American Scientist the Magazine of Sigma Xi, the Scientific Research Society

A reprint from American Scientist the magazine of Sigma Xi, The Scientific Research Society

This reprint is provided for personal and noncommercial use. For any other use, please send a request to Permissions, American Scientist, P.O. Box 13975, Research Triangle Park, NC, 27709, U.S.A., or by electronic mail to [email protected]. ©Sigma Xi, The Scientific Research Society and other rightsholders The Mystery of Masting in Trees

Some trees reproduce synchronously over large areas, with widespread ecological effects, but how and why?

Walter D. Koenig and Johannes M. H. Knops

his autumn, the vast hardwood People have witnessed masting since begin reproducing in the fall—one of a Tforests of North America could be time immemorial; after all, it’s pretty ob- handful of temperate resident species chock full of acorns, a pulse of resources vious when you’re tripping over nuts that will sometimes breed then—and that will cascade through the ecosystem, strewn all over the path some years and resume the following spring. In poor affecting mice, birds and other wildlife you find barely one or two in others. In years, by contrast, birds may be forced for years to come. Over far-flung areas agrarian communities, farmers would to abandon their territories in order to virtually all the oaks of the same species, have been especially aware of the boom search for food elsewhere. The associ- and perhaps more than one species, are and bust cycle because they fed acorns to ated risk of moving into unfamiliar areas already gearing up to produce the seed livestock. Although records are sparse, and consequent starvation results in se- crop of the decade. Or not. In fact, there historical economic data have been used vere population crashes. may be almost no acorns, and a substan- to infer the size of acorn crops from as Direct effects of the variable acorn tial proportion of the wildlife will starve far back as the mid-17th century. Only crop are only the beginning. At the Insti- or have to eat something else. recently, however, have scientists begun tute of Ecosystem Studies in New York, Such highly variable and synchro- to recognize the geographic scope and Richard S. Ostfeld and Clive G. Jones nized reproduction is known as “mast- ecological consequences of masting and have unraveled a series of ecosystem- fruiting” or “masting.” The term comes attempt to explain how this remarkable wide domino effects caused by mast- from the Old English word, mæst, for phenomenon is achieved. So how do all ing. In eastern North America, where nuts of forest trees that have accumu- those millions of trees do it? hardwood forests dominate, large acorn lated on the ground, especially for those In this article, we describe the eco- crops increase populations of mice and used as food for fattening swine. Evolu- logical and economic consequences of deer, two major acorn predators. In turn, tionarily, a significant selective benefit of masting and discuss what causes seed populations of ticks, major ectoparasites masting is “predator satiation.” The idea production to vary so widely. This leads of mice and deer, increase. And again, is that large crops satiate seed eaters so us to review the latest research into pro- since ticks cause Lyme disease, the inci- that some seeds escape being eaten, par- posed mechanisms that govern the syn- dence of Lyme disease in human popu- ticularly in “mast” years with bumper chronous production of seed across wide lations rises, as reported by Ostfeld in crops. To reinforce this effect, small crops geographic areas. Finally, we speculate American Scientist (July–August 1997). keep seed-predator populations so low on the admittedly uncertain effects of Mice also plunder the nests of local bird that there are too few animals to eat all global climate change on ecosystems in species, particularly those that nest on or the seed produced during good years. which masting plays a crucial role. near the ground, and high numbers of Thus, a higher proportion of seeds over- them correlate with decreases in popula- all ultimately escape predation. Mighty Ecosystems from Little Acorns tions of species such as the veery (Catha- Mast crops produce resource pulses that rus fuscescens) and several species of war- Walter D. Koenig received his Ph.D. in zoology in can broadly affect the local ecosystem blers, typically two years following mast 1978 from the University of California, Berkeley, in which they happen. Along the West events. But the effects of big acorn crops where he is currently an adjunct professor in the Coast of North America, variable acorn are not all bad. High mouse densities Department of Integrative Biology and a research crops may be the primary determinant following large acorn crops tend to keep zoologist at the Museum of Vertebrate Zoology. His of increases and decreases in the popula- populations of their major food items in research interests include social behavior of birds, re- tions of certain birds. For example, acorn check. Among the most commercially productive traits of forest trees and patterns of spatial woodpeckers (Melanerpes formicivorus) important is the gypsy moth (Lymantria synchrony in nature. Johannes M. H. Knops received depend almost entirely on acorns for dispar). This pest defoliates a wide vari- his Ph.D. in botany in 1994 from Arizona State food. Uniquely among birds, acorn ety of valuable species, costing an aver- University and is currently an associate professor in woodpeckers store acorns, often by the age $30 million yearly, according to the the School of Biological Sciences at the University of Nebraska. He studies ecosystem nutrient cycling and thousands, in holes they drill in special USDA Forest Service. In short, the pulse how biodiversity relates to ecosystem function. Ad- “storage trees” or “granaries” located in of resources dumped into the ecosystem dress for Koenig: Hastings Natural History Reserva- their territory. In years of bumper crops, during a mast year leads to a “trophic tion, 38601 E. Carmel Valley Road, Carmel Valley, acorn woodpeckers survive well dur- cascade” in which masting affects the CA 93924. Internet: [email protected] ing the winter. Moreover, they tend to numbers of organisms in various levels

