Keywords

1 Introduction

The genus Pterocarya belongs to the family Juglandaceae and was once widely distributed in the Northern Hemisphere, with fossils dating to the early Oligocene (Manchester 1987, 1989; Manos et al. 2007; Kozlowski et al. 2018). Currently, the extant Pterocarya comprise six generally accepted species, with one species in western Eurasia (Turkey, Georgia, Azerbaijan, and Iran) and five others in eastern Asia (China, the Korean Peninsula, Laos, Vietnam, Taiwan, and Japan) (Kozlowski et al. 2018). All of those Pterocarya species are elements of riparian forests in each region. In Japan, Juglandaceae comprises three species in three genera: Pterocarya rhoifolia Sieb. et Zucc., Platycarya strobilacea Sieb. et Zucc., and Juglans mandshurica Maxim. var. sachalinensis (Komatsu) Kitam.

The Japanese wingnut P. rhoifolia is the only volunteer species of Pterocarya in Japan. Although reports refer to the presence of an adjunct P. rhoifolia population from China (Laoshan, East Shandong), there is no reliable record and the presence of this species in China remains unclear (Kozlowski et al. 2018). The standard Japanese name for P. rhoifolia is “Sawagurumi,” which is derived from the fact that this species grows near mountain streams. P. rhoifolia is a representative riparian element in the cool temperate forest of Japan, is widely distributed from southern Hokkaido to Kyusyu, and grows in regions with heavy and light snowfall (Hotta 1975; Kawahara et al. 2009; Nakano and Sakio 2017; Fig. 3.1). In snowy regions on the Japan Sea side and Tohoku region (northeastern Honshu), P. rhoifolia forms the riparian forest canopy with Aesculus turbinata Blume and Cercidiphyllum japonicum Sieb. et Zucc. (Kikuchi 1968; Suzuki et al. 2002). On the Pacific Ocean side, Fraxinus platypoda Oliv. coexists with these species (Sakio 1997; Sakio et al. 2002).

Fig. 3.1
figure 1

(a) Annual maximum snow depth in Japan depth derived from Japanese Meteorological Agency mesh climate data (Nakano and Sakio 2017 revised), (b) map of the P. rhoifolia distribution

Pterocarya rhoifolia forests develop on stream banks, mud flow terraces, and flood terraces in floodplains, and on the alluvial cone and talus in lower hill slopes along mountain streams (Sato 1988; Kaneko and Kawano 2002). These habitats characteristically undergo land-surface disturbance, e.g., floods, debris flows, and landslides. Therefore, P. rhoifolia maintains its populations by adapting its life history to disturbance regimes in riparian areas.

In the Ooyamazawa riparian area on to the Pacific Ocean side of Japan, P. rhoifolia forms the canopy with F. platypoda and C. japonicum. Although P. rhoifolia dominates the canopy with F. platypoda in the Ooyamazawa riparian area, P. rhoifolia canopy trees also grow in patches along streams. In this chapter, we clarify the life history of P. rhoifolia in riparian forest based on studies in the Ooyamazawa riparian forest, and discuss the adaptation of the P. rhoifolia life history to riparian disturbances.

2 Study Species

Pterocarya rhoifolia is a deciduous canopy tree species with straight trunks and can reach 35 m in height and 1.2 m in diameter at breast height (DBH) (Hayashi 1969). It has a single large trunk or a few large trunks in a stool. It often produces sprouts, although not as frequently as C. japonicum. In the Ooyamazawa riparian forest, almost all P. rhoifolia had a single large trunk in a stool. In 2011, we recorded a maximum DBH and tree height of P. rhoifolia of 89.8 cm and 39.9 m, respectively (Fig. 3.2). Its bark is gray and split lengthwise in the mature tree stage. The terminal bud has 1–3 bud scales that fall between autumn and winter and overwinters as a naked bud with pubescence. Its leaves are alternate and imparipinnate compound, 20–30 cm long. The leaves have 9–21 leaflets (Fig. 3.3). The petiole and rachis are finely pubescent. The leaf rachis of the closely related species Pterocarya stenoptera is often winged, while that of P. rhoifolia is not.

