Rapidly changing climatic conditions for wine grape growing in the Okanagan Valley region of British Columbia, Canada

Science of the Total Environment 556 (2016) 169–178

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Rapidly changing climatic conditions for wine grape growing in the Okanagan Valley region of British Columbia, Canada Sierra Rayne a,⁎, Kaya Forest b a b

Chemologica Research, 1617-11th Avenue NW, Moose Jaw, Saskatchewan S6H 6M5, Canada Department of Environmental Engineering Technology, Saskatchewan Polytechnic, 600 Saskatchewan Street West, Moose Jaw, Saskatchewan S6H 4R4, Canada

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Long-term climate changes examined for sites around Okanagan Lake, British Columbia, Canada • Average wine grape growing season temperatures increasing rapidly since 1980s • Dormant season temperatures also rising quickly in post-1970 period • Large increases in growing degree days over past few decades • Formerly cool-climate viticulture regions now moving into intermediate classification

a r t i c l e

i n f o

Article history: Received 3 January 2016 Received in revised form 27 February 2016 Accepted 28 February 2016 Available online 11 March 2016 Editor: D. Barcelo Keywords: Climate change Viticulture Grape growing Wine production Agriculture

a b s t r a c t A statistical analysis was conducted on long-term climate records for sites bordering Okanagan Lake in the Okanagan Valley viticultural region of British Columbia, Canada. Average wine grape growing season temperatures are increasing rapidly in the area over the post-1980 period at rates upwards of 7.0 ± 1.3 °C/century. Similar increases in the average dormant season temperature are evident. These temperature changes are likely some of the most extreme observed among the world's wine producing areas during the past few decades. Growing degree day base 10 °C (GDD10) has increased by nearly 50% at some locations since the 1970s, resulting in major impacts on the corresponding climate classification for viticulture. If current climate trends continue, the southern and central portions of the region will likely enter Winkler region II within the next few decades, placing them in the same category as well-established warmer wine regions from France, Spain, Italy, and Australia. The large dormant season temperature increases over the last several decades have resulted in the area no longer being a cold season outlier when compared to most other cool-climate viticultural areas. Based on average growing season temperatures, the southern end of Okanagan Lake has moved out of the cool-climate viticultural classification and into the intermediate zone, while the central and northern regions are now at the cool/ intermediate viticulture interface, similar to the historical positions of the Rhine Valley in Germany, northern Oregon in the United States, and the Loire Valley, Burgundy-Cote, Burgundy-Beaujolais, and Champagne appelations of France. The corresponding suitable grape species for the area have evolved into warmer region varietals during this time frame, having substantial economic impacts on producers. Increased temperatures are

⁎ Corresponding author. E-mail address: [email protected] (S. Rayne).

http://dx.doi.org/10.1016/j.scitotenv.2016.02.200 0048-9697/© 2016 Elsevier B.V. All rights reserved.

170

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

also expected to bring greater threats from agricultural pests, notably Pierce's disease from the bacterium Xylella fastidiosa. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Concerns over global climate change have led to a critical examination of its potential impacts on wine production during the past several decades. The majority of the world's high quality wine producing regions have seen growing season average temperature increases of about 2 °C over the past century (Jones, 2004), with average winter and summer temperatures increasing 1.3 and 1.5 °C, respectively, between 1950 and 2000 (Fraga et al., 2012; Jones et al., 2005a, 2005b; Schultz and Jones, 2010). The warming trends in recent decades have resulted in wine grapes maturing more rapidly in North America, Europe, and Australia (Duchêne and Schneider, 2005; Duchne et al., 2010; Etien et al., 2008; Jones and Davis, 2000a; Petrie and Sadras, 2008; Sadras and Petrie, 2011; Wolfe et al., 2004). A study of climate and observed changes in European wine grape phenology (Jones et al., 2005a) showed that growing season average temperatures increased by 1.7 °C, maximum temperatures increased by 1.8 °C, minimum temperatures increased by 1.9 °C, with more warming in the spring and summer, and an average increase in growing degree day base 10 °C (GDD10) of 264 across all stations. Between 1972 and 2002, average annual temperatures increased 1.8 °C in the Alsace region of France, where warming trends were highest during the grape ripening phase resulting in 33 more days above 10 °C and harvest dates occurring two weeks earlier (Duchêne and Schneider, 2005). Other work from the Bordeaux region of France found harvest dates were 13 days earlier in 1997 than in 1952 due to regional warming trends (Holland and Smit, 2010; Jones and Davis, 2000a). During the late 1990s and 2000s, mean annual temperatures in the area around Baden, Germany were 1.2 °C higher than the 1961–1990 average, leading to the advancement of maturation by 3 weeks between 1976 and 2006. Similarly, average temperatures increased 1.2 °C from 1970 to 2005 in the Palatinate region of Germany, concommitant with the harvest date moving forward by 2 weeks (Mira de Orduña, 2010). From 1951 to 1997, the wine grape producing regions of coastal California experienced an annual average temperature increase of 1.1 °C which advanced the start of the growing season by 18–24 days (Nemani et al., 2001). In the Veneto region of Italy, average growing season temperatures increased 2.3 °C between 1964 and 2009, with bloom, veraison, and harvest dates moving forward from 13 to 19 days (Tomasi et al., 2011). An eight day earlier harvest was associated with a 1 °C warmer vintage. South Africa's wine regions have also seen significant warming trends throughout the year, on the order of 1 °C from 1967 to 2000 and 0.6 to 1.8 °C between the 1960s/1970s and 1990s/2000s, depending on region and timeframe considered, and an increase in growing season mean temperature for the Stellenbosch district of 0.8 °C from 1967 to 2010 (Vink et al., 2012). Maturation dates advanced eight days per decade between 1985 and 2009 in southern Australia and between 1972 and 2004 in France, as well as four days per decade in Geisenheim, Germany over the period from 1955 to 2004 (Webb et al., 2011, 2012). In the Rheingau region of Germany, harvest dates are now two to three weeks earlier than during the late 18th century through the early 20th century (Stock et al., 2005). For the Chateauneuf du Pape and Tavel regions of southern France, harvest dates moved forwards 18–21 days between 1945 and 2000 (Mira de Orduña, 2010). Warmer temperatures are leading to higher sugar accumulations in the berry over the course of the growing season, which translates into higher alcohol concentrations in the finished wine after fermentation. Approximately half of the increased alcohol contents in wines from Alsace (France), Australia, and the Napa Valley (California, USA) has been attributed to climate change (Jones, 2007). Over the period from 1980