© 2005 Sigma Xi, The Scientific Research Society. Reproduction 340 American Scientist, Volume 93 with permission only. Contact [email protected]. es Nationaux/Art Resource Nationaux/Art es é union des Mus des union é R Figure 1. Oak trees have long provided sustenance for people and their livestock. During a mast year, entire populations of oaks produce huge crops of acorns, and farmers have taken advantage of this bounty since at least medieval times. Painted in 1416, this picture for November from the calendar section of the Très Riches Heures, by the Limbourg brothers from Flanders, shows a peasant throwing a stick to knock down acorns on which his pigs are feeding. Only recently have scientists begun to understand the ecological significance and underlying biology of masting. of the food chain. The resulting ripple this research, recent work has investi- trees: variability and synchrony. That is, effect can influence community compo- gated mast-fruiting in plants at both the trees must synchronize the quantity and sition for years following the event. individual and population levels. Indi- timing of seed production. In this way, vidual plants do not mast. Masting is a by mechanisms that remain poorly un- Top Seed group phenomenon that results when derstood, the masting trees apparently Ostfeld and Jones’s studies demon- plants within a population synchronize come to a consensus as to how bounti- strate that highly variable seed crops their reproductive activity. Thus, mast- ful a crop to produce in a particular have important and long-lasting effects ing takes place as a result of two sepa- year. Let’s look at some ideas of how within their communities. Parallel to rate but related features of reproducing this might come about.

