phenology in the greenhouse, seeds were forced out of dormancy at staggered intervals in a growth chamber at 20C (see experimental design section below for further details).  Seed germination was high across populations, >90% of seeds, and occurred within a week of exposure to growth chamber forcing temperatures. Seedlings were transplanted into the greenhouse upon appearance of the cotyledon leaves.
      A. petiolata, a biennial, and F. verna, a perennial, were collected as mature winter-dormant plants from the field in order to capture performance effects of all species at reproductive maturity. To experimentally manipulate phenology, basal rosettes of A. petiolata and bulbs of F. verna were collected in their dormancy phase during the winter (January  March) of 2021 at staggered intervals (see experimental design section below for further details). Once collected from the field, plant roots or bulbs were washed and dipped in pesticide solution to prevent pest and disease outbreak. Plants were then directly transplanted into pots in the greenhouse and forced out of dormancy at approximate temperatures of 21C. Greenhouse conditions quickly broke plant dormancy for both species, with new leaf formation observed on all plants within a few days of planting. The initial height of all plants was recorded at the time of planting, with height recorded as < 0.1 cm for M. vimineum seedlings. 
      Experimental design: In greenhouse mesocosm pots, we measured the effect of neighboring species interactions on the performance of three invasive plant species. For each mesocosm, species combinations varied by focal species identity (F. verna, FVf; A. petiolata, APf; and M. vimineum, MVf), for which treatment effects were measured, and the number of neighboring species (one and two species). This gave us the following species combination treatments: six pairwise combinations (FVf+AP, FVf+MV, APf+FV, APf+MV, MVf+FV, MVf+AP) and three three-species combinations (FVf+AP+MV, APf+FV+MV, and MVf+FV+AP). Mesocosms also varied across two treatments: the length of phenological separation between species (natural, reduced, or none). and the stem density of neighboring species (one, two, or four individuals), which are described in more detail below. Each treatment group was replicated five times, totaling to 405 mesocosms (9 species combinations x 3 phenological separation treatments x 3 neighbor stem densities x 5 replicate blocks).
      We experimentally altered the length of phenological separation between species by manipulating the timing of when species broke winter dormancy (A. petiolata and F. verna) or seed germination (M. vimineum). To maintain biological realism, all phenological manipulations preserved the phenological order of species emergence in the field (F. verna first, A. petiolata second, and M. vimineum last). Phenological separation between species was manipulated as follows: Natural (mimicking species phenological separation in the field, 100%), reduced (reducing natural phenological separation by 50%), and none (no phenological separation between species, 0%; Figure 2). Within all treatments for a respective focal species, the planting time of the focal species was held constant and phenological separation was manipulating by staggering the planting time of neighboring species. 
      In pairwise mixtures, each pot contained one individual of the focal species and 1, 2, or 4 individuals of the neighboring species. We then manipulated planting time to reflect treatments of natural, reduced or no phenological separation. Specifically, in mixtures containing F. verna and A. petiolata, A. petiolata was planted 4, 2, or 0 weeks after F. verna. In mixtures containing A. petiolata and M. vimineum, M. vimineum was planted 4, 2, or 0 weeks after A. petiolata. In mixtures containing F. verna and M. vimineum, M. vimineum was planted 8, 4, or 0 weeks after F. verna. 
      In three-species mixtures, each pot contained one individual of the focal species and 1, 2, or 4 individuals of each neighboring species (totaling to 2, 4, or 8 neighboring individuals per pot). We then manipulated planting time to reflect phenological separation treatments, with F. verna always planted first, A. petiolata second, and M. vimineum last (Figure 2). In the natural separation treatment, planting time was staggered by four weeks per species. In the reduced separation treatment, planting time was staggered by two weeks per species. In the no separation treatment, all species were planted at the same time.
      Lastly, we planted five additional monoculture replicates of each focal species, with each monoculture pot containing one individual focal stem and no neighbors (n = 15 total monoculture mesocosms). Monocultures were used as a control for calculating the additive effect of neighboring species interactions on focal plant performance.
      All mesocosms were contained in 6 diameter pots. Pots were housed in a greenhouse bay at the University of Pittsburgh that was maintained at an average temperature range of 16-25C. Pots contained identical potting soil (Jolly Gardener Pro-Line C/B growing mix) and were watered as needed. Plants were treated following standard greenhouse phytosanitation polices including the introduction of beneficial insects and insecticide application to control for common greenhouse insect pests. To account for possible environmental variation within the greenhouse, replicates were blocked across lab benches, such that one pot from each treatment group was housed on a single bench. 
      Data Collection: We measured the performance of focal plants by weighing the biomass of roots, shoots, and fruits at the point of senescence, when leaves were no longer green. Plants were washed to remove soil particulates and dried in a convection oven at 65?C for one week prior to weighing. We selected total biomass (combined root and shoot) as our primary performance metric of focal species in the final analysis. 


