Réchauffement climatique, quels impacts probables sur les vignobles ? 1 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
Climate change: observed trends, simulations, impacts and response strategy for the South African vineyards
Valérie BONNARDOT 1∗ and Victoria Anne CAREY 2
1 ARC-Institute for Soil, Climate and Water, Private Bag X79, Pretoria 0001, South Africa 2 University of Stellenbosch, Department of Viticulture and Oenology, Private Bag X1, Matieland 7602, South Africa [email protected] - [email protected]
Abstract Global warming is scientifically and widely accepted (IPCC). Climate change is a reality and its impacts are increasingly felt in South Africa. According to a recent economic impact assessment on climate change in agriculture and particularly the study on farmers’ perceptions of changes in climate, long-term changes in temperature and rainfall have been noticed in all nine provinces of South Africa (Benhin, 2005). This paper focuses on climate change in the Western Cape Province, where most of the traditional South African vineyards lie, examining the observed climatic trends and potential impacts for viticulture. Using the longest data series from weather stations located in the Stellenbosch wine district, major trends in rainfall and temperature over the past 40 years can be drawn for the vineyards of this region: A significant increase in the annual temperature values was noticed. February, ripening period for most of the cultivars in the Western Cape Province, was warmer in terms of minimum as well as maximum temperatures. This may hasten sugar accumulation in wine grapes. With a significant temperature increase from April to July, the winter season has shortened. As a result, lack of sufficient cold during late autumn and beginning of winter may interfere with bud dormancy, which in turn may lead to irregular budburst, resulting uneven growth, berry set, fruit ripening and crop loss. The use of chemical agents to get an even budburst may have to increase. A significant increase in August and September rainfall was noticed. This means that the rainy season tends to occur later than expected. As a result of more spring rainfall, disease pressure during the early season may increase. Assuming an even 0.5°C temperature increase over the year (accepted value for Southern Africa), some global climatic indices for viticulture were calculated for different South African wine regions. Some regions reach the upper level of their group while others change to warmer groups. More heat tolerant varieties would have to be planted. Changes in the areas of suitability for certain cultivars could be of major importance for the regional economy. Adaptation, change in production practices and development of new wine regions are the keys to surviving climate change. The search for cooler locations towards the coast, where the moderating effect of the ocean is expected to temper increasing summer temperatures or at altitude, as well as the development of vineyards towards the eastern part of the Western Cape Province with an evenly distributed rainfall are an existing response to climate change in South Africa.
Key-Words: Climatic trends, vineyards, South Africa
* * *
∗ Current address: BERCAV Bureau d’Etude et de Recherche en Climatologie Appliquée à la Viticulture, 3 479 Route de Thonon, 74380 Cranves-Sales, France. Réchauffement climatique, quels impacts probables sur les vignobles ? 2 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
South Africa is a young wine producing country with an increasing area under vines (110.000 ha in 2005), a production of 950 million litres (3% of the world production) 9th largest overall volume production. Locally, viticulture represents 8.2% of the gross product of the Western Cape Province and there are 257.000 persons working directly or indirectly for the wine industry (SAWIS, 2006). It therefore forms one of the vital sectors of the provincial economy.
Climate is an important environmental factor in viticulture, which, together with soil surveys, is studied in order to help in establishing new locations for vineyards, assessing the potential viability and profitability of the vineyard and managing sites. Global warming is scientifically and widely accepted (IPCC, 2001). Climate change is a reality and its impacts are increasingly felt in South Africa. According to a recent economic impact assessment on climate change in agriculture and particularly the study on farmers’ perceptions of changes in climate, long-term changes in temperature and rainfall have been noticed in all nine provinces of South Africa (Nhemachena et al., 2006). This paper focuses on climate change in the Western Cape Province, where most of the traditional South African vineyards are situated, examining the observed climatic trends and potential impacts for viticulture.
I. Data and method
Since 1940, the Institute for Soil, Climate and Water of the Agricultural Research Council has installed a countrywide network of weather stations (mechanical and automatic) aimed at monitoring the climate and satisfying the climatic requirements of agriculture. Mechanical weather stations were chosen due to their longer time series of daily data (40 in total with at least 13 years of records) and a larger spatial representation of the wine regions compared to the automatic weather station network, which has been established more recently in a reduced number of regions.
Using the longest temperature and rainfall series from these mechanical weather stations located in the different wine regions of South Africa, a descriptive statistical analysis (3-year moving averages, decade averages and linear trends) and the calculation of the Winkler index for viticulture, preliminary results can be drawn for the vineyards of this country over the past 40 years.
II. Results
II.1. Observed climatic trends
A significant increase in the annual temperature values was noticed for the past three decades. Figure 1 shows the example of the annual maximum and minimum temperature for a station in the Stellenbosch wine district from 1967 to 2005 as well as the 3-year moving averages and the linear trends. Except the usual inter-variability between years, both minimum and maximum temperature trends show a positive slope (warming) over the 39-year period. The top ten warmest years in the district and over this period occurred after 1985 except for 1973 (eighth warmest year). The five warmest years were 1999, 2000, 1985, 1993 and 2005 with annual mean temperature between 0.6 and 1°C above the 1967-2005 average (Table 1). Looking at the average per decades for the same weather station (Fig. 2), a 1°C and 0.7°C increase for maximum and minimum temperature respectively were noticeable over the past three decades. Réchauffement climatique, quels impacts probables sur les vignobles ? 3 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
Table 1: Annual temperature (maximum, minimum and mean) and anomaly for the top 12 warmest years over the 1967- 2005 period at Nietvoorbij in the Stellenbosch wine district. Top five warmest years in bold red.