© 2005 Sigma Xi, The Scientific Research Society. Reproduction www.americanscientist.org 2005 July–August 341 with permission only. Contact [email protected]. curve, whereas the annual size of the tioned earlier, predator satiation is an seed crop is not. Also, rainfall one year effective way a tree can keep predators provides no clue as to how much rain- from eating all its seeds. Predator sa- fall there will be in any subsequent year. tiation is an example of “economy of That is, environmental variables usually scale,” to use the terminology of David show no consistent patterns of temporal A. Norton and Dave Kelly of the Uni- autocorrelation, when later measures in versity of Canterbury in New Zealand. a time series are correlated with earlier The idea is that big investments (large measures. seed crops) are more efficient than small In contrast, Victoria Sork, now at the crops because each seed has a higher University of California, Los Angeles, probability of escaping predation. Thus, and her colleagues have shown that it’s more economical for the tree over seed production by individual oak the long run to produce an occasional trees in Missouri is highly autocorre- bumper crop than to produce the same Figure 2. Blue oaks (Quercus douglasii) are lated over time; a good crop one year size crop each year, even though in the classic masting species, producing a highly is generally followed by a relatively other years it produces few or no seeds variable seed crop from one year to the next poor crop the next, a pattern found in that survive. that is synchronized among individuals over several oak species. Thus, knowing the Another challenge for trees is to maxi- a wide area. In a mast year, a mature blue oak may produce more than 100,000 acorns— size of a tree’s crop this year gives you mize pollination efficiency. This is espe- more than 10 times the annual average—and information about how many acorns cially true for wind-pollinated trees that is likely to produce few or no acorns in a it is likely to produce in subsequent are primarily outcrossing (rather than poor year. (Except where noted, all photo- years. This finding suggests that mast- self-fertilizing), such as oaks, pines and graphs courtesy of the authors.) ing trees put so much energy into pro- the majority of tree species in the tem- ducing a good seed crop that they are perate and boreal regions of the North- First, what might cause variability in forced to forgo reproduction in at least ern Hemisphere. This consideration seed crops? One idea is the “resource one subsequent year. contrasts wind-pollinated species with tracking” hypothesis. This mechanism The growth-reproduction tradeoff is animal-pollinated species, as the latter requires that the number of seeds pro- a second line of evidence that masting risk lowering their reproductive success duced each year varies in direct response cannot be fully explained by resource by satiating their pollinators if too many to the amount of resources available to tracking. To evaluate trade-offs more flowers are produced at once. Conse- the tree. For example, warm, wet years precisely we have measured annual quently, masting is expected to be more might provide trees more resources to growth and reproduction in blue oaks widespread among wind-pollinated spe- grow seeds by enhancing photosynthesis (Quercus douglasii) at Hastings Reser- cies. From an evolutionary standpoint, a and increasing the availability of soil nu- vation in central coastal California. To tree should not produce huge quantities trients. In contrast, cold, dry years would quantify growth we use dendrometers, of pollen or flowers that go to waste. By result in less photosynthesis and lower which are stainless-steel bands placed flowering at the same time, trees maxi- nutrient availability, resulting in meager around the trunk. Yearly acorn counts mize their chances of pollination and seed crops. Resource tracking could also have provided estimates of reproductive minimize waste. account for the geographic synchrony effort. The data show that trees tend to The extent to which masting succeeds of seed production observed in masting grow slowly the year that they produce at enhancing pollination is not entire- species, since rainfall and temperature a large acorn crop, but usually grow well ly clear. Pollen is frequently limited in tend to fluctuate in the same direction the next year, after which they will again from one year to the next over large ar- grow relatively poorly. eas. Most important, if resource tracking Producing a large seed crop during can account for masting behavior, then a masting episode apparently takes a evolutionary explanations for masting lot of resources and has long-term con- would not be necessary. Instead, masting sequences for growth and subsequent could merely be a side effect of resources reproduction by individual trees. That that vary from year to year. is, the trees shunt an inordinate amount However, the evidence does not sup- of resources into flower and seed pro- port a simple view of resource tracking duction during mast years, while shift- as a cause of masting. For example, an- ing resources back into growth during nual fluctuations in rainfall and tem- poor seed-crop years. Such switching perature are smaller in magnitude than of resources proves that masting must /VIREO

those of crop sizes of masting species, be an adaptation—an evolutionary re- rette although resource tracking predicts they sponse by trees to some environmental should be at least as great. Moreover, challenge. yearly variation in these environmen- Among the most severe selective pres- tal quantities usually exhibits statistical sures a plant faces is having its seeds eat- Tour La Peter © characteristics distinctly different from en. An adaptation that minimizes such Figure 3. The acorn woodpeckers (Melanerpes those of mast crop sizes. Specifically, an- consumption will be strongly favored formicivorus) is one of many wildlife species nual rainfall and temperature tend to be by natural selection because it directly whose population dynamics are affected by distributed normally, following a bell improves reproductive success. As men- the size of the acorn crop.

