In a warming climate, temperature-sensitive plants must move toward colder areas, that is, higher latitude or altitude, by seed dispersal []. Considering that the temperature drop with increasing altitude (-0.65°C per 100 m altitude) is one hundred to a thousand times larger than that of the equivalent latitudinal distance [], vertical seed dispersal is probably a key process for plant escape from warming temperatures. In fact, plant geographical distributions are tracking global warming altitudinally rather than latitudinally, and the extent of tracking is considered to be large in plants with better-dispersed traits (e.g., lighter seeds in wind-dispersed plants) []. However, no study has evaluated vertical seed dispersal itself due to technical difficulty or high cost. Here, we show using a stable oxygen isotope that black bears disperse seeds of wild cherry over several hundred meters vertically, and that the dispersal direction is heavily biased towards the mountain tops. Mountain climbing by bears following spring-to-summer plant phenology is likely the cause of this biased seed dispersal. These results suggest that spring- and summer-fruiting plants dispersed by animals may have high potential to escape global warming. Our results also indicate that the direction of vertical seed dispersal can be unexpectedly biased, and highlight the importance of considering seed dispersal direction to understand plant responses to past and future climate change.

Main Text

18O and 16O. A decrease in the oxygen isotope ratios (18O/16O) with altitude is known in plant tissues, such as tree rings [ 3 Burk R.L.

Stuiver M. Oxygen isotope ratios in trees reflect mean annual temperature and humidity. 4 Cousens R.

Dytham C.

Law R. Dispersal in Plants: A Population Perspective. 4 Cousens R.

Dytham C.

Law R. Dispersal in Plants: A Population Perspective. To evaluate vertical seed dispersal, we used naturally-occurring stable oxygen isotopes, namelyO andO. A decrease in the oxygen isotope ratios (O/O) with altitude is known in plant tissues, such as tree rings [] (see also Supplemental Experimental Procedures ). If the oxygen isotope ratios of seeds are negatively correlated with the altitude where they are produced, we can determine the altitude of the mother plant of a dispersed seed using its oxygen isotope ratios, and we can estimate vertical seed dispersal, or the vertical distance between a dispersed seed and its mother plant. To test this idea, we targeted an animal-dispersed plant, Prunus verecunda, which has no chance to increase its altitude without the help of animals []. We focused on frugivorous mammals, which are the primary seed dispersers for many plants, and disperse seeds longer distances than other animals []. We conducted this field study in central Japan from 2010–2013.

Figure 1 Vertical seed dispersal by mammals. Show full caption (A) Negative correlation between altitude and the oxygen isotope ratios of Prunus verecunda reference seeds in 2012 and 2013. The lines show the least-squares fit (2012, solid line; 2013, dashed line). Different symbols in the same year and altitude indicate different fruiting trees. The regression line showed an annual variation (see also Supplemental Experimental Procedures ). The oxygen isotope ratios of the seeds did not show large variation among fruiting trees at the same altitude, and among seeds from the same fruiting tree. (B) Vertical seed dispersal distance achieved by Asiatic black bears and (C) Japanese martens (four years data were pooled). Positive and negative values indicate seed dispersal toward the mountain tops and foots, respectively. See also Figure S1 . (D) Vertical seed dispersal toward the mountain tops by mammals that are following the spring-to-summer plant phenology. The spring-to-summer plant phenology proceeds from the foot to the top of mountains. Cherry fruits and young vegetation are no longer available in low altitudes, ripe fruits are available but young vegetation is no longer available in middle altitudes, and ripe fruits are unavailable but young vegetation is available in high altitudes. A strong negative correlation was detected between altitude and the oxygen isotope ratios of non-dispersed reference seeds of P. verecunda ( Figure 1 A), and thus we used the regression lines as calibration lines for estimating vertical seed dispersal. We estimated vertical seed dispersal by Asiatic black bears Ursus thibetanus and Japanese martens Martes melampus, which accounted for 80.3 and 19.6 % of mammal-dispersed seeds, respectively.

5 Koike S.

Masaki T.

Nemoto Y.

Kozakai C.

Yamazaki K.

Kasai S.

Nakajima A.

Kaji K. Estimate of the seed shadow created by the Asiatic black bear Ursus thibetanus and its characteristics as a seed disperser in Japanese cool-temperate forest. 6 Ohdachi S.D.

Ishibashi Y.

Iwasa M.A.

Saitioh T. The Wild Mammals of Japan. 2 Barry R.G.

