Data from: Seed pellets containing activated carbon increase emergence of native plant species used in dryland restoration following herbicide application
Data files
Jun 12, 2026 version files 416.32 KB
-
ac_herbicide_data.csv
344.52 KB
-
composition_data.csv
67.53 KB
-
README.md
4.27 KB
Abstract
Over half of dryland ecosystems worldwide are degraded, making restoration a priority. Most dryland restoration efforts use seed-based approaches, which often result in limited establishment of desirable species. The dual challenge of abiotic stressors and invasive species dominance are key barriers to native plant community re-establishment in degraded drylands. Innovative seeding approaches that help overcome these barriers are needed. We tested seed pellets, a seed enhancement technology, with varying compositions and activated carbon amendments, both for herbicide protection and in the absence of herbicide. We conducted two greenhouse mesocosm experiments. The first experiment tested the protective effect of activated carbon in seed pellets, produced using a low-tech “Bicycle-Powered Seed Pelletizer,” after aminopyralid herbicide (Milestone™) application on diverse seedling emergence. The second experiment examined how seed pellet composition, including variable clay ratios and activated carbon additions in the absence of herbicide, affected seedling recruitment. Seed pellets improved seedling emergence in both experiments, with 136% and 56% higher odds of emergence from seed pellets compared to broadcast seeding in the first and second experiments, respectively. Composition and activated carbon additions without herbicide treatment had limited effects. Following aminopyralid herbicide treatment, we found significantly higher emergence from seed pellets containing activated carbon. Seed pellets with activated carbon may be an effective seeding method in dryland ecosystems where herbicide treatment and reseeding are needed. Varying clay content and activated carbon additions had limited impacts without herbicide treatment. Low-tech seed pellets made using a “Bicycle-Powered Seed Pelletizer,” should be considered in dryland restoration since they improved seedling emergence across all treatments compared to broadcast seeding. Pellet composition had limited effects on emergence under greenhouse conditions, though differences may be more pronounced under low-water field conditions. When invasive species management and seeding are needed simultaneously, seed pellets containing activated carbon provide herbicide protection and offer a low-tech restoration strategy using widely available materials. Without herbicide, activated carbon amendments did not improve emergence. Abiotic stressors and invasion are common challenges; hence, the need for revegetation extends beyond dryland systems. These approaches may be relevant in other degraded systems.
Dataset DOI: 10.5061/dryad.p5hqbzm25
Description of the data and file structure
This data was collected as part of two greenhouse mesocosm experiments investigating seedling emergence using seed pellet technology. The first experiment (activated carbon and herbicide) looked at seedling emergence responses to a fully crossed set of treatments: seeding method (Broadcast or Seed Pellet), activated carbon additions (+/-), and aminopyralid herbicide application (+/-). The second experiment (composition) tested seedling emergence responses to seed pellet composition, including variable clay quantity (low, medium, high) and activated carbon additions (either powdered or granular forms). The response variable in both experiments is seedling emergence as a proportion since a set number of seeds were used in each mesocosm.
Files and variables
File: composition_data.csv
Description: Data collected from the second mesocosm experiment (Composition) that tested seedling emergence from seed pellets with varying clay quantities and activated carbon amendments.
Variables
- observation_date: date when monitoring was completed
- date_seeded: date when mesocosms were seeded initially
- observer_initals: unique identifiers for each observer
- unique_id: identification of each mesocosm, including treatment information, used to account for repeat measures over time
- mesocosm_number: unique number for each mesocosm, treatments were randomly assigned to mesocosms
- treatment: information on experimental treatment, including seeding method (broadcast or seed pellet), clay quantity (no clay, low clay, medium clay, high clay), and activated carbon additions (no carbon, powder carbon, granular carbon)
- bench_location_id: unique id which quadrant on the greenhouse bench the mesocosm was located in, used as a random variable to account for spatial variability in ambient conditions
- seeding_method_clay_amount: Factor variable indicating seeding method (broadcast or seed pellet) and clay quantity (no clay, low clay, medium clay, high clay)
- seeding_method: Factor variable indicating seeding method (broadcast or seed pellet)
- clay_amount: Factor variable indicating clay quantity (no clay, low clay, medium clay, high clay)
- carbon: Factor variable indicating activated carbon additions and form (no carbon, powder carbon, granular carbon)
- proportion_emerged_total_inf: number of seedlings emerged (cumulative over time- seedlings were not removed) represented as a proportion of the number of seeds used in each mesocosm. This has been zero-inflated by adding 0.00000001 to each observation.
File: ac_herbicide_data.csv
Description:
Variables
- unique_id: identification of each mesocosm, including treatment information, used to account for repeat measures over time
- bin_number: unique number for each mesocosm, treatments were randomly assigned to mesocosms
- bench_id: unique id which quadrant on the greenhouse bench the mesocosm was located in, used as a random variable to account for spatial variability in ambient conditions
- carbon: Factor variable indicating activated carbon additions (yes/no)
- herbicide: Factor variable indicating aminopyralid herbicide treatment (yes/no)
- seeding_method: Factor variable indicating seeding method (broadcast/seed pellet)
- treatment: Factor variable that combines carbon, herbicide, and seeding method
- observation_date: date when monitoring was completed
- date_seeded: date when mesocosms were seeded initially
- functional_group: indicates if the emerged seedling was a monocot (grass) or dicot (forb/shrub)
- proportion_emerged: number of seedlings emerged (cumulative over time- seedlings were not removed) represented as a proportion of the number of seeds used in each mesocosm.
- proportion_emerged_inf: number of seedlings emerged (cumulative over time- seedlings were not removed) represented as a proportion of the number of seeds used in each mesocosm. This has been zero-inflated by adding 0.0001 to each observation.