Stable oxygen isotope ratios of plant water (sap water) were observed at Spasskaya Pad experimental forest near Yakutsk, Russia in 1997–1999. The δ18O of sap water in larch trees (Larix gmelinii) decreased soon after leaf unfolding every year, indicating that snowmelt water was used in the beginning of summer. During mid to late summer, a clear difference in the water source used by plants was observed between wet summers and severe drought summers. The δ18O values of water in larch trees were high (−17.8 to −16.1‰) in August 1999 (wet summer), but low (−20.4 to −19.7‰) in August 1998 (drought summer). These results indicated that plants used rainwater during a wet summer, but meltwater from permafrost was used by plants during a drought summer. One important role of permafrost is to provide a direct source of water for plants in a severe drought summer; another role is to keep surplus water in the soil until the next summer. If this permafrost system is disturbed by future global warming, unique monotypic stands of deciduous larch trees in east Siberia might be seriously damaged in a severe drought summer.

Abstract:Soil moisture and its isotopic composition were observed at Spasskaya Pad experimental forest near Yakutsk, Russia, during summer in 1998Russia, during summer in , 1999Russia, during summer in , and 2000. The amount of soil water (plus ice) was estimated from volumetric soil water content obtained with time domain reflectometry. Soil moisture and its υ 18 O showed large interannual variation depending on the amount of summer rainfall. The soil water υ 18 O decreased with soil moisture during a dry summer (1998), indicating that ice meltwater from a deeper soil layer was transported upward. On the other hand, during a wet summer (1999), the υ 18 O of soil water increased due to percolation of summer rain with high υ 18 O values. Infiltration after spring snowmelt can be traced down to 15 cm by the increase in the amount of soil water and decrease in the υ 18 O because of the low υ 18 O of deposited snow. About half of the snow water equivalent (about 50 mm) recharged the surface soil. The pulse of the snow meltwater was, however, less important than the amount of summer rainfall for intra-annual variation of soil moisture.Excess water at the time just before soil freezing, which is controlled by the amount of summer rainfall, was stored as ice during winter. This water storage stabilizes the rate of evapotranspiration. Soil water stored in the upper part of the active layer (surface to about 120 cm) can be a water source for transpiration in the following summer. On the other hand, once water was stored in the lower part of the active layer (deeper than about 120 cm), it would not be used by plants in the following summer, because the lower part of the active layer thaws in late summer after the plant growing season is over.

Distribution patterns along a slope and vertical root distribution were compared among seven major woody species in a secondary forest of the warm‐temperate zone in central Japan in relation to differences in soil moisture profiles through a growing season among different positions along the slope. Pinus densiflora, Juniperus rigida, Ilex pedunculosa and Lyonia ovalifolia, growing mostly on the upper part of the slope with shallow soil depth had shallower roots. Quercus serrata and Quercus glauca, occurring mostly on the lower slope with deep soil showed deeper rooting. Styrax japonica, mainly restricted to the foot slope, had shallower roots in spite of growing on the deepest soil. These relations can be explained by the soil moisture profile under drought at each position on the slope. On the upper part of the slope and the foot slope, deep rooting brings little advantage in water uptake from the soil due to the total drying of the soil and no period of drying even in the shallow soil, respectively. However, deep rooting is useful on the lower slope where only the deep soil layer keeps moist. This was supported by better diameter growth of a deep‐rooting species on deeper soil sites than on shallower soil sites, although a shallow‐rooting species showed little difference between them.

We compared the amount of variation in flower size between autogamous and insect‐pollinated species to examine the hypothesis that pollinator‐mediated selection stabilizes flower size in plant populations. One would expect the flower size variation to be larger in selfing species that are less affected by pollinator‐mediated stabilizing selection than in insect‐pollinated species. The results of phylogenetic comparisons between autogamous and insect‐pollinated flowers supported the pollinator‐mediated stabilizing selection hypothesis, although the non‐phylogenetic comparison did not. According to our results, we discuss the factors influencing the flower size variation.