Fig. 3.2
figure 2

The typical mature tree form of P. rhoifolia (in the Ooyamazawa riparian forest)

Fig. 3.3
figure 3

The leaves of P. rhoifolia

3 Reproductive Characteristics

Pterocarya rhoifolia is a monoecious species with unisexual flowers. It is an anemophilous species, and flowers bloom and new leaves develop simultaneously. P. rhoifolia produces a female catkin hanging from the terminal part of a current-year shoot and some green male catkins hanging from the axil at the base of a current-year shoot (Fig. 3.4). It flowers in late May in the Ooyamazawa riparian forest. Although the male catkins fall off after pollen dispersal, the female catkins develop into infructescences. One infructescence has 20–60 fruit (Kaneko 2009; Figs. 3.5 and 3.6). One P. rhoifolia fruit is a nut with two wings that develop from bractlets; hence, the English name of P. rhoifolia is “Japanese wingnut.” The nut is about 0.8 cm in size, and about 2.2 cm including the wings; the oven-dried nut and nut including wings of P. rhoifolia (mean ± SD) weighed 70 ± 8 and 90 ± 11 mg, respectively (Sakio et al. 2002). A fruit contains one seed. In the Ooyamazawa riparian forest, P. rhoifolia fruit matures and is dispersed by wind from September to November (Fig. 3.7).

Fig. 3.4
figure 4

A flowering branch of P. rhoifolia with male and female catkins

Fig. 3.5
figure 5

Fruiting individual of P. rhoifolia

Fig. 3.6
figure 6

A branch with leaves and a fruiting spike

Fig. 3.7
figure 7

Seasonal change in amount of fallen P. rhoifolia fruit in the Ooyamazawa riparian forest (1995–2005)

Pterocarya rhoifolia can live for up to 150 years (Kaneko 2009) and starts producing fruit at 40–80 years, at which point the DBH is 28 cm and the tree height is 16 m (The Japanese Riparian Forest Research Group 2001). Although P. rhoifolia reaching the canopy layer with stems larger than 30 cm DBH can bloom, small-diameter individuals released from suppression and understory individuals cannot bloom (Kaneko 2009). Therefore, to reach reproductive maturity, P. rhoifolia requires sufficient size and light.

Mature P. rhoifolia produce seeds most years, although the annual seed production fluctuates (Sawada et al. 1998; Kaneko and Kawano 2002; Sakio et al. 2002). In the Ooyamazawa riparian forest, P. rhoifolia produced seeds almost every year between 1995 and 2014, although very few were produced in 1996 and 2006 (Fig. 3.8). Moreover, the annual variation in seed production has tended to alternate yearly since 2005. The coefficient of variation (CV) between 2005 and 2014 was larger than that between 1995 and 2004 (CV1995–2004 = 0.75, CV2005–2014 = 0.89). The variation in annual seed production in P. rhoifolia is intermediate between that of F. platypoda and C. japonicum (Sakio et al. 2002).

Fig. 3.8
figure 8

Annual change in the number of dispersed seeds of P. rhoifolia in the Ooyamazawa riparian forest (1995–2014). Vertical bars indicate the standard deviation

4 Structure and Distribution

According to Sakio et al. (2002), a 4.71 ha study plot in the Ooyamazawa riparian forest contained 112 P. rhoifolia over 4 cm in DBH, of which 80 (71.4%) were in the canopy layer. The mean DBH of the P. rhoifolia canopy trees was 44.6 ± 11.5 cm, with a maximum of 77.7. The DBH distribution of the P. rhoifolia population was continuous from saplings to large individuals (Fig. 3.9), but there were some peaks: one for small trees <5 cm in DBH (suggesting that P. rhoifolia maintains sapling banks) and another at 35–40 cm in DBH (suggesting that synchronous regeneration of P. rhoifolia has occurred several times in the past). The trees form patches of various sizes along a stream, with some large high-density patches on large landslide deposits (Sakio et al. 2002; Fig. 3.10). The mean age of the P. rhoifolia trees forming these patches is about 90 years and most are even-aged. These results suggest that P. rhoifolia invades the sites of large disturbances, such as floods, debris flows, and landslides, where it forms colonies.