to 2001, grape potential alcohol near harvest date increased substantially in southern France, with corresponding decreases in acidity and a higher grape pH (Mira de Orduña, 2010). Maturity normalized for sugar advanced between one-half day up to more than three days per year between 1993 and 2006 in Australia. (Petrie and Sadras, 2008). Asymmetric warming (at night and during the spring), leading to reduced frost occurrence, a longer growing season, and more rapid and advanced spring growth, has been linked to increased yields and quality in California. Average spring-time warming is occurring at twice the rate seen over the rest of the year in this region (Nemani et al., 2001). In general, grapevine phenological timing in Europe over the past 50 years is occurring earlier and with more significant changes being observed in later events than in earlier events. Strong correlations between grapevine phenology and climate parameters were observed, with maximum temperatures being more important for early season events (i.e., budbreak and bloom), and average temperatures and GDD10 being more important for later season events (i.e., veraison and harvest) (Jones, 2005a). Vintage ratings over the past 50 years in most of the world's best wine regions have increased, which has been found to be correlated to increases in growing season temperatures (Jones, 2005b). In the United States, extreme heat impacts may reduce premium grape production by up to 80% in the late 21st century. These heat accumulation increases are predicted to shift production towards warmer climate varieties and/or lower-quality wines while reducing the impacts of frost damage, but the increase in extreme hot days (N35 °C) may eliminate wine grape production across much of the United States (White et al., 2006). A study of the impacts of climate change on the industry in California showed that grape ripening is predicted to occur one to two months earlier and at higher temperatures by the year 2100 (Hayhoe et al., 2004). More recent work suggests that, at the global scale, climate change under the higher RCP 8.5 greenhouse gas concentration pathway will result in a 25 to 73% decline in the area suitable for viticulture, and a 19 to 62% decline under the lower RCP 4.5 pathway, within the major wine producing regions (Hannah et al., 2013). Furthermore, problems with earlier harvest dates brought on by climate change, assuming that grape development has still maintained a balance between sugar, acidity, and flavor, include hot and desiccated fruit if greater irrigation inputs are not applied (Jones, 2005b). Since many of the world's grape growing regions are located in semiarid environments where water scarcity is a major scientific and political issue, increases in irrigation demand to offset the harvesting of hot, desiccated fruit may not be possible. Within this global context, work has only begun to examine the potential climate change effects on viticulture for the emerging wine grape regions in the Canadian province of British Columbia (Belliveau et al., 2006). The economic importance of the Okanagan Valley wine industry to the regional and provincial economy is now valued at several billion dollars per annum, and the region is located within the current effective northern limits of Vitis vinifera wine grape production. Herein we report that portions of this region have undergone – and continue to experience – rapid climatic change over the past several decades at rates likely exceeding those of most other viticultural regions worldwide. 2. Regional description The Okanagan Valley is located in south-central British Columbia, Canada, approximately 300 km east of the Pacific Ocean. The valley is long and narrow and runs northward for 160 km from the US border at 49 to 50°N latitude (Fig. 1). The region lies in a rain shadow between two north-south trending mountain ranges, resulting in low annual

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

171

Fig. 1. Map of the study area showing locations of cities (open squares), viticultural regions, and clusters of vineyards (filled circles) in the study area. Note that all vineyards are not shown for clarity; major groupings are indicated to demonstrate regions of concentrated viticultural activity.

average precipitation (from ca. 320 mm/y at Osoyoos [southern end of the Canadian portion of the valley] to ca. 480 mm/y at Vernon [northern end of the valley]) that is distributed evenly throughout the year, and only modest winds are typical. The area between Oliver and the US border is the northern-most tip of the Sonora Desert, which begins at the Baja Peninsula in Mexico. Summers are generally hot, with average daily temperatures in July and August (ca. 20–22 °C) that are as warm as, and in some cases warmer than, the Napa Valley (California, USA). Daytime maximum temperatures over this period can reach 40 °C, and are often above 30 °C for several consecutive days. In the summer, there are long daylight hours and high light intensity due to the northerly latitude, which helps with prolonged daytime photosynthesis and grape ripening. Two primary modes of natural climatic variability affect the climate of the Okanagan Valley. Both the El Nino/ Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO) have periods of anomalous warming and cooling of the Pacific Ocean, which influence atmospheric circulation. El Nino and its warming of the tropical waters of the central Pacific decreases the strength of the easterly trade winds. The opposite phase of ENSO, La Nina, has a cooling effect on the central Pacific that increases these trade winds. Similarly, the PDO is an alternate warming and cooling of the ocean, but is located in the northern Pacific. During the warm phases of the ENSO and PDO, winters are hotter and drier, with correspondingly colder and wetter winters during the cool phases of these patterns (Mantua et al., 1997; Trenberth and Hurrell, 1994; Wallace and Gutzler, 1981). The ENSO

typically acts on an interannual timescale, with recent strong events during the winters of 1982/1983, 1986/1987, and 1997/1998. In contrast, the PDO (http://research.jisao.washington.edu/pdo/PDO.latest) operates on inter-decadal timescales with at least the following four phase changes since 1900: 1900 to 1925, cold; 1926 to 1946, warm; 1947 to 1976, cold; and 1977 to the late 1990s, warm (Barnett et al., 1999; Latif and Barnett, 1994; Minobe, 1997; Nigam et al., 1999; Trenberth, 1990; Trenberth and Hurrell, 1994; Zhang et al., 1997). During the growing season, there can be a 4 °C average daily difference in temperature between the northern and southern ends of the valley. This results in a preference for red varieties in the south and white varieties in the cooler north (Bowen et al., 2005). Winters are generally cold and temperatures can drop below freezing for long periods, with rare events down to − 25 °C. The valley's extensive lakes are key to moderating the otherwise mountainous/continental climate extremes in summer and winter. However, the main climatic factor limiting V. vinifera grape production in the Okanagan Valley are the low temperatures (critical value range, b−6 °C to b−23 °C) occurring in late October through February (Caprio and Quamme, 2002). As of 2014, there were 10,260 ac of wine grapes under production in British Columbia (BCWI, 2014). Of this, 84.1% (8619 ac) is located in the Okanagan Valley (744 vineyards and 163 wineries), with lesser proportions in the Similkameen Valley (6.4%) to the immediate west of the Okanagan Valley, coastal areas (6.6%) such as Vancouver Island and the Lower Mainland around Vancouver, and other interior regions of

172

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

the province. A summary of the number of vineyards, planted acreage and rate of acreage growth in recent years, and the major wine grape varieties by region within the study area is provided in Table 1. The majority of wine grape production in the Okanagan Valley occurs at the southern end near the Canada-United States border around the communities of Oliver and Osoyoos. There are progressively higher rates of increasing vineyard acreage moving south to north in the valley, with effectively all of the industry expansion in recent years taking place along the margins of Okanagan Lake between Penticton and Vernon. Major varieties within each sub-region reflect a dominant climatic influence, shifting to increasing contributions from cooler-climate varietals (e.g., Pinot Gris, Pinot Noir, Riesling, and Gewurztraminer) moving north up the valley. As the climate has warmed in recent decades, there have been increasing plantings of warmer-climate varietals (e.g., Merlot, Cabernet Sauvignon, Cabernet Franc) further north into regions that were previously too cold to adequately ripen these varieties. 3. Methods Long-term climate data for each station was obtained from the Adjusted and Homogenized Canadian Climate Data (AHCCD) website (http://www.ec.gc.ca/dccha-ahccd/) of Environment and Climate Change Canada. Data used was the “Second Generation of Homogenized Temperature” datasets. In the “First Generation of Homogenized Temperature” datasets, non-climatic shifts were identified in the annual means of the daily maximum and minimum temperatures using a technique based on regression models (Vincent, 1998) and were primarily the result of station relocations, observing practice changes, and automation (Vincent and Gullett, 1999). Identified shifts were adjusted via application to monthly and daily maximum and minimum temperatures (Vincent et al., 2002). For the “Second Generation of Homogenized Temperature” datasets, new adjustments were applied to the daily minimum temperatures in order to address the bias due to a change in observing time (Vincent et al., 2009). In addition, regression models were employed to find non-climatic shifts in temperature monthly series (Vincent, 1998; Wang et al., 2007) and a quantile-matching algorithm was applied to derive adjustments (Vincent et al., 2012; Wang et al., 2010).

