sign in sign up
<<
Home , Autumn leaf color, Quercus alba

-out Dates Highlight a Changing Climate

Caroline Polgar and Richard Primack

he arrival of spring is heralded each year are actually relatively consistent differences by striking displays of flowers on among species in leaf-out dates, as well as large Tand . Perhaps less conspicuous differences from year to year in the timing of than the blooming of flowers, the emergence of leaf out. The study of the timing of leaf out new on woody marks the onset of (and other natural annual phenomena) is known the growing season and controls a host of eco- as phenology. system functions. While to the untrained eye Much of what we know about the physiology it may seem as though the leaves come out at of leaf and development, and the mecha- the same time each year in one big burst, there nisms behind various leaf-out strategies, comes Richa r d P i m ack and C a oline Polga

Clockwise from top left: tend to leaf out early in the spring, while tend to leaf out in late spring. and beeches tend to leaf out in the middle of spring. Leaf-out Dates 15

as red (), white (Quercus alba), and red oak (Quercus C a r oline Polga rubra) (O’Keefe 2010). The Japanese Meteorological Agency has been recording leaf out, flowering times, leaf color, and other phenological data of individual marked plants in phenological gardens at over 100 weather sta- tions since 1953 (Ibanez et al. 2010). Ginkgo (Ginkgo biloba), also known as maidenhair , is one of the species being moni- tored. The International Phenological Gardens (IPG) project, a network of botan- ical gardens across Europe, has been collecting simi- Young red maple (Acer rubrum) leaves begin to expand in the spring. lar data on leaf-out dates of individual plants since from research done in the past by tree physi- 1951 (Menzel 2000). Other leaf-out datasets go ologists and foresters who were interested in back even further, including one that Henry the connections to tree growth and timber har- David Thoreau compiled in the mid-nineteenth vests. Over the past few years, however, the century on plants in Concord, Massachusetts. range of people interested in leaf-out phenol- Findings from these studies indicate that there ogy has grown, as have the methods employed can be large year-to-year variability in the tim- to study it, largely as a result of the relevance ing of leaf out, depending on the weather, and of this phenomenon to global . that there tends to be relative consistency in New technology, including satellite data, is the order of leaf out of species from year to year now being used to monitor leaf-out timing over (Lechowicz 1984). wider areas than was possible in the past. Forest Obtaining annual observations of leaf-out and ecosystem ecologists are connecting these dates can be quite time and labor intensive, leaf-out date observations to larger issues of often limiting studies to a small area around a global climate change, with implications for field station or a small number of species. To carbon cycles, the availability of fresh water, measure leaf out on a larger scale, remote sens- and production. ing has emerged as a valuable new tool that can monitor an entire community or eco- Monitoring leaf out system consisting of many different kinds of The recent resurgence of interest in the phenol- plants. Remote sensing studies typically use ogy of woody plants has led to leaf-out moni- data obtained by sensors on orbiting satellites, toring projects around the world. For example, such as the Advanced Very High Resolution over the past twenty years Dr. John O’Keefe and Radiometer (AVHRR) and the Moderate-reso- other ecologists at the Harvard Forest in central lution Imaging Spectroradiometer (MODIS), or Massachusetts have been recording the dates of equipment on Landsat satellites. Satellite sys- leaf emergence of individual trees and shrubs tems vary in their spatial resolution, frequency each spring, including such common species of coverage, and types of data gathered. Scien- 16 Arnoldia 68/4

A Date of Average Green Leaf Onset (P ) a r d 2007 tists use data transmitted from 110 165 these satellites to calculate the changes in the amount of green vegetation (greenness) there is

