Local Snow Melt and Temperature—But Not Regional Sea Ice—Explain
Total Page:16
File Type:pdf, Size:1020Kb
Received: 27 September 2018 | Revised: 18 January 2019 | Accepted: 3 March 2019 DOI: 10.1111/gcb.14639 PRIMARY RESEARCH ARTICLE Local snow melt and temperature—but not regional sea ice— explain variation in spring phenology in coastal Arctic tundra Jakob J. Assmann1 | Isla H. Myers‐Smith1 | Albert B. Phillimore1 | Anne D. Bjorkman2 | Richard E. Ennos1 | Janet S. Prevéy3 | Greg H. R. Henry4 | Niels M. Schmidt5 | Robert D. Hollister6 1The University of Edinburgh, Edinburgh Abstract 2Senckenberg Biodiversity and Climate Research Centre, Frankfurt, Germany The Arctic is undergoing dramatic environmental change with rapidly rising surface 3Pacific Northwest Research Station, temperatures, accelerating sea ice decline and changing snow regimes, all of which Department of Agriculture – Forest Service, Olympia, Washington influence tundra plant phenology. Despite these changes, no globally consistent di- 4University of British Columbia, Vancouver, rection of trends in spring phenology has been reported across the Arctic. While BC, Canada spring has advanced at some sites, spring has delayed or not changed at other sites, 5Arctic Research Centre, Aarhus University, highlighting substantial unexplained variation. Here, we test the relative importance Roskilde, Denmark 6Grand Valley State University, Allendale, of local temperatures, local snow melt date and regional spring drop in sea ice extent Michigan as controls of variation in spring phenology across different sites and species. Trends Correspondence in long‐term time series of spring leaf‐out and flowering (average span: 18 years) Jakob J. Assmann, The University of were highly variable for the 14 tundra species monitored at our four study sites on Edinburgh, Edinburgh, UK. Email: [email protected] the Arctic coasts of Alaska, Canada and Greenland, ranging from advances of 10.06 days per decade to delays of 1.67 days per decade. Spring temperatures and Funding information Arctic Institute of North America; National the day of spring drop in sea ice extent advanced at all sites (average 1°C per decade Science Foundation, Grant/Award Number: and 21 days per decade, respectively), but only those sites with advances in snow 9714103 0632263 0856516 1432277 1504224; Government of Canada; melt (average 5 days advance per decade) also had advancing phenology. Variation in University of British Columbia; Canadian spring plant phenology was best explained by snow melt date (mean effect: 0.45 days Network for Research and Innovation in Machining Technology, Natural Sciences and advance in phenology per day advance snow melt) and, to a lesser extent, by mean Engineering Research Council of Canada; spring temperature (mean effect: 2.39 days advance in phenology per °C). In contrast Northern Scientific Training Program; Natural Environment Research Council, to previous studies examining sea ice and phenology at different spatial scales, re- Grant/Award Number: NE/L002558/1 and gional spring drop in sea ice extent did not predict spring phenology for any species NE/M016323/1 or site in our analysis. Our findings highlight that tundra vegetation responses to global change are more complex than a direct response to warming and emphasize the importance of snow melt as a local driver of tundra spring phenology. KEYWORDS Arctic tundra, climate change, phenology, sea ice, snow melt, spring, temperature, vegetation 2258 | © 2019 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/gcb Glob Change Biol. 2019;25:2258–2274. ASSMANN ET AL. | 2259 1 | INTRODUCTION 1.3 | Temperature as a driver 1.1 | The importance of phenology and global The environmental variable most widely used to explain variation change in spring phenological events across latitudes and seasons is tem- perature (Post, Steinman, & Mann, 2018; Thackeray et al., 2016), Changing phenology is considered one of the most apparent effects including the phenology of both Arctic and alpine tundra plants of climate change on natural systems worldwide (Cleland, Chuine, (Bjorkman et al., 2015; Huelber et al., 2006; Iler et al., 2017; Kuoo & Menzel, Mooney, & Schwartz, 2007; IPCC, 2014; Menzel et al., 2006; Suzuki, 1999; Molau et al., 2005; Oberbauer et al., 2013; Panchen & Parmesan & Yohe, 2003). Phenological processes control ecosys- Gorelick, 2017; Prevéy et al., 2017; Thórhallsdóttir, 1998; Wheeler tem functions (Ernakovich et al., 2014; Richardson et al., 2013), are et al., 2015). Temperature influences phenology through increasing linked through feedbacks to the climate system (Richardson et al., plant metabolism and development in response to warmer ambi- 2013) and contribute to structuring food webs through trophic ent temperatures (Jones, 2013). Average temperatures over