Fig. 3.9
figure 9

Frequency distribution of the DBH of P. rhoifolia in a 4.71-ha study plot in the Ooyamazawa riparian forest

Fig. 3.10
figure 10

Distribution of three canopy trees along a stream in the Ooyamazawa riparian forest. Patches (A–C) where a large landslide occurred about 90 years ago were dominated by P. rhoifolia canopy trees with a few C. japonicum trees (Sakio et al. 2002)

5 Seedling Regeneration

The dispersed seeds of P. rhoifolia generally germinate in the spring of the following year. P. rhoifolia forms almost no soil seed bank. Kubo et al. (2008) investigated the species composition of the soil seed bank (to 5 cm depth) for the Ooyamazawa riparian forest and found only one buried viable P. rhoifolia seed in 30 L soil. The current-year seedlings of P. rhoifolia are epigeal cotyledons type, which have two opposite cotyledons deeply palmately four-clefted at germination (Fig. 3.11). Their alternate primary leaves are simple leaves until the first or second leaf stage, and become imparipinnate compound leaves at larger leaf stages.

Fig. 3.11
figure 11

Current-year seedling of P. rhoifolia with four-part cotyledons

The germination sites of current-year seedlings of P. rhoifolia in the Ooyamazawa riparian forest are on litter, gravel, mineral soil, and fallen logs (Sakio et al. 2002). On Chichibu Mountain, P. rhoifolia seeds germinate both under closed canopy and in gaps in the riparian forest dominated by F. platypoda and P. rhoifolia (Kisanuki et al. 1995). These results suggest that the germination sites of P. rhoifolia in riparian areas are not markedly limited.

In the Ooyamazawa riparian forest, current-year seedlings that germinate on litter tend to die in the current year because of low light intensity and drought, while current-year seedlings on gravel, mineral soil, and fallen logs die within 3 years (Sakio et al. 2002). Furthermore, almost all current-year P. rhoifolia seedlings that germinate under a closed canopy die in the current year (Kisanuki et al. 1995). Taking this into consideration, under what conditions can current-year seedlings of P. rhoifolia survive and grow? Nursery experiments have indicated that the height of current-year P. rhoifolia seedlings increases with relative light intensity (RLI), reaching about 30 cm at over 40% RLI (Sakio et al. 2008). An investigation of the relationship between the distribution of P. rhoifolia saplings (height >1 m and DBH <4 cm) and topography in the Ooyamazawa riparian forest found that more saplings were distributed in abandoned channels and floodplains than on hillslopes or in active channels (Sakio et al. 2002, Table 3.1; Fig. 3.12). The abandoned channels and floodplains were directly affected by stream flow and lacked an understory. Moreover, the microhabitats in the Ooyamazawa riparian forest in which P. rhoifolia saplings (≥1 years old) tend to thrive consist of deposited gravel with thin soil and no herbaceous layer (Kubo et al. 2000).

Table 3.1 Number of grids with P. rhoifolia saplings (height ≥1 m) and the mean number of saplings per grid (Sakio et al. 2002 revised)
Fig. 3.12
figure 12

Microtopography and distribution of P. rhoifolia saplings in a 0.54-ha study plot in the Ooyamazawa riparian forest (Sakio et al. 2002 revised). This plot was divided into 1350 quadrats (2 × 2 m2). The microtopography was recorded in each quadrat and was classified into the following categories: hillside, active channel, abandoned channel, and floodplain. The figures show the numbers of P. rhoifolia saplings in each quadrat

There are no statistical relationships between canopy gaps and the distribution of P. rhoifolia saplings (height >1 m and DBH <4 cm) in the Ooyamazawa riparian forest (Table 3.1; Fig. 3.13). However, Sato (1992) investigated the regeneration of P. rhoifolia saplings in a P. rhoifolia forest in Hokkaido and found a weak positive correlation between the distribution of P. rhoifolia saplings (2 m ≤ tree height ≤ 4 m) and relative illumination, and reported that P. rhoifolia saplings were distributed contiguously around canopy gaps. Moreover, Sakio (1993) investigated the pattern of leaf expansion and shoot elongation of P. rhoifolia saplings in Ooyamazawa riparian forest, and found that P. rhoifolia saplings (height = 80.3±13.3 cm) in a canopy gap continued to expand their leaves and develop current-year shoots from May to August.

Fig. 3.13
figure 13

Canopy gaps and distribution of P. rhoifolia saplings in a 0.54-ha study plot in the Ooyamazawa riparian forest. This plot was divided into 1350 quadrats (2 × 2 m2)

The factors that enable the survival and growth of P. rhoifolia seedlings and saplings have not been explained completely. However, many investigations have suggested that, for P. rhoifolia to regenerate seedlings, bare ground caused by land-surface disturbance with destruction of the herbaceous layer and a canopy gap are probably needed.