Chorley R.J. Atmosphere, Weather and Climate. 7 IPCC Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Our analysis of the isotope ratios of seeds from bear dung indicates that bears frequently disperse seed over several hundred meters vertically ( Figure 1 B). Furthermore, their dispersal direction was surprisingly biased toward the mountain tops, and this was consistent among years ( Figure S1 ). The corresponding horizontal distance to the vertical distance was 749.5 ± 662.3 m SD ( Supplemental Experimental Procedures ), which is roughly identical to the distance of GPS-tracked bear movement during seed retention times in the bear gut []. Seed dispersal by martens showed similar trends, but with shorter dispersal distance ( Figure 1 C; mean horizontal distance, 460.5 ± 414.8 m SD), probably reflecting their smaller home range due to smaller body size and territorial behavior []. The long-distance seed dispersal toward the mountaintops by bears and martens was considered to be sufficient to keep up with the rate of global warming; mean vertical distance by bears and martens correspond to −2.0 and −1.3°C drops, respectively [], while projected global temperature rise until 2100 is 4.8°C in maximum [].

In particular, bears are very important seed dispersers considering their quantitative and qualitative contributions. It is notable, however, that the effect of vertical seed dispersal on plant escape should be interpreted with care: depending on the height of the mountains, vertical dispersal may ultimately only provide short-term relief and might actually be a trap if the mountains are not high enough.

8 Rötzer T.

Chmielewski F.M. Phenological maps of Europe. 9 Koike S.

Kasai S.

Yamazaki K.

Furubayashi K. Fruit phenology of Prunus jamasakura and the feeding habit of the Asiatic black bear as a seed disperser. 9 Koike S.

Kasai S.

Yamazaki K.

Furubayashi K. Fruit phenology of Prunus jamasakura and the feeding habit of the Asiatic black bear as a seed disperser. 5 Koike S.

Masaki T.

Nemoto Y.

Kozakai C.

Yamazaki K.

Kasai S.

Nakajima A.

Kaji K. Estimate of the seed shadow created by the Asiatic black bear Ursus thibetanus and its characteristics as a seed disperser in Japanese cool-temperate forest. 6 Ohdachi S.D.

Ishibashi Y.

Iwasa M.A.

Saitioh T. The Wild Mammals of Japan. Highly biased seed dispersal toward the mountain tops is probably due to the spring-to-summer phenology of the food plants of bears and martens. The spring-to-summer phenology of plants, including fruit maturation of cherry species, proceeds from the foot to the top of mountains in the temperate zone []. In addition, Asiatic black bears are known to visit fruiting cherry trees at higher altitude later []. Thus, bears and martens would ascend mountains following the phenology of their food plants, resulting in biased seed dispersal toward the mountain tops ( Figure 1 D). However, mammals’ simple tracking of ascending plant phenology does not seem a sole causal factor, considering that seed retention time in the gut of bear and marten is at most 45 hours [], and 45 hours is not long enough for plant phenology to ascend 300 m, the distance achieved by seed dispersal. Another contributing factor may be mammals’ daily foraging pattern, affected by the existence of young vegetation such as herbs, and buds, young leaves and flowers of trees. During spring and summer, this young vegetation becomes another major food source for bears and martens, and is available earlier than cherry fruits []. It is likely that the mammals move to lower altitudes for cherry fruits earlier in the fruiting period when fruits are available, and later move to higher altitudes for young vegetation. These descents and ascents would result in more biased seed dispersal toward the mountain tops, because seed dispersal does not occur during a descent (i.e., no seed in mammal gut during a descent) and because this movement pattern would not stop until either food becomes unavailable.

8 Rötzer T.

Chmielewski F.M. Phenological maps of Europe. 6 Ohdachi S.D.

Ishibashi Y.

Iwasa M.A.

Saitioh T. The Wild Mammals of Japan. 10 Alexander J.M.

Diez J.M.

Levine J.M. Novel competitors shape species’ responses to climate change. Our results suggest that animal-dispersed spring- and summer-fruiting plants in the temperate zone have a high potential to escape global warming when their seeds are dispersed by mammals. However, this may be not the case with many other plants fruiting later; dispersal of seeds to lower altitudes can occur if mammals descend following the autumn–winter phenology, proceeding from the top to the foot of mountains []. In fact, temperate bears, including Asiatic black bears, are known to descend mountains from autumn to winter, especially in poor fruiting years []. Considering that seed dispersal toward the mountain bases under global warming can lead to serious population decrease or extinction of plants due to reduced performance there [], testing this hypothesis is essential from a conservation perspective.