a pre- interactions (Kharouba et al., 2018; Visser & Both, 2005). In high‐ defined period (Bjorkman et al., 2015; Iler et al., 2017; Panchen & latitude ecosystems, the onset of plant growth in spring and senes- Gorelick, 2017; Prevéy et al., 2017) as well as cumulative tempera- cence in autumn is linked with ecosystem net productivity (Forkel tures up to the onset of a phenological event (Barrett, Hollister, et al., 2016; Park et al., 2016; Piao et al., 2008; Xu et al., 2013) and Oberbauer, & Tweedie, 2015; Henry & Molau, 1997; Huelber et al., food availability for herbivores (Barboza, Van Someren, Gustine, & 2006; Kuoo & Suzuki, 1999; Molau et al., 2005; Oberbauer et al., Bret‐Harte, 2018; Doiron, Gauthier, & Lévesque, 2015; Gustine et 2013; Wheeler et al., 2015) have been shown to explain variation in al., 2017; Kerby & Post, 2013a, 2013b; Post et al., 2008). Varying Arctic and alpine plant phenology. Species‐specific minimum heat phenological responses to environmental drivers among species or energy requirements for phenological progress have been suggested taxa, particularly in the highly seasonal Arctic tundra, yield a high for tundra plants (Huelber et al., 2006; Molau et al., 2005), and the potential for phenological mismatch (Doiron et al., 2015; Kerby & sensitivity of flowering to temperature has been shown to vary Post, 2013b; Piao et al., 2008) and shorter flowering seasons with between sites and plots within tundra plant species (Høye, Post, warming (Prevéy et al., 2019). Tundra plants are temperature sen- Schmidt, Trøjelsgaard, & Forchhammer, 2013; Prevéy et al., 2017). sitive, especially at high latitudes (Prevéy et al., 2017), but no net However, in highly seasonal tundra ecosystems, temperature is only advance in leaf or flowering phenology has been observed across one factor determining spring plant phenology. the biome (Bjorkman, Elmendorf, Beamish, Vellend, & Henry, 2015; Oberbauer et al., 2013; Post, Kerby, Pedersen, & Steltzer, 2016) de- spite Arctic surface temperatures rising at twice the global aver- 1.4 | Snow melt as a driver age (IPCC, 2014; Winton, 2006). Instead, a more complex picture is Snow melt timing has been recognized as early as the 1930s as a emerging, highlighting a considerable amount of unexplained vari- primary initiator of plant phenological events in both Arctic and ation in phenology across sites, species and phenological events alpine tundra (Schwartz, 2013; see, e.g., Billings & Mooney, 1968; (Bjorkman et al., 2015; Oberbauer et al., 2013; Post & Høye, 2013; Sørensen, 1941; Wipf & Rixen, 2010), and many recent studies Post et al., 2016; Prevéy et al., 2017). have demonstrated that snow melt date is a key driver explain- ing variation in spring phenology in tundra ecosystems (Bjorkman et al., 2015; Cooper, Dullinger, & Semenchuk, 2011; Cortés et al., 1.2 | Variation in plant phenology—what controls it? 2014; Iler et al., 2017; Molau et al., 2005; Semenchuk et al., 2016; A detailed understanding of which environmental variables serve as Sherwood, Debinski, Caragea, & Germino, 2017; Wipf, 2009; Wipf, cues for Arctic spring phenology is key for explaining the absence Stoeckli, & Bebi, 2009; but see Thórhallsdóttir, 1998). During snow of an overall trend in phenology across sites despite rapid warming melt, tundra plants experience dramatic changes in their immediate and is critical for predicting future responses of Arctic ecosystems environment: light availability increases and plant and soil surfaces to the effects of climate and environmental change (Richardson et are exposed to atmospheric temperatures and CO2 concentrations al., 2013). Interannual variation in tundra phenology has been attrib- (Starr & Oberbauer, 2003), which in turn stimulate plant meta- uted to variation in temperature (Bjorkman et al., 2015; Iler, Inouye, bolic and developmental activities (Jones, 2013). In addition, snow Schmidt, & Høye, 2017; Molau, Nordenhäll, & Eriksen, 2005; melt may act as an indicator for suitable growing conditions to Oberbauer et al., 2013; Panchen & Gorelick, 2017; Prevéy et al., come as the growing season advances (Wheeler et al., 2015). Prior 2017; Wheeler, Høye, Schmidt, Svenning, & Forchhammer, 2015), to melt, the insulation of the snow layer protects the plants from snow melt (Bjorkman et al., 2015; Iler et al., 2017; Semenchuk et al., frost damage, desiccation and photoinhibition (Lundell, Saarinen, & 2016) and sea ice (Kerby & Post, 2013a; Post et al., 2016). To date, Hänninen, 2010; Mølgaard & Christensen, 2003; Sherwood et al., no study has combined all three environmental variables to test the 2017; Wheeler et al., 2015; Wipf & Rixen, 2010; Wipf et al., 2009) degree to which local snow melt, temperature and regional sea ice and reduces early‐season herbivory (Wheeler et al., 2016), while melt influence spring