Each time-series was analysed for first-order autocorrelation using the Durbin-Watson statistic (Durbin and Watson, 1950, 1951; Watson and Durbin, 1951). No autocorrelation problems were observed. Possible trends were examined using both parametric and non-parametric (Spearman and Kendall rank correlations) linear regression models using the KyPlot v2.0b15 software package (Yoshioka, 2002). For the wine grape industry, climate is a dominant contributor to quantity and quality (Maltman, 2008; van Leeuwen et al., 2004; White et al., 2009). Major climatic conditions required for producing wine grapes with balanced composition and varietal typicity include low frost damage in mild winters; early and even budbreak, flowering, and development during warm springs; and optimal maturation with low summer temperature variability, adequate heat accumulation, and a lack of extreme heat (Gladstones, 1992; Nemani et al., 2001; White et al., 2006). For the northern hemisphere, the four seasonal stages of grapevine development, together with the associated general climate influences, can be summarized as follows (Nemani et al., 2001). Stage 1 includes budbreak and leaf development, begins between midMarch to the first week of April, and requires sufficient soil moisture and sunshine with temperatures N10 °C. Frost or freeze occurrence during this stage can reduce bud fruitfulness, leading to poor yields and quality. In stage 2 (floraison/flowering) during early through midJune, dry, stable conditions are needed to avoid delaying flower growth. Stage 3 (veraison, development of the grape, and its maturation) typically begins between the end of July and the first week of August, where dry conditions with moderate temperatures are preferred to limit moisture induced grape rot. In addition, high levels of insolation and low temperature variability at this time help concentrate sugars before harvest. Finally, during stage 4 which begins between late September through mid-October (harvest through dormancy), the vine requires sufficiently cold temperatures to initiate latent bud hardening with limited freeze damage. While these four stages describe preferred seasonal climatic conditions, daily variation within each stage also helps contribute to optimum wine grape quality. Assessing viticultural suitability with mean climatological parameters often neglects daily weather events that significantly influence the quantity and quality of harvest (Jones and Davis, 2000b).

Table 1 Summary of viticultural activities by region within the study area. Data from BCWI (2014). Region

No. of vineyards

Planted acres

Acreage growth rate since 2011

Main varieties by acreage

Osoyoos

108

1545

+2%

Oliver

203

3639

+3%

Okanagan Falls

35

594

+10%

Penticton (incl. Kaleden)

222

1144

+13%

Peachland/Summerland

59

405

+14%

Kelowna (incl. West Kelowna)

85

1037

+15%

Vernon (incl. Lake Country and Spallumcheen/Shuswap)

46

351

+30%

Merlot (29.2%) Cabernet Sauvignon (18.6) Syrah (12.5%) Cabernet Franc (8.7%) Merlot (19.0%) Chardonnay (12.5%) Pinot Gris (11.4%) Cabernet Sauvignon (9.5%) Chardonnay (17.0%) Pinot Noir (15.6%) Gewurztraminer (15.5%) Pinot Gris (14.0%) Merlot (16.5%) Pinot Noir (12.6%) Pinot Gris (12.2%) Chardonnay (9.9%) Gewurztraminer (20.6%) Pinot Noir (19.7%) Pinot Gris (19.4%) Chardonnay (11.4%) Pinot Noir (25.0%) Gewurztraminer (15.4%) Riesling (12.8%) Pinot Gris (12.1%) Pinot Gris (16.1%) Pinot Noir (15.8%) Riesling (10.6%) Gewurztraminer (8.1%)

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

For example, maximum daily temperatures during the summer months can influence grapevine growth both positively and negatively. Shortterm periods with temperatures N30 °C can assist ripening potential, but extended periods can cause heat stress in the vine, premature veraison, berry abscission, enzyme inactivation, and reduced flavour development (Mullins et al., 1992). Similarly, while there is general agreement in using 10 °C as the base viticultural temperature (i.e., GDD10) for total growing season heat units when comparing different climatic regions, and for assessing a region's suitability for grapevine production (Oliveira, 1998; Shaw, 2001, 2005), upper limits on heat summation ranging from 19 °C to 21 °C have been used because of the negative effect of sustained higher average temperatures on wine quality (Gladstones, 1992; Winkler et al., 1974). We do note, however, that some have advocated caution in the use of relating GDD10 to vine physiology, production, and quality in some regions (Jones and Davis, 2000a, 2000b). In the maturation stage of late summer and early autumn, significant diurnal temperature ranges increase both sugar and tannin production in the fruit, helping preserve the balance for wine production (Gladstones, 1992). These lower autumn temperatures are important in the production of quality table wines, ideally with prolonged cool temperatures before the first frost. Cool temperatures at this time of the late growing season, just before the mean monthly temperature decreases to b 10 °C, slow berry development and allow for a better balance of sugar and acidity and increased flavour and aroma constituents (Gladstones, 1992; Shaw, 2001). A minimum temperature threshold also exists during the winter months, where effective chilling units are required to ensure uniform budbreak in the spring (Jones, 2005b). Grape hardiness generally increases through the winter, and daytime temperatures b−9 °C in late November and early December increase grape production, likely due to the prevention of vine deacclimation (Caprio and Quamme, 2002). However, most V. vinifera cultivars are injured at temperatures b−20 °C, and temperatures b−12 °C before mid-November can result in freeze damage, particularly to the primary buds. At b−25 °C, even matured canes and young vines can be severely damaged, and at b−29 °C, entire vines may be killed (Gladstones, 1992). Northern hemisphere grape growing periods were defined as follows (White et al., 2006): growing season, April 1 to October 31; dormant season, November 1 to March 31. Average growing and dormant season temperatures were obtained by taking a simple unweighted average of mean monthly temperatures for each season. Monthly GDD10 were calculated by subtracting 10 °C from the mean monthly temperature and multiplying by the number of days in the month for each month in the growing season. Negative values were assigned as zero, and the monthly GDD10 summed to yield a corresponding growing season total GDD10. GDD10 measures heat accumulation by the grapevine over the course of a growing season, and – absent disease, drought, or other extreme conditions – the vine grows in a cumulative and sequential manner with various developmental and maturation benchmarks that can be predicted by following GDD10. Growth stages, susceptibility to pests, and the dates of fertilizer and pesticide applications, pruning, thinning, and other management activities can be estimated by the use of correlations with GDD10. Individual grape varieties each have their own unique GDD10 benchmarks throughout the growing season, thereby allowing the use of GDD10 for assessing the suitability of a site for a particular varietal and the effects of a changing climate on the corresponding varietal suitability over time. The latitude–temperature index (LTI) was calculated as follows: LTI = mean temperature of the warmest month × (60-latitude [in degrees]). The LTI is employed to assess the capacity of a region to ripen wine grapes (Jackson and Cherry, 1988) and to assess varietal suitability (Kenny and Shao, 1992). Individual varietals are generally found to have optimal LTI indices within which they will produce the highest quality wine grapes.

