Skip to main content
Dryad logo

Data and code from: Spring phenological escape is critical for the survival of temperate tree seedlings


Lee, Benjamin; Ibanez, Ines (2021), Data and code from: Spring phenological escape is critical for the survival of temperate tree seedlings, Dryad, Dataset,



Understory plants in deciduous forests often rely on access to ephemeral light availability before the canopy closes in spring and after the canopy reopens in fall, a strategy commonly referred to as phenological escape. Although there is evidence for a relationship between understory plant phenology and demographic performance, a mechanistic link is still missing.

In this study, we bridged this gap by estimating annual carbon assimilation as a function of foliar phenology and photosynthetic capacity for seedlings of two temperate tree species that commonly co-occur across eastern North America. We then modeled the relationship between estimated carbon assimilation and observed seedling survival and growth.

Our results indicate that seedlings of both species strongly depend on spring phenological escape to assimilate the majority of their annual carbon budget and that this mechanism significantly affects their likelihood of survival (but not growth). Foliar desiccation also played a strong role in driving patterns of seedling survival, suggesting that water availability will also help shape seedling recruitment dynamics. We found only weak associations between seedling senescence in fall and annual carbon assimilation, suggesting that phenological escape in fall plays a relatively minor role in seedling demographic performance.

Our results indicate that spring phenological escape is critical for survival of these temperate tree species, and thus any changes to this dynamic associated with climate change could strongly impact these species’ recruitment.



These data were collected between 2014-2019 by Benjamin R. Lee and include model code for the photosynthesis, survival, and growth models outlined in the corresponding article. Models were conducted in OpenBUGS which is free to download online.

Data for all models are included in the code documents in collapsible data-frames that are clearly labeled. Initial values are also recorded at the bottom of each model.

Posterior parameter estimates are available in the supplemental materials of the corresponding article.

Usage Notes

All models were run separately for each species.

Parameter definitions for the photosynthesis model are provided in the supplemental materials for this article along with an in-depth description of the model structure. 

Curve-level variables for photosynthesis models: plant = individual replicate identifier, curveid = individual curve replicate identifier, AQ = 1 indicates ACi curve and 2 indicates AQ curve, PS = centered soil moisture value (see notes on soil moisture sub-model in supplement), VS = centered Vapor pressure deficit values.

Observation-level variables for photosynthesis models: curve = curve identifier (same as curveid above), Obs = individual gas exchange measurement (multiple in each curve), Aobs = observed photosynthetic rate (A), Q = light level (PPFD) inside the cuvette, Pressure = atmospheric pressure, CiP = interstitial pressure, O = partial pressure of O2

Survival/growth model variables: Site = (1 = George Reserve, 2 = Radrick Forest, 3 = Saginaw), surv = recorded as either alive (1) or dead (0), growth = centered annual growth value, AnnC = centered estimate of annual carbon assimilation, PctDam = percent of leaf recorded as damaged by end of growing season, Des = binary desiccation damage indicator (1 = present, 0 = absent), Deer = binary deer herbivory indicator (1 = present, 0 = absent), SdlgA = individual identifier for Acer seedlings, SdlgQ = individual identifier for Quercus seedlings, Year starts at 1 indicating data collected in 2015, GSF = Global Site Factor (centered around mean), Cohort = Seedling cohort (1-3 = seedlings planted in 2014-2016, respectively), Age = age of seedling in years at time of measurement