Differential growth rate, water-use efficiency and climate sensitivity between males and females of Ilex aquifolium in north-western Spain

Abstract

Background and aims: Dioecious plant species, i.e. those in which male and female functions are housed in different individuals, are particularly vulnerable to global environmental changes. For long-lived plant species, such as trees, long-term studies are imperative to understand how growth patterns and their sensitivity to climate variability affect the sexes differentially.

Methods: Here, we explore long-term intersexual differences in wood traits, namely radial growth rates and water-use efficiency quantified as stable carbon isotope abundance of wood cellulose, and their climate sensitivity in Ilex aquifolium trees growing in a natural population in north-western Spain.

Key results: We found that sex differences in secondary growth rates were variable over time, with males outperforming females in both radial growth rates and water-use efficiency in recent decades. Summer water stress significantly reduced the growth of female trees in the following growing season, whereas the growth of male trees was favoured primarily by cloudy and rainy conditions in the previous autumn and winter combined with low cloud cover and warm conditions in summer. Sex-dependent lagged correlations between radial growth and water availability were found, with a strong association between tree growth and cumulative water availability in females at 30 months and in males at 10 months.

Conclusions: Overall, our results point to greater vulnerability of female trees to increasing drought, which could lead to sex-ratio biases threatening population viability in the future.

Keywords: Dendroecology; carbon isotope discrimination; dioecy; sexual dimorphism; tree growth; water-use efficiency.

Fig. 1.

(A) Location of the study area in the Sierra de Ancares, Galicia, north-western Spain. (B) Climatic diagram of the Sierra de Ancares for the period 1901–2000. Maximum (Tmax), mean (Tmean) and minimum (Tmin) monthly temperatures and total monthly precipitation (Prec) series are displayed along with mean annual temperature (T) and total annual precipitation (P). (C) Temporal variation of the mean annual temperature in the period 1901–2000, with its linear fit, r² and P-values are shown.

Fig. 2.

(A) Annual variation in basal area increment (BAI; in centimetres squared per year) for Ilex aquifolium males and females. (B, C) Variation at 5-year intervals of mean (±s.e.) BAI (B) and mean (±s.e.) carbon isotope discrimination (∆¹³C; per mille) (C) for males and females.

Fig. 3.

Adjusted linear predictions of the year × sex interaction term from basal area increment basal area increment (BAI) (A) and carbon isotope discrimination (∆¹³C) (B) best models (Table 2). The shaded area shows 95 % confidence intervals.

Fig. 4.

Relationships between mean basal area increment (BAI; in centimetres squared per year) and carbon isotope discrimination (∆¹³C; per mille) at 5-year intervals for males and females. The obtained linear fits, their coefficients of determination and their statistical significance are shown.

Fig. 5.

Correlations of annual standardized basal area increments (BAI) with the monthly climatic variables: (A) mean temperature (Tmean); (B) precipitation (Prec); (C) cloud cover (Cld); and (D) potential evapotranspiration (PET) for males (black bars) and females (grey bars) over the period 1942–2001. Dashed lines indicate P < 0.05 and dotted lines P < 0.01. Jan to Dec refers to the months of the year, from January to December; (−1) indicates the year before the trees grew.

Fig. 6.

Correlation (r) of the standardized BAI chronologies of female (A) and male (B) Ilex aquifolium trees in Sierra de Ancares explained by monthly standardized precipitation–evaporation index (SPEI) series at different time scales for the period 1942–2001. The analyses were calculated from September of the previous year to December of the current growth year (previous year months are depicted in capitals). Insets in the graphs (dotted lines and black diamonds) represent the month and time lag at which the maximum correlation occurs.