Fi s he r and Mu st in a certain area over a growing season. Analysis of graphs of greenness over time can be used to quantify important dates in the growing season, such as date of first leaf out in spring, the date at which half of the leaf cover has developed, and canopy senescence in autumn. Several recent research papers have shown that regional leaf-out data from satellites accurately match ground observations. This is particularly important because there is concern that 0 25 50 100 Km different topographic features, such as mountains, fields, cities, B and lakes, might create errors in the detection of green-up dates. In one study from Rhode Island, researchers using Landsat data were able to incorporate land- scape features into their analysis and detect a delay in leaf out at the base of hills due to cold air drainage, a delay in coastal areas due to the cooling effects of the ocean, and a one-week delay in leaf out for forests in rural areas compared to those in the nearby urban area of Provi- dence (Fisher et al. 2006). While such changes in leaf-out dates are already known from specific ground observations, the abil- ity to detect such effects using remote sensing greatly extends 0 50 100 200 Km our ability to map leaf out over The average onset of leaf out in (A) southern New England from Landsat (1984– large areas. 2002) and (B) the northeastern using MODIS (2000–2005). These Another interesting remote images demonstrate that later phenology occurs at higher elevations, such as the sensing approach for monitoring Adirondacks and White Mountains; at higher latitudes; and in coastal areas that leaf out uses phenocams, which experience moderating ocean effects, such as Cape Cod and the Islands. The refers to cameras placed in fixed and New York metropolitan areas leaf out earlier because of higher temper- atures associated with the urban heat island effect; earlier leaf out is also seen in locations that are used to record warm river valleys. Colors indicate the date on which half of the tree canopy has images of the leaf canopy at leafed out (from day 110 [April 20] to day 165 [June 15]), with earlier onset shown regular intervals, such as every by blue and later onset by and red. (image from fisher and mustard 2007) hour or once a day, throughout Leaf-out Dates 17 a r d Fo e st e s y o f John O ’ K ee and t he H a rv C ou rt

A sequence of two photos taken at the same spot by a pheno- cam at the Harvard Forest showing leaf out over a one-week period. The picture on the top left was taken on April 30, 2009, the one on the top right was taken on May 7, 2009. The photograph on the bottom was taken on April 29, 2009 and shows a view over the canopy of the Harvard Forest with the phenocam below.

the growing season. Dr. Andrew Richardson and others have set up phenocams in the can- opy at Harvard Forest and 11 other forests in the northern continental United States. These images can be analyzed using computer pro- grams to determine the seasonal trajectory of budburst, green-up, and senescence. Networks of these phenocams can fill in the spatial and temporal gaps between plant monitoring by human observers and regional remote sensing have advanced their leaf out images. Seven of these sites in the United States by an average of one week over the past fifty also have towers that monitor the exchange of years (Menzel 2000). In , woody plants

(CO2) and water between the such as forsythia (Forsythia viridissima var. atmosphere and the forest. This combination of koreana), ginkgo, mulberry ( bombycis), data from webcams, satellites, and gas sensors and various species (Prunus spp.) leafed is providing crucial information on the relation- out an additional 2 to 7 days earlier for each ship between phenology and ecosystem pro- 1°C increase in temperature between 1953 and cesses, especially carbon uptake (Richardson 2005. At a few sites, however, ginkgo trees were et al. 2009a). actually leafing out later than they did in the past, contrary to expectations (Ibanez et al. 2010). Leaf out and climate change At the Hubbard Brook Experimental For- Climate change is already affecting many eco- est in , the leaf out onset of logical processes, and leaf out is no exception three native species—American beech (Fagus (Ibanez et al. 2010; Menzel 2000; Richardson et grandifolia), sugar maple (Acer saccharum), al. 2006). By analyzing long-term data on leaf- and (Betula alleghaniensis)— out dates, much can be learned about how the has advanced an average of 5 to 10 days over onset of spring has changed over time as tem- the past five decades (Richardson et al. 2006). peratures have increased. From data collected Historical datasets, such as those recorded by at the IPG, researchers determined that trees in Henry David Thoreau and Aldo Leopold, can 18 Arnoldia 68/4