© 2005 Sigma Xi, The Scientific Research Society. Reproduction 342 American Scientist, Volume 93 with permission only. Contact [email protected]. animal-pollinated species, but scientists have generally thought that pollen was �� for all practical purposes unlimited in �� wind-pollinated species. From this, it �� was assumed that increased pollination � � �� efficiency was unlikely to play an impor- � tant role in natural selection for masting. �� However, this premise has been over- turned in the last several years. Evidence � that wind-pollinated species are indeed pollen-limited has begun to accumulate from field experiments. These data are �� consistent with conclusions from wind- � pollination models. Theoretical work, largely by Akiko Satake and Yoh Iwasa of Kyushu University in Japan, demon- strates that pollen coupling, the depen- �� dence of individual trees on pollen from members of the same species, can be suf- ficient to drive the highly variable seed production that is characteristic of mast- � �� ing species. ��

Going to Seed But how do trees synchronize flowering, �� and hence their production of seeds? It is �� one thing for trees to produce more seeds some years than others, but how do they manage to coordinate their activities, � �� given that masting involves switching � �� resources between growth and reproduction? For that matter, how many trees are � involved, and over how large an area? � Answering the last question requires accumulating data on a large geographic scale over many years, so it’s not surpris- Figure 4. Multiple ecological interactions between species in a model Eastern deciduous for- ing that no one has attempted to address est result in a trophic cascade in which the effects of acorn crop size ripple through the food it until recently. Our own work has ap- web, setting off a chain of events that cause animal populations to change. Large amounts of energy (arrows) flow through the system during a mast year (right) when a large acorn crop proached the problem in two ways. First, is produced by an oak (center), causing populations of deer, mice and ticks to increase and we have analyzed data extracted from raising the incidence of tick-borne Lyme disease (+). Populations of ground-nesting birds and the literature, and second, we have col- gypsy moths decrease (–) owing to predation by mice. During a mast failure (left) effects of an lected field data from California oaks. acorn shortage on the food web are negative for acorn-eaters and their predators and positive The literature search involved lots of for birds and moths, whose eggs and larvae are eaten by mice.

��

� � � � � � � ��

Figure 5. According to the resource tracking hypothesis, environmental fluctuations cause seed crops to vary. If these fluctuations account for masting, theoretical frequency distributions for environmental measures and crop size are both expected to be normal, following a bell curve (left). Ac- tual data (right) for environmental fluctuations (annual mean temperature) are distributed this way. Crop-size data (annual acorn crop) are not. The difference implies that resource tracking alone cannot account for variation in seed production and that other factors initiate and regulate masting.

© 2005 Sigma Xi, The Scientific Research Society. Reproduction www.americanscientist.org 2005 July–August 343 with permission only. Contact [email protected]. Figure 6. Newly emerged catkins of valley oak will soon begin to shed Figure 7. Dendrometers, such as this one on a blue oak, facilitate pollen. Pollination efficiency increases during a good acorn year, offering precise measurement of a tree’s radial growth. A stainless steel a selective advantage for masting in addition to predator satiation—when band held together by a spring expands along with the tree. The so many acorns are produced that some escape being eaten. expansion is measured using notches on the band. library time, but yielded results much spruces and pines. The data set repre- or to get total acorn weights. Because faster. We found more than 400 rele- sents 72 Northern Hemisphere sites and the estimate includes acorns in the vant publications on various tree spe- includes more than 5,000 data points. crown, this technique also minimizes cies, including hardwoods such as oaks Each point corresponds to one species’ under-sampling caused by not count- and beeches, and conifers such as firs, seed crop from one site for a particular ing acorns that are eaten by arboreal year. We crunched the numbers and seed eaters. Such bias is a significant �� found strong evidence for spatial syn- source of error in counts from traps chrony in seed production among sites. that catch fallen acorns. �� Results of the meta-analysis indicate From 19 sites scattered throughout �� that conifer genera at sites as far as 2,500 California, our data currently encom- kilometers apart spatially synchronize pass more than a thousand trees divided �� their seed production. This implies that among 50 populations of nine oak species trees are synchronizing reproductive (eight Quercus species and the tanbark �� activities over an area of at least six mil- oak, Lithocarpus densiflorus). After 11 years �� lion square kilometers. Even the switch- we’re at last getting a picture of the ex- ing of resources between reproduction tent to which California oaks synchronize �� and growth appears to be coordinated their reproductive activities. over this vast area—crop sizes and ra- The results vary among species but �� dial growth are inversely correlated clearly show considerable synchrony. �� over distances of up to 1,000 kilometers. Consider blue oaks, which are one of �� (Reproduction was inferred with the the most extensively distributed trees �� data from our literature search, and rain California. They grow in foothill re- �� dial growth was measured by tree-ring gions around the Central Valley over an �� chronologies of conifers obtained from area of some 20,000 square kilometers, �� the International Tree-ring Databank.) about five percent of the state’s total land �� A longer-term undertaking has been area. Eleven years of data from 10 sites Figure 8. Correlation between reproduction the gathering of field data on the spa- between 80 and 700 kilometers apart (measured by annual acorn surveys) and tial synchrony of seed production. To demonstrate strong synchrony between concurrent growth in blue oaks is negative, accomplish this, we initiated the Cali- sites. In the absence of synchrony, we reflecting the trade-off between these two fornia Acorn Survey in 1994. Each Sep- would expect the average pairwise cor- biological imperatives. That is, when the re- tember we spend ten days counting relation in the mean acorn crop between sources invested in reproduction are signifi- acorns and driving 3,500 kilometers sites, r, to approach zero; with perfect cant, growth tends to be small, and vice versa. around the state. To quantify the acorn synchrony, it would approach 1. In our Interestingly, if one compares reproduction crops, we scan the crowns of each data, r averages about 0.6. It decreases in one year with growth in the previous year marked tree, counting as many acorns slightly as sites get farther apart but re- and calculates a correlation coefficient using observations from all the survey years for as possible for 30 seconds. Since a larg- mains well above zero even between which such pairings are available, a positive er crop will result in a higher density of the sites farthest apart. In other words, value results, indicating that these trees tend acorns, this method is an efficient way synchrony in acorn production extends to switch between masting and growing from to sample relative crop sizes, rather to pretty much every blue oak, a popula- one year to the next. than attempting to count every acorn tion of 100 to 200 million individuals.

