Tree Coring Tree cores will be collected in the field from 198 red maples, sugar maples, and red oaks, with two cores taken per tree to account for uneven growth. Approximately half of the trees are located at sites in Southern Michigan, the other half at Northern sites near the UM Biological Station. All living adult trees of the three species present in the designated study sites were sampled. This does leave room for variation in age, size, and health amongst individuals sampled, as well as some variation in environmental factors such as access to light. However, such variation will largely be overcome by the following dendrochronological analysis. Cores will be collected using an increment borer at approximately 1.4 m above the ground (Tree Coring and Interpretation). Cores will then be placed in plastic straws for protection in the field and labelled with species and location information. Once all cores are collected, they will be sanded with increasingly fine materials until all tree rings are clear and can be analyzed by a specialized computer program at UM. Dendrochronological Analysis Tree cores will be analyzed by using a dendrochronological computer program, which will measure the width of each ring. Tree rings will also be analyzed in order to create a more coherent timeline across individuals and align the core data with environmental and seed data. By assuming that a high or low growth year will occur across a site in all individuals due to similar environmental conditions (Bernabei and Bontadi, 2012), the program can attach dates to each ring. Though not infallible, as conditions can vary even within a site, generally covariation among trees allows for dating of each ring within one year of accuracy (Helama et al, 2003). This allows for comparison between species, as there is an array of ages across all sites and enables the connection of the growth data to the corresponding annual environmental conditions and reproductive efforts. Seed Collection Seed collection began in 2008 for some sites, 2014 for others. 15 nets were set up at sites in a grid pattern, with 3 columns of 5. Nets were balanced on metal rods approximately 1 foot off the ground. The use of such traps is an established method for long-term sampling of seed production within a forest community (Caignard et al, 2017). Twice a year, once in the spring and once in the fall, the contents of each net were collected. All material was discarded with the exception of seeds, which were identified and tallied, allowing for an approximation of seed production across species at each site. There did not appear to be common species-level masting events captured by these traps, as is not uncommon and has been seen previously in studies in temperate forests (Koenig et al 1994). This re-enforces the focus on non-masting years, given the apparent rarity of a true, synchronous masting event. Some data was lost due to the partial or destruction of the nets or rods attributed to herbivory, weather, or human disruption of the study site. In these cases, all contents were discarded given the unknown nature of the individual damaging event. While this method does not allow for a exact precision in terms of what tree produced each seed, we are able to best estimate using the location of each trap and plotting the trees in turn. Environmental Data Collection Temperature, soil moisture, photosynthetically active radiation (PAR), and relative humidity (RH) were collected at each site beginning in either 2008 or 2014 using HOBOware sensors. Data was downloaded from sensors twice a year, once in the fall, once in the spring. This will allow for a greater understanding of change over time at all study sites in an array of environmental conditions. Many of these shifts, namely those in temperature and the length of growing seasons, are attributed to climate change and highly influential on plant growth (Way and Montgomery, 2015). As such, being able to view heightened temperature over time or elongating spring and summer conditions will help solidify these greater climate changes at our study sites. It will also allow us to compare more accurately between sites, especially those in different parts of the state. It is unclear how Northern sites will respond as their growing season begins to further resemble that of Southern sites, nor is it known if the difference in growing seasons will remain constant as it lengthens equally in both parts of the state. Data was then processed to remove any incorrect data, such as error messages or data point not representative of reality (i.e., negative values for soil moisture) Statistical Modeling While still unclear what the modeling process will look like exactly, the hope is to look at the associations between growth and reproduction and determine if the former can be used a predictor for the latter. This will likely take the form of a multiple linear regression involving some combination of growth, reproductive, and environmental data. It is unclear yet the variations between tree species, and if they can be looked at as a whole or will be categorized by species for the entirety of the project. The environmental data collected will serve as both a constant, in terms of allowing us to compare between sites based on the similarity of external factors but will also help illuminate potential changes over time that could be attributed to climate change. This can include the ever-increasing temperature and the continuation of what may be labelled “spring” or “summer” conditions, such as elevated temperatures or rainfall, even as photoperiods shift towards fall. Using this information, we might be able to create a more complex model to forecast future reproduction rates based on predicted growth patterns, which are highly related to environmental conditions. This again could be done on a species-basis, but would be more valuable if we can find a more universal pattern and create a more general model for the entire communities we are studying. This model would hopefully allow for the prediction of reproductive effort based on predicted environmental factors and growth.