also be used for these types of studies by com- In eastern Massachusetts, species leaf out paring their records to contemporary observa- over a 4 to 6 week period. Among the first plants tions from the same place, even if there is a lack to leaf out in the spring are such introduced of data between the two time periods (Bradley ornamental shrubs as common lilac (Syringa et al. 1999; Miller-Rushing and Primack 2008). vulgaris), honeysuckles (Lonicera spp.), and Jap- We know that leaf out has become earlier anese barberry (Berberis thunbergii), and non- in many areas in recent years, largely because native fruit trees such as apple ( spp.). Of of warmer temperatures, but what about the native species, meadowsweet (Spiraea alba var. future? Will the advance in spring’s onset latifolia), quaking ( tremuloides), continue for all species? To answer these ques- black cherry (Prunus serotina), and grey birch tions, it is necessary to both be familiar with (Betula populifolia) are among the first species the physiology behind leaf out, and to build on to leaf out. Consistently among the last species what we already know about the response of to leaf out are white ash ( americana), leaf out to temperature. white oak, and black (Nyssa sylvatica), with poison sumac ( vernix) and Variation in leaf-out times buttonbush (Cephalanthus occidentalis) often Trees and shrubs vary widely in leaf-out times, being the last of all. The pattern of leaf out is both among and within species. For instance, fairly consistent across the temperate zone of individuals will leaf out earlier in a warm, Europe and North America. Certain groups of sunny location, such as a south-facing hill, plants tend to leaf out early (birches, willows, than individuals of the same species located in , many poplars and ) and others late a cold, shady location. Similarly, all individu- (hickories, walnuts, and ashes). als of a given species will leaf out later during a So why do some species leaf out so early and cold spring than in a warm spring. Sometimes other species leaf out so late? Since the func- when a tree is growing on the edge of field, the tion of leaves is to carry out photosynthesis and exposed sunny side will leaf out earlier than the provide sugars for the tree, in general it should shady side. These differences aside, there is a benefit a tree to leaf out as early as possible to fairly consistent pattern in the leaf-out timing get the longest growing season. A tree species of trees, shrubs, and vines from year to year. that leafs out in early April has four additional weeks to photosynthesize compared to a tree species that leafs out in early to mid May. How- ever, the early-leafing tree faces the danger of a late frost that will kill its leaves and damage r Maye Ro b e rt its vessel elements, the chief water conducting tissue. This trade-off between the advantages of early growth and of late growth provides a good explanation of why certain species leaf out when they do. The stem anatomy supports this explanation, with early species tending to have smaller vessel elements that are less prone to frost damage than the larger vessel elements of later species (Lechowicz 1984; Miller-Rushing and Primack 2008). Also important is the evo- lutionary history of a plant group: If it orginated in a warmer climate, it may not have fully- adapted mechanisms for dealing with extreme cold and therefore may have different factors regulating leaf out than a plant group originat- Buttonbush (Cephalanthus occidentalis) is one of the ing in a colder climate. latest native shrubs to leaf out in the spring. It is seen The vulnerability of trees and other plants to here blooming in midsummer. frost damage was recently demonstrated when Leaf-out Dates 19 t e U ni v rs i y a a S t I o w

A late frost killed the new foliage on this oak tree. two weeks of abnormally warm weather in ally a minimum of 0 to 10°C [32 to 50°F]) days March 2007 triggered early leaf out all across in winter are required before the are able eastern and central North America. A return of to break dormancy. The exact number of chill- freezing weather from April 5 to 9 killed young ing units required depends both on species and leaves and flowers, and caused the die back of on the weather of the preceding growing season tree canopies across the region (Gu et al. 2008). (Hunter and Lechowicz 1992; Perry 1971). Once This frost damage was an example of the type this requirement has been fulfilled, a certain of episodes of mismatches between plants and number of warm days above a certain tempera- climate that may become increasingly common ture threshold are then needed for leaf develop- as climate change continues. ment to begin and buds to open. This pattern is seen in both deciduous and evergreen species. What triggers leaf out? In addition, some species also have a photope- Leaf out is predominantly controlled by temper- riod requirement, meaning that they will only ature, with plants generally leafing out earlier leaf out once daylength reaches a certain num- in warmer conditions, but warm temperature ber of hours in the spring. In particular, long- is the not only factor. In fact, for many species lived trees of mature forests, such as American it is a combination of warm and cold tempera- beech, some oak species, and hackberry (Celtis tures along with day length that dictates when occidentalis), often rely on a combination of the leaves will emerge from the bud. Most photoperiod and temperature cues to break dor- temperate species, including sugar maple and mancy. For these species, budbreak only occurs quaking aspen, have a chilling requirement, after specific photoperiod and temperature meaning that a certain number of cold (gener- requirements have been met. This holds even 20 Arnoldia 68/4

