Phenological sensitivity as a mediator of plant interactions

Abstract

For temperate plants, spring phenology, or the timing of annual life cycle events such as seed germination, leaf emergence and flowering, is primarily controlled by temperature and light cues. Across a diversity of plant taxa, increasing global temperatures due to anthropogenic climate change have produced corresponding shifts in spring phenology. Yet variation in the strength of these shifts is substantial among species, populations, and even between separate phenological phases in the same organisms. This variation in phenological sensitivity at multiple scales of biological organization will impact a myriad of ecological and evolutionary processes. In this dissertation, I address the physiological drivers and ecological consequences of interspecific differences in phenological sensitivity across two phenological events: the flower-leaf sequences of temperate woody plants and the germination phenology of woodland herbs.

In Chapter 1, I evaluate several long-standing hypotheses about the adaptive evolution of flower-leaf phenological sequences in temperate woody plants. Through modeling the covariation between flower-leaf sequence patterns and traits associated with these hypotheses, I demonstrate that treating flower-leaf sequences as a continuous, quantitative measure substantially improves our understanding of the function of flower-leaf sequence variation over the traditional, categorically descriptive, framework. I find that the evolution of flower-leaf sequences is phylogenetically structured and associated with wind-pollination, early flowering, and—within biotically-pollinated taxa—drought tolerance. These associations suggest there are multiple drivers of flower-leaf sequence variation in temperate woody plants.

In Chapter 2, I compare the phenological responses of flowers and developing leaves for 10 temperate woody species to varying levels of temperature and photoperiod in an experiment to test competing hypotheses regarding how environmental cues determine flower-leaf sequence variation. Specifically, I ask whether forcing alone drives variation or differential sensitivity to chilling and photoperiod influence sequence variability. I find that flower and leaf phenology responded with differential sensitivity to environmental cues, with differences in their response to chilling being the dominant driver of flower-leaf sequence variation. I also find that the largest shifts in flower-leaf sequences due to climate change will be for species that flower before leafing.

In Chapter 3, I simulate winter and spring temperature variation in growth chambers to evaluate how differences in phenological sensitivity to temperature influences relative germination timing—termed “phenological advantage”—among 11 herbaceous woodland plants. I then leverage this interspecific variation to indirectly manipulate the strength of germination priority effects in a pair-wise competition trial to quantify how much phenological differences contribute to the competitive dynamics between an invasive and a native herb. I find that phenological priority effects strongly influence the competitive dominance of the invader, doubling its negative impact on native plant biomass relative to its intrinsic competitive ability. Differences in phenological sensitivity to cold stratification between the native and invasive taxa in my germination assays suggest that climate change may favor rapidly germinating invasive species, further augmenting their phenological advantage over native plant communities.

In Chapter 4, I present an experimental framework for robustly evaluating the interactive effects of temperature and photoperiod sensitivity in phenology experiments. In particular, I detail a common experimental design that co-varies thermo- and photo-periodicity, and demonstrate, using both a mathematical proof and comparative data analysis, how this experimental artifact results in routinely over- or under- estimating phenological sensitivity to forcing and photoperiod. I present several experimental designs that can correct this problem, providing a path forward for improving experimental inference about phenological sensitivity to environmental cues.