173

4. Results and discussion Historical trends in average growing season (GST,avg) and dormant season (DST,avg) temperatures were examined for the three long-term AHCDD climate stations (Penticton, southern end; Kelowna, approximate mid-point; and Vernon, northern end) surrounding Okanagan Lake (Table 2). Since 1980, there have been rapid increases in GST,avg at each site (Fig. 2), particularly at Kelowna (Table 3), whereas no significant trends in GST,avg existed at any of the sites for the 1900–1980 period. While the post-1980 rates of increase in GST,avg are substantial at Penticton (2.4 ± 1.2 °C/century) and Vernon (1.8 ± 1.2 °C/century), Kelowna is experiencing an extreme rate of temperature increase during the past 35 years (7.0 ± 1.3 °C/century). During the dormant season, only Kelowna displays a significant increase in DST,avg over the 1900–1970 period (1.9 ± 1.0 °C/century), with no corresponding significant trends at the southern and northern ends of Okanagan Lake. However, in the post-1970 period, Kelowna's DST,avg is increasing at very high rates on the order of 7.3 ± 1.9 °C/century, far outpacing the still substantial increasing DST,avg trends from 1970 to 2014 at Penticton (2.5 ± 1.5 °C/century) and Vernon (2.4 ± 1.7 °C/century). For context, the Northern Hemisphere land temperature anomaly (https://www.ncdc.noaa.gov/cag/) rate of increase since 1980 for the April to October growing season period has been 3.2 ± 0.3 °C/century, while the North American and European land temperature anomalies increased over this timeframe at rates of 2.7 ± 0.6 and 4.4 ± 0.5 °C/century, respectively. The November to March dormant season rate of increase in the Northern Hemisphere, North American, and European land temperature anomalies from 1970 to 2014 were 3.1 ± 0.3, 3.6 ± 0.8, and 2.5 ± 0.8 °C/century, respectively. Consequently, the central Okanagan Valley's corresponding rate of change is up to approximately three-fold higher than that of the North American, Northern Hemispheric, and European averages. In addition, the California climate divisions (1 and 2) encompassing and adjacent to the Napa Valley wine region are only exhibiting – albeit still substantial – post-1980 GST,avg increasing trends of 1.9 ± 0.8 and 2.1 ± 0.8 °C/century, respectively, and post-1970 DST,avg increasing trends of 1.6 ± 0.8 and 1.7 ± 0.9 °C/century, respectively. The cause of the rapidly rising temperatures in the study area, especially the Central Okanagan region near Kelowna, is not known. We also examined GST,avg and DST,avg trends since 1980 and 1970, respectively, at a number of other AHCCD climate stations throughout southcentral British Columbia (Table 4). The south Okanagan station at Oliver, which is the largest viticultural area in the province, is also experiencing a large GST,avg increase in recent decades (5.8 ± 1.0 °C/century) but a more modest DST,avg increase (2.6 ± 1.6 °C/century) that is essentially equivalent to the trends at Penticton and Vernon. With the exception of a lower growing season temperature trend at Cranbrook (1.0 ± 1.2 °C/century), all other stations in south-central British Columbia have post-1980 GST,avg and post-1970 DST,avg trends on the order of 3 to 5 °C/century, generally well above the corresponding Northern Hemisphere and North American land temperature anomaly trends. Recent work has shown that the world's lakes are warming rapidly (O'Reilly et al., 2015), and the area represented by the Kelowna climate station is likely more influenced by its proximity to a large lake than any of the other sites examined either within or outside the Okanagan Valley. Consequently, if Okanagan Lake – and some of the other smaller lakes between Kelowna and Vernon – is experiencing rapid warming, this could be responsible for the unusually high temperature trends

Table 2 Details for the climate stations under consideration. Station name

Penticton

Kelowna

Vernon

Station ID Latitude (°N) Longitude (°W) Elevation (masl)

1,126,150 49.47 119.60 344

1,123,993 49.87 119.48 430

1,128,582 50.23 119.20 427

174

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

Fig. 2. Time series for average growing and dormant season temperatures at each climate station. Solid lines are parametric linear regression fits from 1900 to 1980 and 1980–2014 for the growing season and 1900–1970 and 1970–2014 for the dormant season. Dash–dot–dot lines and the shaded regions encompass the 95% confidence interval about the regression fits.

located in this sub-region. Due to the stations at Penticton and Vernon being potentially less influenced by the Okanagan Lake's thermal regime, any lake warming signals would likely be muted at these other sites, thereby explaining their smaller warming trends. However, the absence of long-term water temperature data for lakes in the area prevents us from more definitively determining causation. Correlations between the growing and dormant season PDO indices and the corresponding post-1980 GST,avg and post-1970 DST,avg were also investigated at each site in the study area. With the sole exception of the post-1970 DST,avg (p = 0.04, r = +0.32 against the dormant season PDO index) at Vernon, no other site/temperature combinations displayed a significant correlation (p-values ranging from 0.11 to 0.88) with the PDO index. As a result, variations in the PDO do not appear to play a significant role in the temperature trends throughout the region over the past few decades.

Several decades ago, the three Okanagan sites had DST,avg well below the lower range of major wine producing regions worldwide (Fig. 3). This gap has been closed to such an extent that Penticton now has an average DST,avg equivalent to the Alsace region in France. Kelowna and Vernon are also approaching DST,avg equivalence with the Alsace. During the 1970s, all three Okanagan sites would have been considered cool regions based on GST,avg. Penticton is now ranked as an intermediate climatic region, with Kelowna and Vernon moving up to the coolintermediate interface, placing these locations in the same GST,avg neighborhood as the Rhine Valley in Germany, northern Oregon in the United States, and the Loire Valley, Burgundy-Cote, BurgundyBeaujolais, and Champagne appelations of France. If current rates of temperature increase continue, by the end of the 21st century the central Okanagan could see growing season temperatures rise to such an extent (i.e., N21 °C) that they exceed the current upper boundaries of

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

175

Table 3 Regression statistics for the growing and dormant season average temperature time series at each climate station. Timeframe

Site

Pentictone Kelowna Vernon Penticton Kelowna Vernon

1900–1980

1980–2014

Timeframe

Site

1900–1970

Penticton Kelowna Vernon Penticton Kelowna Vernon

1970–2014

a b c d e

Trend in growing season average temperature p-Valuea

rb

DFc

m (°C/y) [±SE]d

0.33/0.50/0.56 0.35/0.31/0.33 0.87/0.85/0.81 0.053/0.067/0.069 2 × 10−5/4.5 × 10−4/2.2 × 10−4 0.14/0.083/0.078

0.12 0.11 0.018 0.34 0.69 0.26

73 77 79 31 30 33

0.0035 ± 0.035 0.0028 ± 0.0030 0.0005 ± 0.0030 0.024 ± 0.012 0.070 ± 0.013 0.018 ± 0.012

p-Value

r

DF

m (°C/y) [±SE]

0.35/0.29/0.32 0.052/0.045/0.042 0.17/0.18/0.17 0.096/0.16/0.20 4 × 10−4/9 × 10−4/7 × 10−4 0.17/0.19/0.19