Metadata:
Filename: experimental_design.csv
Column descriptions:
ID: Mesocosm identification number.
Focal: focal species code. AP = Alliaria petiolata, FV = Ficaria verna, MV = Microstegium vimineum.
Species_A: Neighbor species code.
Species_B: In three-species mesocosms, second neighboring species code.
Separation: phenological separation in weeks.
Separation_Level: Phenological separation category. Low = no separation. Med = Reduced separation, High = natural separation.
Separation_A: Exact planting time difference (in weeks) of Species A from the focal species. Positive values indicate planting time later than the focal, while negative values indicate planting time earlier than the focal.
Separation_B: Exact planting time difference (in weeks) of Species B from the focal species. Positive values indicate planting time later than the focal, while negative values indicate planting time earlier than the focal.
Focal_week: Planting week of the focal species within the experiment (1-8).
A_week: planting week of species A within the experiment (1-8).
B_week: planting week of species B within the experiment (1-8).
Focal_pop: population code (1,2,3) of the focal species.
Density_A_B: Mesocosm density of species A and B.
A.pop: Population code (1,2,3) of Species A.
B.pop: Population code (1,2,3) of Species B.
Replicate: Treatment replicate number (1-5).
two_way_pair: Species pair code for two-species mesocosms (focal first, neighbor second).
focal_timing: Phenological timing of the focal species, relative to Species A, in two-species mesocosms.

Filename: final_focal_measurements.csv
Column descriptions:
Date: Harvesting date.
ID: Mesocosm identification number.
species: Focal species code. AP = Alliaria petiolata, FV = Ficaria verna, MV = Microstegium vimineum.
Focal_or_competitor: Focal species label only (F).
height: final recorded height in cm.
num_fruits: final fruit count.
num_bulblets: For FV only, number of bulblets.
num_leaves: final leaf count.
Biomass.R: Root biomass in g.
Biomass.S: Shoot biomass in g.
Biomass.Fruit: fruit biomass in g.

Filename: focal_initial_and_biweekly_measurements.csv
Date: measurement date.
ID: mesocosm identification number.
Focal: focal species code. AP = Alliaria petiolata, FV = Ficaria verna, MV = Microstegium vimineum.
Num_leaves: number of leaves.
num_buds: number of buds.
num_flowers: number of open flowers.
num_fruits: number of fruits.
num_bulblets: For FV only, number of visible bulblets aboveground.

Filename: neighbor_initial_measurements.csv
Column descriptions:
Date: measurement date.
ID: Mesocosm identification number.
Species: Neighbor plant species code. AP = Alliaria petiolata, FV = Ficaria verna, MV = Microstegium vimineum.
Competitor_num: individual identification number (1-4) of the stem within a mesocosm.
Num_leaves: initial leaf count.
Height: initial height of plant. For MV seedlings, height was recorded as 0.
 
Filename: Analysis_Code_Reeb_2023.Rmd
Description: R code for data cleaning, data analysis, and supplementary material.







