Modeling size-based reproductive output for dominant perennial graminoids in southwestern ponderosa pine forest ecosystems

—Understanding plant population dynamics requires demographic data on survival, growth, and reproduction. While survival and growth are relatively straightforward to measure, reproduction can be difficult to quantify, especially for graminoids. To increase our understanding of reproductive output (seed production) for 10 dominant perennial graminoids in southwestern ponderosa pine forest ecosystems, we modeled seed production as a function of 3 plant traits: plant size (basal area cm²), number of flower stalks (log flower stalks), and maximum flower stalk height (cm). We developed generalized linear models to predict the reproductive output (seeds produced) and flowering probability by species as a function of plant size. Then, we demonstrate how these seed production data can be used to parameterize the reproductive output models (fecundity kernels) within a size-structured plant population model (integral projection model). To parameterize the fecundity kernel, we leveraged 19 years (2002–2020) of long-term permanent quadrat vegetation map data to estimate annual and mean recruitment rates and recruit size distributions as a function of plant size for each species. As expected, the number of seeds as a function of plant size exhibited strong positive relationships for 9 of the 10 perennial graminoid species. Recruitment rates for all species over the 19-year period were low (<3%). The fecundity kernels demonstrated that some perennial graminoid species produced recruits from individuals across their size distribution, while other species only produced recruits from large individuals. Our results support the hypothesis that plant size is a reliable predictor of the number of flowering stalks and seed output in long-lived perennial graminoids. These fecundity estimates can then be integrated into long-term demographic studies that track individual growth and survival to generate fully parameterized structured plant population models in the southwestern ponderosa pine understory.