© 2005 Sigma Xi, The Scientific Research Society. Reproduction 344 American Scientist, Volume 93 with permission only. Contact [email protected]. Tree’s Company that the environmental cue is spatially synchronous on large geographic scales. Scientists recognize at least three possible synchronous, the trees would then be Synchrony does decline with distance mechanisms by which trees synchronize synchronized as well. but remains statistically correlated up seed production: chemical signaling, re- The principle of synchronization of to distances of thousands of kilometers. sponse to environmental cues and pol- population dynamics by environmental To drive the spatial synchrony observed len coupling. In the first case, a chemi- fluctuations is termed the “Moran ef- in masting, such environmental factors cal signal could transmit the necessary fect,” from work in the 1950s by Patrick must be synchronized at least as closely information from one tree to the next. Moran, an Australian statistician. Mo- as the trees’ reproduction. That is, plenty Molecules might travel via a physical ran showed that an external factor, such of confounding influences can and prob- connection or through the air. Herbivory as weather, that acts across separated ably do reduce reproductive synchrony is known to induce chemical signals in populations with similar ecophysiol- below that of the critical environmental some plants. Combined root and fun- ogy would tend to produce correlated factor (whatever it may be). But, in order gus structures share fungal filaments, or changes in their abundance and hence for the Moran effect to be the driving mycelia, that pass nutrients from one tree synchronize their population cycles. The force, spatial synchrony in the environ- to the next, providing a potential signal- Moran effect has since been shown to mental factor must be at least as great ing mechanism. However, both myce- generate spatial synchrony in the pop- as that observed in the character under lial connections and airborne chemicals ulation dynamics of a wide variety of consideration. would facilitate synchrony only among organisms. Well-established examples Preliminary evidence suggests that trees that were in close proximity, and include human viral pathogens, sheep this is indeed true in the case of masting it’s unlikely they could work effectively on the St. Kilda archipelago and cari- by California oaks. In California, for ex- across a large area. These mechanisms bou in Greenland. The Moran effect is ample, mean temperature in April is spa- are especially implausible given that behind spatial synchrony in many bird tially synchronized more strongly than many individuals are in fragmented populations, and there are good reasons acorn production by blue oaks. (It has populations, and therefore out of contact to suspect it plays a crucial role in syn- a higher correlation coefficient among with trees in other parts of the range. chronizing masting behavior as well. sites over similar distances, as shown If chemical signaling is inadequate to Several weather phenomena have ef- in Figure 11.) This is consistent with the explain masting, how could masting be fects on wide geographic scales; some hypothesis that trees use fluctuations in mediated by an environmental factor approach the global level. The best- temperature as the cue for whether to such as rainfall, mean spring tempera- known is the El Niño (more properly invest their resources in growth or repro- ture, or perhaps even sunspot activity? called the El Niño/Southern Oscillation, duction that year, and that such fluctua- This may sound rather like a reworking ENSO). This climatic anomaly starts tions result in geographic synchrony of of the discredited resource tracking hy- with unusually warm temperatures in masting in this species—a Moran effect. pothesis, but the basic idea is different. the central and equatorial eastern Pacific What is the biological basis for focus- In this instance, the amount of resources Ocean that cause changes in ocean cur- ing on mean April temperatures as a devoted to reproduction does not de- rent direction, precipitation and air tem- synchronizing cue? This is when blue pend on metabolic processes limited by peratures across a significant proportion oaks in California flower and when pre- the environment, but rather on the trees’ of the globe. vailing conditions appear to influence use of the environment as a stimulus But spatial synchrony is not limited to the number of flowers that get fertilized that provides information about how sporadic patterns; even everyday tem- and produce acorns. There are no ex- they will allocate resources. To the extent perature and rainfall can be spatially periments to demonstrate how the effect