0.12 0.24 0.17 0.20 0.54 0.21

61 68 66 42 38 44

0.010 ± 0.011 0.019 ± 0.010 0.014 ± 0.010 0.025 ± 0.015 0.073 ± 0.019 0.024 ± 0.017

Trend in dormant season average temperature

Values presented as parametric linear regression/Spearman rank correlation/Kendall rank correlation. Correlation coefficient. Degrees of freedom. Slope based on parametric linear regression of the form y = mx + b. Climate record starts in 1908 for Penticton.

hot climate viticulture – thereby posing a risk to continued premium wine grape production. On the other hand, even large increases in dormant season temperatures in this region would result in the sites still remaining well within the current range for viticultural activities, although with the lessening of winter damage from cold temperatures would come an increased risk from pests that would be able to better overwinter within the vineyards. Several climate based indices have been developed for wine grape production. One of the earliest indices developed for agriculture is the heat unit concept using GDD10 (Amerine and Winkler, 1944). Correlating the GDD10 with quality wine grape production has led to Winkler heat unit regions, whereby increasing Winkler region numbering (e.g., 1, 2, 3, 4) denotes increased heat units accumulating during the growing season (Winkler et al., 1974). However, the GDD10 based heat unit approach for discriminating among wine grape growing regions has been found to be most accurate for intermediate through hot regions (e.g., southern Europe, California, Australia, and South America), and less successful for analyzing cool through intermediate zones (e.g., northern Europe, northeastern and northwestern North America, and New Zealand). For this reason, the LTI (Jackson and Cherry, 1988) was developed and is generally thought to better discriminate between cool climate viticulture areas (i.e., those in Winkler regions 1 and 2) than the GDD10. Cool climate regions are generally defined as those where the GDD10 is b1390 (Amerine and Winkler, 1944) and/or where the LTI is b 270 (Jackson and Cherry, 1988).

Compared to the 1970s, average decadal GDD10 values increased 30% by the 2000s at Kelowna, 13% at Penticton, and 9% at Vernon (Fig. 4). The first few years of the 2010s have seen a 48% increase in GDD10 for Kelowna above that of the 1970s. Although all three sites remain in Winkler region 1 along with similar sites throughout Canada, the northern United States, Germany, and New Zealand, these large climate shifts suggest that – if current trends continue – Penticton and Kelowna will likely enter Winkler region II (that includes regions from France, Spain, Italy, and Australia) within the next few decades. All three Okanagan sites are now firmly within LTI region B placed near Freiburg, Germany, whereas in the 1970s Kelowna and Vernon were at the interface of LTI regions A and B closer in climatic characteristics to Geisenheim, Germany. The extrapolation of current trends through the remainder of this century would see the area enter Winkler region 3, thereby potentially threatening sustainable viticultural activities for premium wine production over this time frame, in large part because the rate of change would lead to serious financial pressures for nearcontinuous replanting. Growing season average temperatures have been correlated to the ability to ripen the major wine grape varietals (Fig. 5). In the 1970s, Kelowna was – according to this system – unable to fully ripen almost all varietals with the sole exceptions of the lower boundaries for the cool climate members Muller-Thurgau, Pinot Gris, and Gewurtztraminer. Just 40 years later, this site is now (along with Vernon) at the upper boundary for these varieties, and is currently in the optimum range

Table 4 Trends in average growing season (GST,avg) and average dormant season (DST,avg) temperatures during recent decades at other climate stations in south-central British Columbia surrounding the study area. Station name

Station ID

Latitude (°N)

Longitude (°W)

Elevation (masl)

1980–2014 GST,avg trend (°C/century) [±SE]

1970–2014 DST,avg trend (°C/century) [±SE]

Cranbrook Grand Forks Kamloops Merritt Oliver Princeton Salmon Arm

1,152,102 1,133,270 1,163,780 1,125,079 1,125,766 1,126,510 1166R45

49.62 49.03 50.70 50.12 49.18 49.47 50.68

115.78 118.47 120.45 120.80 119.55 120.52 119.23

940 532 345 609 297 700 527

1.0 ± 1.2a 3.8 ± 1.4⁎ 2.6 ± 1.1⁎ 3.5 ± 1.0⁎⁎ 5.8 ± 1.0⁎⁎⁎ 2.9 ± 1.1⁎ 3.5 ± 1.4⁎

3.3 ± 1.6⁎ 4.4 ± 1.9⁎ 4.6 ± 1.9⁎ 4.1 ± 1.7⁎

a Not significant (p N 0.10). ⁎ p b 0.05 significance level. ⁎⁎ p b 0.01 significance level. ⁎⁎⁎ p b 0.001 significance level.

2.6 ± 1.6a 3.3 ± 1.6⁎ 4.9 ± 1.8⁎⁎

176

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

Fig. 3. Comparison of average growing and dormant season temperatures for the Okanagan Valley climate stations with conditions at established wine regions worldwide. Adapted from prior work (Jones et al., 2005b) to include the current dataset. Growing season is April through October in the Northern Hemisphere and October through April in the Southern Hemisphere. Dormant season is November through March in the Northern Hemisphere and May through September in the Southern Hemisphere. Climate maturity groupings are based upon the average growing season temperatures and ability to ripen a variety. Site codings are as follows: MV-DE (Mosel Valley, Germany); A-FR (Alsace, France); C-FR (Champagne, France); RV-DE (Rhine Valley, Germany); NO-US (Northern Oregon, USA); LV-FR (Loire Valley, France); BC-FR (Burgundy-Cote, France); BB-FR (Burgundy-Beaujolais, France); CH (Chile); EW-US (Eastern Washington, USA); BO-FR (Bordeaux, France); CW-US (Central Washington, USA); R-ES (Rioja, Spain); SO-US (Southern Oregon, USA); CC-US (Coastal California, USA); SA (South Africa); NC-US (Northern California, USA); NRV-FR (Northern Rhone Valley, France); NP (Northern Portugal); B-IT (Barolo, Italy); SRV-FR (Southern Rhone Valley, France); MR-AU (Margaret River, Australia); C-IT (Chianti, Italy); HV-AU (Hunter Valley, Australia); BV-AU (Barossa Valley, Australia); SP (Southern Portugal); SC-US (Southern California, USA); P-BC-CA(70s) (Penticton, BC, Canada; 1970s average); P-BC-CA(00s) (Penticton, BC, Canada; 2000s average); K-BC-CA(70s) (Kelowna, BC, Canada; 1970s average); K-BC-CA(00s) (Kelowna, BC, Canada; 2000s average); V-BC-CA(70s) (Vernon, BC, Canada; 1970s average); and V-BC-CA(00s) (Vernon, BC, Canada; 2000s average).

