COS 18-6 - Seed production response to climate and its effect on forests species composition.

Background/Question/Methods

Tree recruitment can be a major bottleneck for determining species composition in forests. The number of seeds produced varies among species, as well as being sensitive to weather and tree size within a species. Thus, it can be challenging to quantify the relative role of seed production on forest species composition and dynamics. To address this, we created seed production equations as a function of weather and tree size and integrated them into the forest gap model ForClim. Generalized linear mixed-effects models were made for 15 of the most common tree species found in the Pacific Northwest (PNW). The predictor variables for seed production were diameter at breast height (DBH), spring minimum temperature, spring temperature anomaly, summer mean temperature, summer mean temperature anomaly, and moisture deficit. Seed production data and DBH of individual trees came from the Mast Inference and Prediction (MASTIF) network. Parameter-elevation Regressions on Independent Slopes Model (PRISM) data provided the temperature and precipitation data. The species composition from the simulated forests both with and without the seed equations were used to determine the effects of seed production on species composition.

Results/Conclusions

The seed production response to spring and summer temperatures varied between species. Some species experienced an increase in seed production when both the spring and summer temperatures increased. For example, seed production of Abies grandis and Pinus ponderosa increased with both warmer spring and summer temperatures. Alternatively, seed production in Thuja plicata and Tsuga mertensiana decreased as spring and summer temperatures increased. Other species, such as Pinus contorta, have different seasonal temperature responses where seed production had a positive relationship with spring temperatures and a negative relationship with summer temperatures. Seed production was also strongly influenced by DBH and typically peaked at species-specific sizes. For example, Pseudotsuga menziesii seed production peaked at ~150 cm DBH and Abies amabilis peaked at ~75 cm DBH. Predicting the seed production of species in a future climate can help predict species composition. Tsuga mertensiana is found in forest sites that have low temperatures and will experience a decrease in seed production if temperatures continue to increase. Producing fewer seeds can decrease the chance for the recruitment of that species. Understanding these species-specific relationships for seed production can help us predict future forest species composition.