��

Figure 9. The California Acorn Survey, begun by the authors in 1994, is an annual assessment of acorn production. More than 1,000 oaks at 19 sites (red dots) from across the range of oaks in the state (shaded area) are sampled. During the survey, one of the authors (Koenig) counts acorns through a pair of binoculars at the Santa Rosa Plateau Ecological Reserve, an undisturbed tract of Engelmann oaks (Quercus engelmannii), in Riverside County.

© 2005 Sigma Xi, The Scientific Research Society. Reproduction www.americanscientist.org 2005 July–August 345 with permission only. Contact [email protected]. further demonstrated that pollen cou- over a large enough area to isolate the pling, in the complete absence of envi- synchronizing effects of pollen cou- ronmental fluctuations, can potentially pling. However, it is possible to gauge synchronize reproduction among trees the environmental component, and in throughout an entire forest. The key is effect eliminate pollen coupling, by the degree to which individual trees looking at synchrony between species. depend on pollen produced by conspe- Assuming that the species do not usu- cifics a long distance away as opposed ally hybridize, then pollen coupling to nearby trees. This question has been cannot lead to interspecific synchrony addressed in terms of pollen flow—or between the two species, since pollen how far pollen moves—in wind-pol- being produced by one of the species linated species. cannot be having any influence on fer- Until recently, pollen dispersal dis- tilization efficiency or seed production tances were something of a mystery— of the other. This approach won’t work how can a microscopic pollen grain be in cases where inter-species hybridiza- tracked across a wide geographic area? tion is common, because the two spe- Scientists have worked on the prob- cies could be jointly “pollen coupled,” lem through a combination of field since pollen being produced by one Figure 10. Blue oak, pictured here in Lake studies on blue oaks by Eric E. Knapp, of the species could be fertilizing the County, is widely distributed in California. now at the U.S. Forest Service Pacific flowers of the other. If hybridization Spatial synchrony of acorn production in this Southwest Research Station, and mo- is absent, and if no interspecific syn- species appears to encompass its entire range lecular work on valley oaks by Peter chrony is found, it could be because of 20,000 square kilometers. E. Smouse at Rutgers and Victoria L. the biology underlying the population Sork. Their results suggest that the dynamics of the two species is not the might work, but warm, dry conditions mean distance of effective pollination same, violating the primary assump- during flowering do correlate with a (over which pollen actually fertilizes tion of the Moran effect. However, sig- larger acorn crop in several oak species, flowers that produce acorns) may be nificant interspecific synchrony pretty including valley oak (Quercus lobata) and on the order of 60 meters—surprising- much eliminates all potential mecha- blue oak in California and white oak (Q. ly short. If these findings prove gen- nisms for synchronization except for alba) in eastern North America. erally applicable to wind-pollinated the Moran effect. Ironically, the possibility that weather trees, they imply that pollen coupling As it happens, synchrony of masting conditions enhance pollination, leading could play a role in synchronizing between different species is relatively to greater seed production, suggests the trees over modest distances, perhaps common. For example, Eric M. Schau- third mechanism for how masting trees as much as a few kilometers. However, ber, now at Southern Illinois University, synchronize their reproduction: pol- they would rule out pollen coupling and Kelly studied masting in 17 New len coupling. As mentioned, enhanced as a mechanism for synchronizing re- Zealand tree and grass species. Schaub- pollination efficiency is an alternative productive timing over thousands of er and Kelly found that seed production hypothesis to predator satiation as a square kilometers. is synchronized over an area of some selective advantage of mast fruiting. Unfortunately, environmental fluc- 150,000 square kilometers. Moreover, However, Satake and Iwasa’s work has tuations cannot be physically excluded the masting is associated with unusu- ��

��

�� Figure 11. Constancy of the correlation coefficient of acorn production across distances up to 800 kilometers (left), as indicated by the shallow slope of the trend line, provides evidence for spatial synchrony of blue oak acorn production in California. (Data are based on the mean acorn crop of 10 sites surveyed annually since 1995.) How the trees coordinate their reproduction so closely is unknown, but the spatial synchrony of mean temperature in April (right) across sites is consistent with the possibility that synchrony in environmental fluctuations (known as the Moran effect) drives synchrony in seed production.

© 2005 Sigma Xi, The Scientific Research Society. Reproduction 346 American Scientist, Volume 93 with permission only. Contact [email protected]. �� unusually high summer temperatures Bibliography prompt heavy flower and fruit produc- Curran, L. M., I. Caniago, G. D. Paoli, D. Astianti, tion by several species of ecologically M. Kusneti, M. Leighton, C. E. Nirarita and �� dominant New Zealand Chionochloa H. Haeruman. 1999. Impact of El Niño and logging on canopy tree recruitment in Borneo. grasses, an event that governs popula- Science 286:2184–2188. tions of several insect seed-predator �� Elkinton, J. S., W. M. Healy, J. P. Buonaccorsi, species. Mark J. McKone at the Uni- G. H. Boettner, A. M. Hazzard, H. R. Smith versity of Canterbury and other New and A. M. Liebhold. 1996. Interactions among Zealand scientists have suggested that gypsy moths, white-footed mice, and acorns. �� Ecology 77:2332–2342.