for Pinot Noir, Riesling, Chardonnay, and Sauvignon Blanc. Penticton has transitioned from a position during the 1970s of where Kelowna is currently to a GST,avg that is potentially able to fully ripen Semillon and Cabernet Franc, and is approaching the lower boundaries for such warm-climate wine grape varieties as Merlot, Malbec, and Syrah, among others. The current cooler-climate varietals planted in the study area (see Table 1) will be unsuitable for wine production within several decades at the rate for which growing season temperatures are increasing. As evidenced by the increasing presence of warmerclimate varietals (e.g., Merlot, Cabernet Sauvignon, Cabernet Franc) being planted in the central and northern portions of the valley, growers are already responding to the climate changes that have taken place over the past few decades. If current trends continue, over the coming decades significant portions of the Okanagan Valley study area will likely move from Winkler heat unit region 1 to 2, changing its classification from a region similar to the Niagara Peninsula (Canada) and Geneva (New York, USA) closer to the current climates in Sienna (Italy), Napa Valley (California, USA), and the Margaret River region of Australia. Increasing temperatures throughout the region are also expected to result in increased risk of pests such as Pierces disease. The rapidly increasing growing and dormant season temperatures will have major economic impacts on producers who will be forced to address over-ripening concerns for current varietals and the costs of repeated replanting to keep pace with the fast-changing climate. Over the short through medium-term, the increased heat units during the growing season will likely improve wine grape quality, but over the longer term the upper boundaries for premium wine grape production may be encountered, posing a risk to the industry itself. In addition, large regions of the southern and central valley walls and the northern valley floor and sidewalls will become

Fig. 4. Comparison of growing season conditions for the Okanagan Valley climate stations with conditions at established wine regions elsewhere in Canada and worldwide. Adapted from prior work (Shaw, 2001) to include the current dataset. GDD10 use the April through October growing season in the Northern Hemisphere and the October through April growing season in the Southern Hemisphere. Latitude-temperature index (LTI) classes are based on those of Jackson and Cherry (1988)). Winkler heat unit groupings based on those of Winkler et al. (1974). Site codings are as follows: K-NS-CA (Kentville, NS, Canada); P-NS-CA (Pugwash, NS, Canada); I-PQ-CA (Iberville, PQ, Canada); V-ON-CA (Vineland, ON, Canada); H-ON-CA (Harrow, ON, Canada); K-ON-CA (Kingsville, ON, Canada); P-ON-CA (Pelee Island, ON, Canada); L-ON-CA (Leamington, ON, Canada); RON-CA (Ridgetown, ON, Canada); W-ON-CA (Windsor, ON, Canada); C-BC-CA (Cowichan Bay, BC, Canada); M-FR (Montpellier, France); BM-FR (Bordeaux/Medoc, France); S-IT (Sienna, Italy); L-ES (Logrono, Spain); F-DE (Freiburg, Germany); G-DE (Geisenheim, Germany); N-CA-US (Napa Valley, CA, USA); G-NY-US (Geneva, NY, USA); Y-WA-US (Yakima Valley, WA, USA); M-AU (Margaret River, Australia); H-NZ (Hawkes Bay, New Zealand); B-NZ (Blenheim, New Zealand); P-BC-CA(70s) (Penticton, BC, Canada; 1970s average); P-BC-CA(00s) (Penticton, BC, Canada; 2000s average); K-BC-CA(70s) (Kelowna, BC, Canada; 1970s average); K-BC-CA(00s) (Kelowna, BC, Canada; 2000s average); V-BCCA(70s) (Vernon, BC, Canada; 1970s average); and V-BC-CA(00s) (Vernon, BC, Canada; 2000s average).

climatically suitable for viticulture. This will place substantial agricultural pressure on the fragile landscape – which is concurrently under threat from rapid population growth – that contains the highest number of endangered, threatened, and rare species in British Columbia. 5. Conclusions Average growing and dormant season temperatures are increasing rapidly for the wine grape growing regions surrounding Okanagan Lake in south-central British Columbia, Canada since the post-1980 period at rates upwards of 7.0 ± 1.3 °C/century, representing some of the most rapid increases seen among the world's wine producing areas over the past few decades. Due to dramatic changes in the average growing season temperature, growing degree days and the latitudetemperature index are also experiencing substantial increases, thereby altering the climate classifications for viticulture in the study area. Climate changes during the past several decades have led to the southern portion of the region moving out of the cool-climate viticultural classification and into the intermediate zone, while the central and northern regions are now at the cool/intermediate viticulture interface. Grape growers in the area have responded to these changes by increased plantings of warmer-climate wine grape varietals during the past two decades. If current rates of warming continue, the quality of wine production may increase in the short- through medium term because of the increased growing season heat units available, but over the longterm, the sustainability of the industry may be threatened from increased pest pressures, additional extreme heat days, and excessive growing season temperatures that move the region's climate classification beyond current limits for global premium wine grape production.

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

177

Fig. 5. Wine grape climate maturity groupings based on average growing season temperatures and the estimated span of varietal ripening potential that occurs within and across groups for the Okanagan Valley climate stations. Adapted from prior work (Jones et al., 2005b) to include the current dataset.

Acknowledgments We thank the anonymous reviewers for their comments which greatly improved the manuscript.

References Amerine, M.A., Winkler, A.J., 1944. Composition and quality of musts and wines of California grapes. Hilgardia 15, 493–675. Barnett, T.P., Pierce, D.W., Latif, M., Dommenget, D., Saravanan, R., 1999. Interdecadal interactions between the tropics and midlatitudes in the Pacific basic. Geophys. Res. Lett. 26, 615–618. BCWI, 2014. 2014 B.C. Wine Grape Acreage Report. British Columbia Wine Institute, Kelowna, BC, Canada. Belliveau, S., Smit, B., Bradshaw, B., 2006. Multiple exposures and dynamic vulnerability: evidence from the grape industry in the Okanagan Valley, Canada. Glob. Environ. Chang. 16, 364–378. http://dx.doi.org/10.1016/j.gloenvcha.2006.03.003. Bowen, P.A., Bogdanoff, C.P., Estergaard, B.F., Marsh, S.G., Usher, K.B., Smith, C.A.S., Frank, G., 2005. Geology and wine 10: use of geographic information system technology to assess viticulture performance in the Okanagan and Similkameen valleys, British Columbia. Geosci. Can. 32, 161–176. Caprio, J.M., Quamme, H.A., 2002. Weather conditions associated with grape production in the Okanagan Valley of British Columbia and potential impact of climate change. Can. J. Plant Sci. 82, 755–763. http://dx.doi.org/10.4141/P01-160. Duchêne, E., Schneider, C., 2005. Grapevine and climatic changes: a glance at the situation in Alsace. Agron. Sustain. Dev. 25, 93–99. Duchne, E., Huard, F., Dumas, V., Schneider, C., Merdinoglu, D., 2010. The challenge of adapting grapevine varieties to climate change. Clim. Res. 41, 193–204. http://dx. doi.org/10.3354/cr00850. Durbin, J., Watson, G.S., 1950. Testing for serial correlation in least squares regression. I. Biometrika 37, 409–428. http://dx.doi.org/10.1093/biomet/37.3-4.409. Durbin, J., Watson, G.S., 1951. Testing for serial correlation in least squares regression. II. Biometrika 38, 159–177. http://dx.doi.org/10.2307/2332325. Etien, N., Daux, V., Masson-Delmotte, V., Mestre, O., Stievenard, M., Guillemin, M.T., Boettger, T., Breda, N., Haupt, M., Perraud, P.P., 2008. Summer maximum temperature in Northern France over the past century: instrumental data versus multiple proxies (tree-ring isotopes, grape harvest dates and forest fires). Clim. Chang. 94, 429–456. http://dx.doi.org/10.1007/s10584-008-9516-8. Fraga, H., Malheiro, A.C., Moutinho-Pereira, J., Santos, J.A., 2012. An overview of climate change impacts on European viticulture. Food Energy Secur. 1, 94–110. http://dx. doi.org/10.1002/fes3.14. Gladstones, J., 1992. Viticulture and environment. Winetitles, Broadview, SA, Australia. Hannah, L., Roehrdanz, P.R., Ikegami, M., Shepard, A.V., Shaw, M.R., Tabor, G., Zhi, L., Marquet, P.A., Hijmans, R.J., 2013. Climate change, wine, and conservation. Proc. Natl. Acad. Sci. 110, 6907–6912. http://dx.doi.org/10.1073/pnas.1210127110. Hayhoe, K., Cayan, D., Field, C.B., Frumhoff, P.C., Maurer, E.P., Miller, N.L., Moser, S.C., Schneider, S.H., Cahill, K.N., Cleland, E.E., Dale, L., Drapek, R., Hanemann, R.M., Kalkstein, L.S., Lenihan, J., Lunch, C.K., Neilson, R.P., Sheridan, S.C., Verville, J.H., 2004. Emissions pathways, climate change, and impacts on California. Proc. Natl. Acad. Sci. U. S. A. 101, 12422–12427. http://dx.doi.org/10.1073/pnas.0404500101. Holland, T., Smit, B., 2010. Climate change and the wine industry: current research themes and new directions. J. Wine Res. 21, 125–136. http://dx.doi.org/10.1080/ 09571264.2010.530095. Jackson, D.I., Cherry, N.J., 1988. Prediction of a district's grape-ripening capacity using a latitude–temperature index (LTI). Am. J. Enol. Vitic. 39, 19–28. Jones, G.V., 2004. Making wine in a changing climate. Geotimes (August, 24–27). Jones, G.V., 2005a. How hot is too hot? Wine business monthly February, 1–4.

