A number of years ago, I wrote an essay on the potential impact climate change might have on fall leaf color displays (see http://biology.appstate.edu/fall-colors/will-global-climate-change-affect-fall-colors). This was based on a review of the scientific literature at that time, and also on anecdotal information about how the weather and climate affect color and duration and timing. Now, researchers at Harvard University have published the first scientific analysis of the impacts of weather on the timing and duration of fall color, at least for trees that turn red in the fall. They were able to do this because one of their colleagues, John O'Keefe, kept track of when the leaves came out in the spring, and when the turned color and senesced (fall off) in the autumn for the past 18 years, making this dataset perhaps the longest continual such one in the world. The paper I will discuss in this essay is available free to anyone at this website: http://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0057373&type=printable . Granted, it is not easy reading unless you happen to be a statistics whiz, so below, I distill the main points for you.
The authors collated data on leaf color change and senesce from 1993 through 2010 on 3-5 individuals of each of eight different tree species (black cherry, black gum, red and sugar maples, white ash, and several oak species: red, white and black. Observations were made every 3-7 days in the spring during leaf out, and in the fall when leaves senesced, and all the observations were made by O'Keefe, thus limiting variation due to sampling by different individuals. O'Keefe measured the percent of leaves on each tree that had turned color by a certain date, and the percent that had fallen off. These data are publically available at: http://harvardforest.fas.harvard.edu:8080/exist/apps/datasets/da-browse-id.html (datasets HF000, Hf001 and Hf003) should anyone want to see them.
I won't go into the statistical details here, but suffice it to say that they developed two types of models to predict fall color and duration: one involved relating various weather variables, such as monthly temperature and precipitation to leaf color and senescence, and the other used cooling-degree days, a common horticultural index that involves summing up temperatures below a threshold point. Once that threshold is passed, many plants will either sprout, germinate, or flower and it is a component of the USDA's cold hardiness maps for plants.
So, what did the researchers find? First, there were obvious species differences in the timing of peak leaf color, by up to 23 days in some years. The trees that consistently turned early were red maple, white ash, and black cherry. The later turning species were all the oaks, black gum and surprisingly, sugar maple. I say surprisingly, because so many sugar maples here in the southern Appalachians turn early. Perhaps there are genetic differences in phenology (the study of the timing of life cycle events in living organisms, such as when plants leaf out, flower, and set seed, or when animals migrate to and from areas) between northern and southern sugar maples! Check out the National Phenology Network at this website: https://www.usanpn.org/ .
The researchers in this study found that leaf color was first detectable in early September, which sort of jives with what we are seeing here, although I first noticed leaf coloration changes on our sugar maples in early August, nearly a month ahead of the trees at Harvard. The Harvard scientists also noticed large changes from year to year in the timing of fall color: trees turned early in 1994 and later in 2002, years that were cool and warm, respectively.
For red maple, leaf color and leaf fall were both delayed in years when autumn temperatures were warm (similar to what I've been saying for years, so it's nice to see my conjectures confirmed by "hard" data!). Early onset of fall colors, though, sometimes occurred in years with warm spring temperatures. Similar patterns were found for sugar maple, black gum and black cherry trees. The physiological mechanism responsible for this, however, remains a mystery. For white ash, warm temperatures, in contrast to the other species, move up the dates for peak leaf color and leaf fall. It is not clear why this species behaves so differently from the others.
For black oak, warm summer and autumn temperatures improved fall color, although it also delayed its onset to later in the year. Among all species, September temperatures appeared the most influential in determining variations in leaf color and leaf fall. Precipitation, on the other hand, was rarely found to have any effects. That contradicts some of my assumptions about the influence of cloudy rainy weather on anthocyanin production (the pigment that causes leaves to turn red), and so perhaps this is an area still ripe for more study. The authors of this study also did not consider severe drought effects, from which we know in the south through experience can cause premature leaf fall and ruin that year's fall color display. Perhaps there were no droughts up at Harvard during the years of their study.
In the second portion of this study, the authors used future climate scenarios in the Harvard Forest up to the year 2099 to predict the impacts on autumnal leaf color and they obtained some surprising results. Color (as measured by the duration and proportion of leaves that are red, and not by the intensity of the red color on an individual leaf) is predicted to get better, not worse, by the end of this century. The onset of the autumnal display will shift to later in the season, but given the warmer temperatures by then, the duration will increase also. In other words, future scenarios suggest that fall leaf displays will happen later (perhaps by up to two weeks), last longer (at least the early phases) and be more colorful (due to having more leaves in color at any one time, but not necessarily because each leaf has more intense color).
But will this also happen in the southern Appalachians, where significant climate change has yet to occur, at least in most areas? One exception may be at the Coweeta Hydrologic Lab, near Franklin, NC, where researchers have recorded significant rises in temperature for the past two decades. But in the Boone area and in Great Smoky Mountains National Park, there are few if any significant temperature or precipitation trends over the past 70 years. This does not mean they won't eventually occur, and when they do, we may see different trends, since we have so many additional species contributing color down here compared to New England.
I should add a note that this study also does not consider how climate-induced species migrations may affect autumnal colors. Studies by the US Forest Service, for example, suggest that certain trees like sugar maples will migrate northward to avoid excessive temperatures, and their loss from the forest could diminish the intensity and variety of colors that we now observe. This aspect of the climate change story certainly deserves more study.
Given that the fall "leaf peeping" industry during the months of September-November contributes $600-800 million to the economy of North Carolina, and up to $25 billion to all the eastern and Midwest states (see my essay on the economics of fall leaf color: http://biology.appstate.edu/fall-colors/economics-fall-foliage-tourism-north-carolina), changes in the timing, duration and intensity could have very large societal and economic impacts. This latest study by the Harvard scientists offers some glimmer of optimism though, in that the autumnal fall color display, which is certainly one of the most spectacular acts of nature anywhere in the world, will continue to delight us each year for the foreseeable future, even as we humans continue to modify the climate in ways that could threaten our very existence. And if that should happen, and we're no longer here, will the trees still change color in the woods?