�� ongoing global temperature increases Jones, C. G., R. S. Ostfeld, M. P. Richard, E. M. will cause considerable decreases in Schauber and J. O. Wolff. 1998. Chain re- the year-to-year variation in flowering � actions linking acorns to gypsy moth out- of these species. The effects of smooth- breaks and Lyme disease risk. Science 279: � �� ing out such variation on seed survival 1023–1026. �� Kelly, D., and V. L. Sork. 2002. Mast seeding in �� are unknown but potentially disastrous because fluctuations in seed-predator perennial plants: Why, how, where? Annual Review of Ecology and Systematics 33:427–447. Figure 12. High correlation (r = 0.88) of the populations will be less pronounced mean number of acorns per tree for valley oaks Knapp, E. E., M. A. Goedde and K. J. Rice. 2001. in the absence of masting. Similar det- Pollen-limited reproduction in blue oak: Impli- versus blue oaks at Hastings Reservation over rimental effects could become wide- cations for wind pollination in fragmented 25 years (1980 to 2004) indicates that the two spread among ecosystems in which populations. Oecologia 128:48–55. species strongly synchronize their production masting trees play an important role. Koenig, W. D., and J. M. H. Knops. 1998. Scale of acorns. Data points toward the upper right of mast-seeding and tree-ring growth. Nature corner of the plot represent mast years when These biomes include the majority of boreal and temperate woodlands and 396:225–226. both species were highly productive. Koenig, W. D., and J. M. H. Knops. 2000. Pat- probably a substantial proportion of terns of annual seed production by North- ally high temperatures during the sum- tropical forests. ern Hemisphere trees: a global perspective. mer, a phenomenon linked to the La American Naturalist 155:59–69. Niña phase of ENSO events. Rise and Fall … Liebhold, A., W. D. Koenig and O. N. Bjørnstad. The work of Lisa M. Curran, from the Masting among trees and other plants 2004. Spatial synchrony in population dy- University of Michigan, showcases an- probably confers significant adaptive namics. Annual Review of Ecology, Evolution, and Systematics 35:467–490. other example of interspecific synchrony. benefits. The boom and bust cycle of McKone, M. J., D. Kelly and W. G. Lee. 1998. She has investigated masting events in- seeds available to seed predators im- Effect of climate change on mast-seeding spe- volving at least 50 species of Dipterocar- poses starvation-satiation regimes that cies: frequency of mass flowering and escape paceae, the dominant family of canopy help control predator populations and from specialist insect seed predators. Global trees in Borneo. This scale of masting ensure that when seeds are produced Change Biology 4: 591–596. presents a scenario in which an entire plenty of them go uneaten. Addition- Satake, A., and Y. Iwasa. 2002. The synchronized and intermittent reproduction of forest trees forest comprises dozens of species all ally, masting may enhance pollination is mediated by the Moran effect, only in asso- flowering and fruiting at the same time. efficiency. Plants that synchronously ciation with pollen coupling. Journal of Ecology Based primarily on records of diptero- produce masses of flowers and pollen 90:830–838. carp seed export by timber concessions, once every few years are likely to experi- Schauber, E. M., D. Kelly, P. Turchin, C. Simon, W. she concluded that these trees produce ence higher pollination rates than plants G. Lee, R. B. Allen, I. J. Payton, P. R. Wilson, P. seed crops only every three to four years, that produce average amounts of flow- E. Cowan and R. E. Brockie. 2002. Masting by 18 New Zealand plant species: The role of again during ENSO events. ers and pollen each year. The benefits so temperature of a synchronizing cue. Ecology These examples prove that an envi- outweigh the costs that masting species 83:1214–1225. ronmentally driven Moran effect alone more or less forgo reproduction entirely Sork, V. L., J. Bramble and O. Sexton. 1993. can be a strong synchronizing agent in non-masting years. Teasing apart the Ecology of mast-fruiting in three species of in masting species, although they do relative importance of these ideas will be North American deciduous oaks. Ecology not entirely eliminate the possibility the focus of future research into masting. 74:528–541. Sork, V. L., F. W. Davis, P. E. Smouse, V. J. Apsit, that pollen coupling plays a role in However, scientists must also under- R. J. Dyer, J. F. Fernandez-M. and B. Kuhn. synchronizing reproduction. For our stand how masting will be affected by 2002. Pollen movement in declining popula- California oaks, the presence of high the worrying decline in California oak tions of California valley oak, Quercus lobata: interspecific synchrony in acorn pro- populations and also by climate change, Where have all the fathers gone? Molecular duction between valley and blue oaks, problems that portend dramatic changes Ecology 11:1657–1668. which are closely related but rarely hy- in the food and nutrient resources made bridize, suggests that, in these species available by masting. Predicting the eco- as well, the Moran effect is a dominant system-wide effects of these risk factors force behind masting. is a major challenge to ecologists. For relevant Web links, consult this So next fall, don’t take those acorns issue of American Scientist Online: Masting and Climate Change for granted. The fact that they are every- The prevalent role of weather as a cue where, or nowhere, as the case may be, http://www.americanscientist.org/ for mast fruiting raises the question illustrates one of the more spectacular IssueTOC/issue/741 of how global climate change might ecological mysteries that you’ll encoun- affect masting events and subsequent ter as you walk through the woods. En- ecosystem functioning. For example, joy them while you have the chance.