Jones, G.V., 2005b. Climate change in the western United States grape growing regions. Acta Horticult. 41–60 http://dx.doi.org/10.17660/ActaHortic.2005.689.2. Jones, G.V., 2007. Climate change: Observations, projections, and general implications for viticulture and wine production. Whitman College, Walla Walla, WA, USA. Jones, G.V., Davis, R.E., 2000a. Climate influences on grapevine phenology, grape composition, and wine production and quality for Bordeaux, France. Am. J. Enol. Vitic. 51, 249–261. Jones, G.V., Davis, R.E., 2000b. Using a synoptic climatological approach to understand climate–viticulture relationships. Int. J. Climatol. 20, 813–837. http://dx.doi.org/10. 1002/1097-0088(20000630)20:8b813::AID-JOC495N3.0.CO;2-W. Jones, G.V., Duchene, E., Tomasi, D., Yuste, J., Braslavska, O., Schultz, H., Martinez, C., Boso, S., Langellier, F., Perruchot, C., Guimberteau, G., 2005a. Changes in European winegrape phenology and relationships with climate. Presented at the XIV International GESCO Viticulture Congress. Groupe d'Etude des Systemes de Conduite de la vigne (GESCO), Geisenheim, Germany, pp. 54–61. Jones, G.V., White, M.A., Cooper, O.R., Storchmann, K., 2005b. Climate change and global wine quality. Clim. Chang. 73, 319–343. http://dx.doi.org/10.1007/s10584-0054704-2. Kenny, G.J., Shao, J., 1992. An assessment of a latitude–temperature index for predicting climate suitability for grapes in Europe. J. Hortic. Sci. 67, 239–246. http://dx.doi.org/ 10.1080/00221589.1992.11516243. Latif, M., Barnett, T.P., 1994. Causes of decadal climate variability over the North Pacific and North America. Science 266, 634–637. http://dx.doi.org/10.1126/science.266. 5185.634. van Leeuwen, C., Friant, P., Choné, X., Tregoat, O., Koundouras, S., Dubourdieu, D., 2004. Influence of climate, soil, and cultivar on terroir. Am. J. Enol. Vitic. 55, 207–217. Maltman, A., 2008. The role of vineyard geology in wine typicity. J. Wine Res. 19, 1–17. http://dx.doi.org/10.1080/09571260802163998. Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C., 1997. A Pacific interdecadal climate oscillation with impacts on salmon production. Bull. Am. Meteorol. Soc. 78, 1069–1079. http://dx.doi.org/10.1175/1520-0477(1997)078b1069:APICOWN2.0.CO; 2. Minobe, S., 1997. A 50–70 year climatic oscillation over the North Pacific and North America. Geophys. Res. Lett. 24, 683–686. http://dx.doi.org/10.1029/97GL00504. Mira de Orduña, R., 2010. Climate change associated effects on grape and wine quality and production. Food Res. Int. Clim. Chang. Food Sci. 43, 1844–1855. http://dx.doi. org/10.1016/j.foodres.2010.05.001. Mullins, M.G., Bouquet, A., Williams, L.E., 1992. Biology of the grapevine. Cambridge University Press, Cambridge, United Kingdom. Nemani, R.R., White, M.A., Cayan, D.R., Jones, G.V., Running, S.W., Coughlan, J.C., Peterson, D.L., 2001. Asymmetric warming over coastal California and its impact on the premium wine industry. Clim. Res. 19, 25–34. http://dx.doi.org/10.3354/cr019025. Nigam, S., Barlow, M., Berbery, E.H., 1999. Analysis links Pacific decadal variability to drought and streamflow in United States. EOS Trans. Am. Geophys. Union 80, 621–625. http://dx.doi.org/10.1029/99EO00412. Oliveira, M., 1998. Calculation of budbreak and flowering base temperatures for Vitis vinifera cv. Touriga Francesa in the Douro region of Portugal. Am. J. Enol. Vitic. 49, 74–78. O'Reilly, C.M., Sharma, S., Gray, D.K., Hampton, S.E., Read, J.S., Rowley, R.J., Schneider, P., Lenters, J.D., McIntyre, P.B., Kraemer, B.M., Weyhenmeyer, G.A., Straile, D., Dong, B., Adrian, R., Allan, M.G., Anneville, O., Arvola, L., Austin, J., Bailey, J.L., Baron, J.S., Brookes, J.D., de Eyto, E., Dokulil, M.T., Hamilton, D.P., Havens, K., Hetherington, A.L., Higgins, S.N., Hook, S., Izmest'eva, L.R., Joehnk, K.D., Kangur, K., Kasprzak, P., Kumagai, M., Kuusisto, E., Leshkevich, G., Livingstone, D.M., MacIntyre, S., May, L., Melack, J.M., Mueller-Navarra, D.C., Naumenko, M., Noges, P., Noges, T., North, R.P., Plisnier, P.-D., Rigosi, A., Rimmer, A., Rogora, M., Rudstam, L.G., Rusak, J.A., Salmaso, N., Samal, N.R., Schindler, D.E., Schladow, S.G., Schmid, M., Schmidt, S.R., Silow, E., Soylu, M.E., Teubner, K., Verburg, P., Voutilainen, A., Watkinson, A., Williamson, C.E., Zhang, G., 2015. Rapid and highly variable warming of lake surface waters around the globe. Geophys. Res. Lett. 42. http://dx.doi.org/10.1002/2015GL066235 (2015GL066235).

