Temporal patterns of seed germination in early spring-flowering temperate woodland geophytes are modified by warming

Abstract

Background and aims: Understorey species in temperate deciduous woodlands such as wild daffodil (Narcissus pseudonarcissus) and common snowdrop (Galanthus nivalis) have complex dormancy: seeds that are shed in late spring require warm summer temperatures for embryo elongation and dormancy alleviation, but then cooler temperatures for germination in autumn. As seasons warm and tree canopies alter, how will different seasonal temperature sequences affect these complex dormancy responses?

Methods: The effect of different sequences of warmer (+5 °C), current or cooler (-5 °C) seasons (summer to spring) on seed germination patterns over seven successive seasons were investigated, with all sequences combined factorially to determine the consequences of differential seasonal temperature change for the temporal pattern of germination (and so seedling recruitment).

Key results: Little (<1 %, G. nivalis) or no (N. pseudonarcissus) seed germination occurred during the first summer in any treatment. Germination of N. pseudonarcissus in the first autumn was considerable and greatest at the average (15 °C) temperature, irrespective of the preceding summer temperature; germination was also substantial in winter after a warmer autumn. Germination in G. nivalis was greatest in the warmest first autumn and influenced by preceding summer temperature (average > warmer > cooler); the majority of seeds that germinated over the whole study did so during the two autumns but also in year 2's cooler summer after a warm spring.

Conclusions: Warmer autumns and winters delay first autumn germination of N. pseudonarcissus to winter but advance it in G. nivalis; overall, warming will deplete the soil seed bank of these species, making annual seed influx increasingly important for recruitment and persistence. This study provides a comprehensive account of the effects of temperature changes in different seasons on seed germination in these early spring-flowering geophytes and consequently informs how these and other temperate woodland species with complex seed dormancy may respond to future climate change.

Keywords: Galanthus nivalis L; Narcissus pseudonarcissus L; Amaryllidaceae; climate change; daffodil; geophyte; morphophysiological dormancy; seed germination; snowdrop; temperate woodland; temperature.

Fig. 1.

Cumulative seed germination (%, n = 50 seeds) of Narcissus pseudonarcissus under a (A) cool (5 °C), (B) average (10 °C) or (C) warm (15 °C) spring with all possible combinations of summer [cool (15 °C), average (20 °C) or warm (25 °C)], autumn [cool (5 °C), average (10 °C) or warm (15 °C)] and winter [cool (0 °C), average (5 °C) or warm (10 °C)] temperatures over 616 d with 84 or 112 d duration of the seasons. See the Materials and Methods for durations of the seasons. Year 1 (1 in column) comprised four seasonal periods from summer onwards; year 2 (2 in column) ended at the end of winter. The control (average seasonal temperatures) result is highlighted by light grey shading (centre of figure).

Fig. 2.

Cumulative seed germination (%) of Narcissus pseudonarcissus at the end of summer (nt = 27 treatments), autumn (nt = 9 treatments) and winter (nt = 3 treatments) in year 1 for average (A), cooler (C) or warmer (W) seasonal temperatures (T). Data are from Fig. 1. Germination was not significantly different (∙) or was significantly lower (−) or higher (+) than the control (logistic regression, P = 0.05). The control (average seasonal temperatures) result is highlighted by light grey shading (centre of figure).

Fig. 3.

Cumulative seed germination (%, n = 50 seeds) of Galanthus nivalis under a (A) cool (5 °C), (B) average (10 °C) or (C) warm (15 °C) spring with all possible combinations of summer [cool (15 °C), average (20 °C) or warm (25 °C)], autumn [cool (5 °C), average (10 °C) or warm (15 °C)] and winter [cool (0 °C), average (5 °C) or warm (10 °C)] temperatures over 588 d with 84 d duration of seasons. Note the restricted y-axis. Year 1 (1 in column) comprised four seasonal periods from summer onwards; year 2 (2 in column) ended at the end of winter. The control (average seasonal temperatures) result is highlighted by light grey shading (centre of figure).

Fig. 4.

Cumulative seed germination (%) of Galanthus nivalis at the end of summer (nt = 27 treatments), autumn (nt = 9 treatments) and winter (nt = 3 treatments) for average (A), cooler (C) or warmer (W) seasonal temperatures (T). Data are from Fig. 3. Germination was not significantly different (∙) or was significantly lower (−) or higher (+) than the control (logistic regression, P = 0.05). The control (average seasonal temperatures) result is highlighted by light grey shading (centre of figure).