when individuals from these species are planted of leaf-out dates. If certain early successional in subtropical climates with exceptionally high species with minimal photoperiod and chilling temperatures (Korner and Basler 2010). requirements continue to leaf earlier in the In contrast, many opportunistic species that spring, they may increase their abundance and are found early in forest succession, such as distribution to become the dominant species, birches, hazelnuts (Corylus spp.), and poplars, and shift the leaf-out time of the whole forest. do not have a photoperiod requirement to The unmet chilling and photoperiod require- break winter dormancy. This somewhat risky ments of other species may significantly slow strategy allows trees to respond more quickly the advance of leaf out at the whole forest to episodes of warm temperature in the early level. These two scenarios have consequences spring, but also creates more susceptibil- for many ecosystem processes, including the ity to late frosts. Yet a third group of species, uptake of carbon dioxide, tree growth, forest which includes mostly ornamental plants from temperature, and water movement. warmer climates, has a leaf-out strategy linked The earlier leaf-out times of many escaped to spring temperature with minimal chilling ornamental shrubs, such as Japanese barberry requirements and no photoperiod requirement. and several honeysuckle species, may help to The common lilac is a local example, and is one explain why these species are increasing so of the first plants to leaf out each spring. greatly in abundance in our forests. Their ear- lier leaf-out times may give them a competi- Learning from the past, predicting the future tive advantage over native species with more Using information about past phenological restrictive requirements for leaf out (Willis et responses to temperature and future climate al. 2010). scenarios, scientists can develop models to pre- dict future phenological changes both at the spe- Ecological interactions cies and ecosystem levels. One modeling study The onset of spring affects not only plants and found that the advance in leaf-out time for most ecosystem processes, but also organisms that species and places is likely to continue in com- depend on those plants. Leaf-out timing deter- ing decades as the climate continues to warm mines the availability of food and shelter for (Morin et al. 2009). Many temperate tree species many species, particularly insects. This timing will show large advances in leaf out at higher is especially important for species that have latitudes, including the northern United States gone through a long winter with little avail- and Canada. Delays in leaf out, or abnormal able food, or for bird species completing an leaf-out events, could occur at the southern end energy-demanding migration north. From his of species ranges in the southern United States close observations of nature in Concord, Henry for some species including black ash (Fraxinus David Thoreau was aware of the ecological nigra) and sugar maple if those species fail to importance of the emergence of leaves in the meet their winter chilling requirement. Species spring, writing in 1854: “To-day the air is full with photoperiod requirements are also unlikely of birds; they attend the opening of the buds. to continue to show linear advancements in The trees begin to leaf, and the leaf-like wings leaf-out dates with increasing temperatures of birds are in the air. The buds start, then the since photoperiod will not change. insects, and then the birds.” Because there are a host of complicated fac- Thoreau was aware of order of events based on tors involved in leaf-out phenology, it is hard what he had experienced in previous years and to predict whether leaf out will continue to took for granted that the same pattern would advance linearly with changes in tempera- persist, even with the large inter-annual varia- ture at the whole forest level. The possibil- tion in weather. In the twenty-first century, we ity of shifts in species composition resulting can no longer take for granted that this order of from climate change, as some species expand natural events will continue each year. While their range and others contract theirs, adds yet plants are responsive to changes in tempera- another layer of uncertainty to the prediction ture, other organisms that interact with plants Leaf-out Dates 21 Richa r d P i m ack

Clockwise from top left: Photos of the Old North Bridge in Concord, Massachusetts, showing the development of the leaf canopy in the spring of 2010 (April 15, April 20, May 3, May 13). The meadow in the foreground is flooded in the first two photos, and dried out in the second two photos.

in an ecosystem may not be quite so quick to If trees are leafing out earlier in the spring and respond. For instance, while certain species of dropping their leaves later in the autumn, they birds arrive earlier in warmer years, other birds are likely increasing the net amount of carbon do not change their arrival dates, and some spe- being sequestered in biomass. This possibility cies are even arriving later (Miller-Rushing et is supported by the work done by Andrew Rich- al. 2008). Insects are involved also: If certain ardson and others at the Harvard Forest and the kinds of insects feed only on the young leaves of Howland Forest in , showing that the a particular plant species that are present for a earlier onset of spring in New England results limited time in the spring, those insect species in an increase of carbon sequestered in decid- may decline in abundance if they emerge too uous forest and somewhat less in coniferous early or late in the spring relative to their food forest (Richardson et al. 2009b). If this is occur- resource. Birds that depend on those insects for ring over a large area and over many years, the food may similarly decline in abundance. longer growing season could allow temperate Much more work is needed to understand how forests to withdraw more carbon dioxide from climate change and rising carbon dioxide con- the atmosphere. The longer growing season centrations are affecting ecosystem processes. and warmer temperatures could also mean that 22 Arnoldia 68/4