178

S. Rayne, K. Forest / Science of the Total Environment 556 (2016) 169–178

Petrie, P.r., Sadras, V.o., 2008. Advancement of grapevine maturity in Australia between 1993 and 2006: putative causes, magnitude of trends and viticultural consequences. Aust. J. Grape Wine Res. 14, 33–45. http://dx.doi.org/10.1111/j.1755-0238.2008. 00005.x. Sadras, V.O., Petrie, P.R., 2011. Climate shifts in South-Eastern Australia: early maturity of chardonnay, Shiraz and Cabernet Sauvignon is associated with early onset rather than faster ripening. Aust. J. Grape Wine Res. 17, 199–205. http://dx.doi.org/10. 1111/j.1755-0238.2011.00138.x. Schultz, H.R., Jones, G.V., 2010. Climate induced historic and future changes in viticulture. J. Wine Res. 21, 137–145. http://dx.doi.org/10.1080/09571264.2010.530098. Shaw, A.B., 2001. Pelee Island and Lake Erie north shore, Ontario: a climatic analysis of Canada's warmest wine region. J. Wine Res. 12, 19–37. http://dx.doi.org/10.1080/ 09571260120069732. Shaw, A.B., 2005. The Niagara peninsula viticultural area: a climatic analysis of Canada's largest wine region. J. Wine Res. 16, 85–103. http://dx.doi.org/10.1080/ 09571260500327630. Stock, M., Gerstengarbe, F.W., Kartschall, T., Werner, P.C., 2005. Reliability of climate change impact assessments for viticulture. Acta Horticult. 29–40. http://dx.doi.org/ 10.17660/ActaHortic.2005.689.1. Tomasi, D., Jones, G.V., Giust, M., Lovat, L., Gaiotti, F., 2011. Grapevine phenology and climate change: relationships and trends in the Veneto Region of Italy for 1964–2009. Am. J. Enol. Vitic. http://dx.doi.org/10.5344/ajev.2011.10108 (ajev.2011.10108). Trenberth, K.E., 1990. Recent observed interdecadal climate changes in the Northern Hemisphere. Bull. Am. Meteorol. Soc. 71, 988–993. http://dx.doi.org/10.1175/15200477(1990)071b0988:ROICCIN2.0.CO;2. Trenberth, K.E., Hurrell, J.W., 1994. Decadal atmosphere–ocean variations in the Pacific. Clim. Dyn. 9, 303–319. http://dx.doi.org/10.1007/BF00204745. Vincent, L.A., 1998. A technique for the identification of inhomogeneities in Canadian temperature series. J. Clim. 11, 1094–1104. http://dx.doi.org/10.1175/ 1520-0442(1998)011b1094:ATFTION2.0.CO;2. Vincent, L.A., Gullett, D.W., 1999. Canadian historical and homogeneous temperature datasets for climate change analyses. Int. J. Climatol. 19, 1375–1388. http://dx.doi. org/10.1002/(SICI)1097-0088(199910)19:12b1375::AID-JOC427N3.0.CO;2-0. Vincent, L.A., Milewska, E.J., Hopkinson, R., Malone, L., 2009. Bias in minimum temperature introduced by a redefinition of the climatological day at the Canadian synoptic stations. J. Appl. Meteorol. Climatol. 48, 2160–2168. http://dx.doi.org/10.1175/ 2009JAMC2191.1. Vincent, L.A., Wang, X.L., Milewska, E.J., Wan, H., Yang, F., Swail, V., 2012. A second generation of homogenized Canadian monthly surface air temperature for climate trend analysis. J. Geophys. Res. Atmos. 117, D18110. http://dx.doi.org/10.1029/ 2012JD017859.

Vincent, L.A., Zhang, X., Bonsal, B.R., Hogg, W.D., 2002. Homogenization of daily temperatures over Canada. J. Clim. 15, 1322–1334. http://dx.doi.org/10.1175/15200442(2002)015b1322:HODTOCN2.0.CO;2. Vink, N., Deloire, A., Bonnardot, V., Ewert, J., 2012. Climate change and the future of South Africa's wine industry. Int. J. Clim. Chang. Strateg. Manag. 4, 420–441. http://dx.doi. org/10.1108/17568691211277746. Wallace, J.M., Gutzler, D.S., 1981. Teleconnections in the geopotential height field during the Northern Hemisphere winter. Mon. Weather Rev. 109, 784–812. http://dx.doi. org/10.1175/1520-0493(1981)109b0784:TITGHFN2.0.CO;2. Wang, X.L., Chen, H., Wu, Y., Feng, Y., Pu, Q., 2010. New techniques for the detection and adjustment of shifts in daily precipitation data series. J. Appl. Meteorol. Climatol. 49, 2416–2436. http://dx.doi.org/10.1175/2010JAMC2376.1. Wang, X.L., Wen, Q.H., Wu, Y., 2007. Penalized maximal t-test for detecting undocumented mean change in climate data series. J. Appl. Meteorol. Climatol. 46, 916–931. http://dx.doi.org/10.1175/JAM2504.1. Watson, G.S., Durbin, J., 1951. Exact tests of serial correlation using noncircular statistics. Ann. Math. Stat. 22, 446–451. Webb, L.B., Whetton, P.H., Barlow, E.W.R., 2011. Observed trends in winegrape maturity in Australia. Glob. Chang. Biol. 17, 2707–2719. http://dx.doi.org/10.1111/j.13652486.2011.02434.x. Webb, L.B., Whetton, P.H., Bhend, J., Darbyshire, R., Briggs, P.R., Barlow, E.W.R., 2012. Earlier wine-grape ripening driven by climatic warming and drying and management practices. Nat. Clim. Chang. 2, 259–264. http://dx.doi.org/10.1038/nclimate1417. White, M.A., Diffenbaugh, N.S., Jones, G.V., Pal, J.S., Giorgi, F., 2006. Extreme heat reduces and shifts United States premium wine production in the 21st century. Proc. Natl. Acad. Sci. 103, 11217–11222. http://dx.doi.org/10.1073/pnas.0603230103. White, M.A., Whalen, P., Jones, G.V., 2009. Land and wine. Nat. Geosci. 2, 82–84. http://dx. doi.org/10.1038/ngeo429. Winkler, A.J., Cook, J.A., Kliewer, W.M., Lider, L.A., Cerruti, L., 1974. General Viticulture. University of California Press, Berkeley, CA, United States. Wolfe, D.W., Schwartz, M.D., Lakso, A.N., Otsuki, Y., Pool, R.M., Shaulis, N.J., 2004. Climate change and shifts in spring phenology of three horticultural woody perennials in Northeastern USA. Int. J. Biometeorol. 49, 303–309. http://dx.doi.org/10.1007/ s00484-004-0248-9. Yoshioka, K., 2002. KyPlot as a tool for graphical data analysis. In: Härdle, W., Rönz, B. (Eds.), Compstat: Proceedings in computational statistics. Springer, New York, pp. 37–46. Zhang, Y., Wallace, J.M., Battisti, D.S., 1997. ENSO-like interdecadal variability: 1900–93. J. Clim. 10, 1004–1020. http://dx.doi.org/10.1175/1520-0442(1997)010b1004: ELIVN2.0.CO;2.