Of the many plant phenology studies recently undertaken worldwide, most have been concentrated in moist temperate zones, places such as western Europe, eastern North America, and Japan. At these locations, temperature is the most common controlling factor in determining the timing of spring flowering and leaf bud break, whereas precipitation, and resulting soil moisture, is not often a limiting factor (Ibanez et al., 2010; Schwartz & Hanes, 2010). The findings from these studies are not universal, however, as there are large areas of the world covered by deserts, tropical deciduous forests, tropical and temperate grasslands, and other dryland habitats in which precipitation plays a much more important role in dictating the timing of biological processes (Peñuelas et al., 2004). Phenology is relatively understudied in these systems, especially considering their contributions to global biodiversity and ecological processes. Consequently, the paper on the phenology of Arizona desert plants by Crimmins et al. in this issue of New Phytologist (pp. 468–479) is especially welcome. Their paper describes a 26-yr record of summer flowering phenology along an elevational gradient in the mountains of southeastern Arizona (USA). Their study shows that in this system and season, the timing and amount of rainfall are the dominant factors in determining flowering phenology. That finding may not be particularly surprising to many, but its documentation is important, as are many of the implications that the authors identify for the ecology of the area and the study of phenology more broadly.
‘These one-of-a-kind historical data sets will complement well-organized, national phenology monitoring programmes …’
Many fields of ecological research have been re-invigorated and re-analysed from the perspective of a changing climate: phenology is an exemplar of such a revitalized field (Cleland et al., 2007). In the past, phenology researchers most often simply characterized phenological schedules or calendars, or investigated how climatic variation was linked to plant or animal behaviours, such as flowering or bird migration times (Bradley et al., 1999; Miller-Rushing & Primack, 2008). However, researchers are now investigating how climate change is causing, or not causing, species to change the timing of their life cycles and their ability to survive in their present ranges (Chuine, 2010; Willis et al., 2010). Researchers are then using those data to predict changes in the abundance and distribution of species in response to coming decades of climate change, with associated implications for ecosystem processes and conservation (Morin et al., 2009). New perspectives and insights concerning the role of phenology in a changing climate are forcing a rethinking of phenology’s place in ecological and evolutionary theory and conservation practice (Forrest & Miller-Rushing, 2010). Research from Arizona presented by Crimmins et al. provides a valuable contrast with this growing body of phenology research mainly based on temperate systems, and yields important insights into some of the challenges facing researchers in this field.
First, and perhaps most fundamentally, the paper highlights the difficulty of measuring phenology in a generalized way (Parmesan, 2007). Recent publications (drawing from a temperate-system concept of phenology, in which species typically flower in a single intense period) have encouraged investigators to favour analyses of mean flowering dates rather than first observations (Miller-Rushing et al., 2008). However, in many arid systems, where bouts of flowering can occur multiple times throughout the season, sometimes for extended periods, changes in mean flowering dates have much less meaning. Characterizing changes in phenology in arid regions is much more complex than in most temperate systems. Similarly, in most discussion of phenology, leafing out is also assumed to occur in one single burst in the spring, whereas in dry habitats, leaf out can be a complex process that is linked to rainfall. This sets up particular challenges for national and international efforts to measure and evaluate changes in phenology across diverse habitats (Cleland et al., 2007; Parmesan, 2007), where the distribution of phenology may be fundamentally different.
Second, the paper highlights the need for models used to forecast changes in phenology (e.g. Morin et al., 2009; Ibanez et al., 2010) to take a broader perspective and include variables that are important in a variety of systems. Recent papers have shown that phenology models must consider winter chilling requirements, rather than simply spring warming as measured by growing degree days or average spring temperature (Morin et al., 2009; Schwartz & Hanes, 2010; Yu et al., 2010). Work like that of Crimmins et al. pushes the need to consider more climate variables in phenology models even farther – highlighting the need to consider both the amount and timing of precipitation and perhaps other variables as well, such as extreme temperature and precipitation events.
Third, Crimmins et al. identify several intriguing patterns in the observed changes in phenology, patterns that may be important for forecasting ecological consequences of climate change. For example, different life forms, such as woody shrubs, annuals, and herbaceous perennials, differ in their flowering phenology and responses to changes in precipitation. Also, the total flowering season was compressed in drier years, meaning that there was more overlap in flowering times. As precipitation changes, its effects on these different life forms and the interactions among species, including other trophic levels, will likely have important consequences for the ecology of species of this system, potentially leading to losses of historical mutualistic or competitive relationships and gains in new ones. Patterns from observational data like these will be important to follow up with more targeted analyses of interacting species, and field and glasshouse experiments that examine more closely the mechanisms driving changes in phenology, especially experiments that manipulate water availability. Such experiments can also be used to evaluate resulting changes in ecological interactions and processes. In addition, more work is needed on the physiological perception of the environmental signals and genetic and molecular responses that lead to leafing out and flowering.
Fourth, there is still a strong need for monitoring phenology and discovering and analysing historical data sets in a variety of locations. Our understanding of phenological changes is very limited relative to their potential impact on ecology and biophysical systems worldwide (Forrest & Miller-Rushing, 2010). The need for more information is strongest in nontemperate systems like the arid environment studied by Crimmins et al., but we know relatively little about the role of phenology in the ecology of even the best-studied systems. There are several well-known phenology records that have received a disproportionate amount of attention to date. These data sets – such as the > 1000-yr record of cherry blossom times in Kyoto (Aono & Kazui, 2008), Marsham family records from the UK (Sparks & Carey, 1995), grape harvest records from France (Chuine et al., 2004), and observations of Aldo Leopold (Bradley et al., 1999) and Henry David Thoreau (Miller-Rushing & Primack, 2008) in the United States – are certainly invaluable for their contributions to science and communication of climate change. However, there are many more data sets out there collected by other less well-known individuals in less famous places. Most will not reflect the level of work put into the data set collected by David Bertelsen in Arizona (Crimmins et al., this issue), which contains data collected during 623 hikes of a 16-km trail up Mt Kimball in Arizona over a 26-yr period, but much of the data will be valuable, or at least useful, nonetheless. These one-of-a-kind historical data sets will complement well-organized, national phenology monitoring programmes in Japan, Europe, the United States, and several other countries and regions across the world.
The story of changes in phenology is complex – and it is much more than just a trend toward earlier springs, the most common change communicated to the public. The phenologies of each species are changing in unique ways, and such changes depend on a multitude of factors, making forecasts of future changes difficult. The diversity of changes will likely alter ecosystems in substantial ways as species change in phenology, abundance and distribution. Crimmins et al. make an important contribution to this field – broadening the study of phenology and climate change. We hope that their work in Arizona continues and that others follow their lead.