trees are losing more water vapor to the atmo- timing of budburst in temperate trees. Journal sphere during the process of photosynthesis; in of Applied Ecology 29: 597–604. consequence, forest ecosystems could possibly Ibanez, I., R. B. Primack, A. J. Miller-Rushing, E. Ellwood, H. Higuchi, S. D. Lee, H. Kobori et al. 2010. release less water to streams and aquifers, with Forecasting phenology under global warming. major implications for drinking water supplies, Philosophical Transactions of the Royal Society flood control, and ecology of aquatic organisms. B-Biological Sciences 365: 3247–3260. Korner, C., and D. Basler. 2010. Phenology under global Conclusion warming. Science 327: 1461–1462. Throughout the world, forests are being altered Lechowicz, M. J. 1984. Why do temperate deciduous trees leaf out at different times? Adaptations in many ways by the rising temperatures associ- and ecology of forest communities. American ated with global climate change, and the earlier Naturalist 124: 821–842. leaf-out dates of trees and shrubs is one such Menzel, A. 2000. Trends in phenological phases in Europe example. Earlier leaf-out dates are expected to between 1951 and 1996. International Journal continue in coming decades across much of of Biometeorology 44: 76–81. North America. Over a longer period of time, Miller-Rushing, A. J., T. L. Lloyd-Evans, R. B. Primack, and P. Satzinger. 2008. Bird migration times, many tree species will likely be extirpated on climate change, and changing population sizes. a local scale and shift their ranges in response Global Change Biology 14: 1959–1972. to the changing climate. Because there is a Miller-Rushing, A. J., and R. B. Primack. 2008. Global wide variation among species in leaf-out times, warming and flowering times in Thoreau’s changes in the species composition of a forest concord: A community perspective. Ecology 89: 332–341. will also mean changes in leaf-out dates at the Morin, X., M. J. Lechowicz, C. Augspurger, J. O’ Keefe, D. level of the whole forest. Disentangling the sep- Viner, and I. Chuine. 2009. Leaf phenology in arate effects of changing species composition 22 North American tree species during the 21st and changing climate is one of the great chal- century. Global Change Biology 15: 961–975. lenges of detecting leaf-out trends using remote O’Keefe J. 2010. Phenology of Woody Species. Harvard sensing. Botanical gardens such as the Arnold Forest Data Archive: HF003. Arboretum can contribute to these efforts by Perry, T. O. 1971. Dormancy of trees in winter. Science 171: 29–36. quantifying the differences among species in Richardson, A. D., A. S. Bailey, E. G. Denny, C. W. Martin, leaf-out dates for trees, shrubs, and vines all and J. O’Keefe. 2006. Phenology of a northern growing at one location, which for many spe- hardwood forest canopy. Global Change Biology cies is outside of their native range. Such infor- 12: 1174–1188. mation can then aid in calibrating leaf-out dates Richardson, A. D., B. H. Braswell, D. Y. Hollinger, J. P. over a large area using remote sensing. Jenkins, and S. V. Ollinger. 2009a. Near-surface remote sensing of spatial and temporal variation in canopy phenology. Ecological Applications References 19: 1417–1428. Bradley, N. L., A. C. Leopold, J. Ross, and W. Huffaker. Richardson, A. D., D. Y. Hollinger, D. B. Dail, J. T. Lee, 1999. Phenological changes reflect climate J. W. Munger, and J. O’Keefe. 2009b. Influence change in Wisconsin. Proceedings of the of spring phenology on seasonal and annual National Academy of Sciences of the United carbon balance in two contrasting New England States of America 96: 9701–9704. forests. Tree Physiology 29: 321–331. Fisher, J. I., J. F. Mustard, and M. A. Vadeboncoeur. 2006. Willis, C. G., B. R. Ruhfel, R. B. Primack, A. J. Miller- Green leaf phenology at Landsat resolution: Rushing, J. B. Losos, and C. C. Davis. 2010. Scaling from the field to the satellite. Remote Favorable Climate Change Response Explains Sensing of Environment 100: 265–279. Non-Native Species’ Success in Thoreau’s Fisher, J.I. and J.F. Mustard. 2007. Cross-scalar satellite . PLoS ONE 5(1): e8878. doi:10.1371/ phenology from ground, Landsat, and MODIS journal.pone.0008878. data. Remote Sensing of Environment 109: 261–273. Caroline Polgar is a graduate student at Boston University, Gu, L., P. J. Hanson, W. Mac Post, D. P. Kaiser, B. Yang, where Richard Primack is a professor. For the past eight R. Nemani, S. G. Pallardy et al. 2008. The 2007 years, Richard Primack and his students have been eastern US spring freezes: Increased cold damage investigating the impact of climate change on the plants in a warming world? Bioscience 58: 253–262. and animals of Massachusetts, with much of the focus at Hunter, A. F., and M. J. Lechowicz. 1992. Predicting the the Arnold Arboretum and Concord.