Drought, temperature, and moisture availability: understanding the drivers of isotopic decoupling in native pine species of the Nepalese Himalaya

Abstract

The Himalayas experienced long-term climate changes and recent extreme weather events that affected plant growth and the physiology of tree species at high-elevation sites. This study presents the first statistically robust δ¹⁸O_TR chronologies for two native pine species, Pinus roxburghii, and Pinus wallichiana, in the lower Nepalese Himalaya. The isotope chronologies exhibited 0.88‰ differences in overall mean isotope values attributed to varying elevations (460-2000 m asl). Comparative analysis of climate response using data sets from different sources and resolutions revealed the superiority of the APHRODITE (Asian Precipitation - Highly-Resolved Observational Data Integration Towards Evaluation) data set calibrated for the South Asian Summer Monsoon (SASM)-dominated region. Both species exhibited negative correlations with monsoon precipitation and positive correlations with temperature. However, during the peak monsoon season (July-August), daily resolved climate data disentangled statistically insignificant relationships, and revealed that δ¹⁸O_TR is influenced by atmospheric moisture. Both congeneric species showed a decoupling between the chronologies after 1995. However, no significant change in air moisture origin and monsoon regime between the study sites was observed, indicating a consistent dominant moisture source during different monsoon seasons. Besides, we also observed the decreased inter-series correlation of both δ¹⁸O_TR chronologies after 1995, with P. wallichiana experiencing a steeper decrease than P. roxburghii. The weakening correlations between and within the chronologies coincided with a regional drought during 1993-1995 in both sites, highlighting the strong regulation of local climate on the impact of regional extreme climate events. Our findings emphasise the importance of employing climate data with optimal spatial and temporal resolution for improved δ¹⁸O_TR-climate relationships at the intra-annual scale while considering the influence of site-specific local environmental conditions. Assessing climate data sets with station data is vital for accurately interpreting climate change's impact on forest response and long-term climate reconstructions.

Keywords: Isotopic decoupling; Local environmental effects; Native pine species; Nepalese Himalaya; Precipitation; Stable oxygen isotopes; Temperature.

Fig. 1

Map of the study area indicating study sites (blue and red circles), climate stations (blue and red triangles), and elevation (colour in the background)

Fig. 2

Average climate conditions and recent climate trends in both study sites. (a) and (b) represent the monthly temperature and precipitation patterns in the Hetauda and Daman stations. (c) and (d) describe the annual mean temperature and precipitation trends —source: Government of Nepal, Department of Hydrology and Meteorology

Fig. 3

(a) Stable oxygen isotope chronologies of P. roxburghii and P. wallichiana, including the 5% confidence intervals and mean values represented by dashed coloured lines. (b) The 11-year moving mean inter-series correlation of the isotope series. (c) The 11-year running bootstrapped correlation between the PIRO and PIWA isotope series with a 5% confidence interval. The grey-shaded vertical column represents the drought period from 1993 to 1995

Fig. 4

Correlation diagram between δ¹⁸O_TR chronologies of the two studied pine species and climate variables from CRU, ERA5, and APHRODITE data sets. TEM, PRE, scPDSI, and RH denote the average temperature, total precipitation, self-calibrated PDSI, and relative humidity. Cells including a value represent significant correlations at p < 0.05

Fig. 5

30-year moving correlation analysis of PIWA and PIRO chronologies using 30-day moving climate windows of APHRODITE data set. Figures (a) and (b) represent correlations with running sums of daily precipitation, (c) and (d) represent correlations with a running mean of daily temperatures. Shaded areas represent significant correlations at p < 0.01 (dotted areas) and p < 0.05 (hatched areas) levels, respectively

Fig. 6

30-year moving correlation analysis of PIWA and PIRO stable oxygen isotope chronologies using 30 days moving averages of vapour pressure deficit (VPD) of (a) PIWA and (b) PIRO site using ERA5-land data set. Shaded areas represent significant correlations at p < 0.01 (dotted areas) and p < 0.05 (striped areas)

Fig. 7

Comparison of linear regression coefficients between CRU TS, APHRODITE, and ERA5-land reanalysis data sets and instrumental climate data (1982–2012). DAM and HET represent the climate of the local meteorological stations at Daman and Hetauda. ‘***’ means indicates regressions significant at the 99% confidence level

Fig. 8

Air parcel trajectories to the sites from 1991–1995. The upper panels (a and b) show the trajectory path to the PIRO site for DOY 105–181 and DOY 243–304, whereas the lower panel (c and d) shows the same for the PIWA site. The polar plots inside each sub-figure indicate the percentage of trajectories from different directions (N: North, E: East, S: South, and W: West). The values in parenthesis represent the percentage of precipitation from each bearing (direction). (Please see Fig. S5-S8 in supporting documents for all analysed 5-year periods.)

Fig. 9

(a) Annual scPDSI of both investigation sides, indicating a severe drought during 1993–1995, indicated by grey bars in all panels. (b) and (d) total precipitation trends from days 105–181 (representing the early monsoon season) and 243–304 (= late monsoon season), respectively. (c) and (e) average temperature trends for the same periods. Red and blue colours represent the PIRO and PIWA sites. Dashed lines represent the locally weighted scatter plot smoothing (LOWESS) lines