6 Sprouting Traits

Pterocarya rhoifolia produces some sprouts from the root collar and lowest part of the stem, but lacks root suckers. In the Ooyamazawa riparian forest, P. rhoifolia has a maximum of 10 sprouts/individual and there are no significant correlations between the number of sprouts and DBH of the main stem (Sakio et al. 2002). In an unpublished study, overall, 79.2% of individuals had sprouts, 95.3% of which were less than 50 cm high (Y. Nakano, unpublished data). In the Ooyamazawa riparian forest, such sprouts have no role in maintaining the population (Nakano and Sakio 2017).

However, the sprouting traits of P. rhoifolia change in response to maximum snow depth. Nakano and Sakio (2017) comparing several areas with maximum snow depths from 30 to 480 cm in a cool temperate mountainous area in central Japan, which includes the Ooyamazawa riparian forest as a low-snow region, and reported that the number of sprouts per P. rhoifolia individual increased with maximum snow depth (Fig. 3.14). In deep snow, the sprouts of P. rhoifolia may play a role in repairing individuals damaged by the snowpack to maintain the population (Nakano and Sakio 2018). They also found that, with increasing maximum snow depth, the DBH decreased, maximum stem length and tree height shortened, trees tended toward a “dwarf shrub” form, and seed production decreased (Nakano and Sakio 2017). These results suggest trade-offs between clonal growth (producing sprouts) and sexual reproduction (seed production) and between producing sprouts and height growth (Nakano and Sakio 2017; Fig. 3.15). This sprouting ability could conceivably be due to P. rhoifolia growing in an environment that tends to be influenced by disturbances specific to riparian areas, such as floods and landslides. Therefore, the sprouting ability of P. rhoifolia is insurance enabling it to survive when threatened.

Fig. 3.14
figure 14

Relationship between the maximum snow depth and number of sprouts per mature P. rhoifolia. N number of individuals (Nakano and Sakio 2017 revised)

Fig. 3.15
figure 15

Schematic diagram relating the life history traits of P. rhoifolia and snowfall conditions (Nakano and Sakio 2017 revised)

Ito (1992) investigated dry matter partitioning in seedlings and sprouts of P. rhoifolia and showed that the slopes of the regression lines of the allometric relationship between basal trunk diameter and tree height were greater in sprouts than in seedlings, suggesting a change in growth characteristics from the “waiting” type (diameter-preferred growth) in seedlings to the “competing” type (elongation-preferred growth) in sprouts.

7 Conclusion

In the Ooyamazawa riparian forest, P. rhoifolia forms the canopy with F. platypoda and C. japonicum. Sato (1992) compared the distribution, growth, and survival of saplings of the main canopy tree species (P. rhoifolia, Acer mono, Alnus hirsute, and Ulmus laciniata) in a P. rhoifolia forest in Hokkaido. Defining a pioneer species by its high germinating capacity, high growth rate, and high survival rate under sufficient light, they concluded that P. rhoifolia is not a typical pioneer species, but is intermediate between a pioneer and non-pioneer species. In the Ooyamazawa riparian forest, P. rhoifolia is more of a pioneer than F. platypoda and C. japonicum, particularly in terms of the growth rate during the seedling and sapling stages (cf. Chap. 7).

Mature P. rhoifolia produce and disperse seeds almost every year. P. rhoifolia seeds invade the sites of large disturbances, including floods, debris flows, landslides, and canopy gaps, and establish there. Moreover, P. rhoifolia can grow rapidly with sufficient light. Consequently, this tree species can form the canopy and even-aged populations in the Ooyamazawa riparian forest. In this forest, P. rhoifolia does not produce many sprouts. Hence, sprouts may not contribute to maintaining P. rhoifolia individuals for a long time, such as C. japonicum (cf. Chap. 4), and the lifespan of P. rhoifolia appears to be about 150 years (Kaneko 2009). Consequently, the dominance ratio of late-successional tree species in the canopy such as F. platypoda and C. japonicum may increase after the lifespan of P. rhoifolia without large disturbances. However, if large disturbances occur along rivers and generate suitable sites for P. rhoifolia regeneration, even-aged P. rhoifolia forest may reestablish (Sato 1988). Therefore, the life history traits of P. rhoifolia are adapted to disturbances in riparian areas and the species may be able to maintain its population in the Ooyamazawa riparian forest.