Anomaly (°C) Warmest Years MaxT (°C) MinT (°C) MeanT (°C) year MaxT MinT MeanT (rank) 1973 23,4 11,9 17,7 0,6 0,2 0,5 7th 1984 23,2 12,0 17,6 0,4 0,3 0,4 10th 1985 23,6 12,2 17,9 0,8 0,5 0,7 3rd 1986 23,3 12,2 17,7 0,5 0,5 0,5 8th 1988 23,3 11,8 17,6 0,5 0,1 0,4 9th 1993 23,4 12,2 17,8 0,6 0,5 0,6 5th 1994 23,4 12,0 17,7 0,6 0,3 0,5 6th 1999 24,1 12,4 18,2 1,3 0,7 1,0 1st 2000 23,6 12,1 17,9 0,8 0,4 0,7 2nd 2001 23,1 12,1 17,6 0,3 0,4 0,4 11 th 2004 23,1 11,9 17,5 0,3 0,2 0,3 12th 2005 23,4 12,3 17,8 0,6 0,6 0,6 4th
24 23 22 R2 = 0,32 21 p < 0,01 20 19 18 17 16 15 Temperature (°C) 2 14 R = 0,29 p < 0,01 13 12 11 10 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Years
MaxT MinT Linear MaxT Linear MinT 3-year moving average
Figure 1. Minimum (blue) and maximum (red) annual temperature at Nietvoorbij weather station in the Stellenbosch Wine District (1967-2005 period).
Réchauffement climatique, quels impacts probables sur les vignobles ? 4 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
25 24 23,0 23,1 23,0 22,5 23 22,1 22 R2 = 0,82 21 p < 0,01 20 19 18 17 16
Temperature (°C) 15 R2 = 0,78 14 p < 0,01 13 11,6 11,7 11,8 11,8 12 11,1 11 10 1967-70 1971-80 1981-90 1991-00 2001-05 Decades
MaxT MinT Linear (MaxT) Linear (MinT)
Figure 2. Minimum (blue) and maximum (red) annual temperature per decade at Nietvoorbij weather station in the Stellenbosch Wine District (1967-2005 period).
Considering an earlier annual temperature time series (1942-1965) at another weather station (Lourensford) in the same district (Stellenbosch), the slopes were slightly positive but not statistically significant (Fig. 3). A slight temperature increase (0.4°C and 0.5°C for maximum and minimum temperature respectively could be noticed but it does not indicate a significant warming. The warming as noticed previously over the 1967-2005 period had not appeared yet over the 1942-1965 period. What was significant over the 1942-1965 period, was the increase in winter maximum temperature. The increase of 1.5°C identified for August maximum temperature over the 24-year period from 1942 to 1965 could be the first marker of a warming in this region (Fig. 4). For Nietvoorbij during the most recent decades (1967-2005), the winter temperature increase is clearly noticeable from the beginning of the 1970’s. Using July data, mean increases of more than 2°C for maximum and up to 1°C for minimum temperatures over the past four decades were noticed (Fig. 5).
Réchauffement climatique, quels impacts probables sur les vignobles ? 5 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
24 23 22 21 R² = 0,02 20 19 18 17 16 15 14 13 Temperature (°C) 12 11 10 9 R² = 0,02 8 7 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 Years
MaxT MinT Linear MaxT Linear MinT 3-year moving average
Figure 3. Minimum (blue) and maximum (red) annual temperature at Lourensford weather station in the Stellenbosch Wine District (1942-1965 period).
22 21 R² = 0,2 20 p < 0,3 19 18 17 16 15 14 13 12 11
Temperature (°C) 10 9 8 7 6 5 R² = 0,02 4 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 Years
MaxT MinT Linear MaxT Linear MinT 3-year moving average
Figure 4. August minimum (blue) and maximum (red) temperature at Lourensford weather station in the Stellenbosch Wine District (1942-1965 period).
Réchauffement climatique, quels impacts probables sur les vignobles ? 6 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
19 18,4 18 17,2 17,2 17,4 R² = 0,86 17 16,1 p < 0,01 16 15 14 13 12 11 Temperature (°C) Temperature 10
9 8,0 8,0 7,6 7,8 8 7,0 7 R² = 0,78 p < 0,05 6 1967-70 1971-80 1981-90 1991-00 2001-05 Decades
MaxT MinT Linear (MaxT) Linear (MinT)
Figure 5. Minimum (blue) and maximum (red) July temperature per decade at Nietvoorbij weather station in the Stellenbosch Wine District (1967-2005 period).
The most significant trend was found during the growing period, especially during February, the ripening period of most of the cultivars in the Western Cape. February month was warmer in terms of minimum as well as maximum temperatures, which may have hastened sugar accumulation in wine grapes. As for example, for Nietvoorbij in the Stellenbosch district, there has been a greater than 2°C increase for maximum temperature and up to 1.3°C increase for the minimum temperature since the beginning of the 1980’s (Fig. 6).
31 29,5 30 28,4 28,7 29 27,4 27,5 R2 = 0,95 28 p < 0,01 27 26 25 24 23 22 21 20 Temperature (°C) R2 = 0,79 19 p < 0,05 18 16,1 17 15,9 15,8 16 14,8 14,7 15 14 1967-70 1971-80 1981-90 1991-00 2001-05 Decades
MaxT MinT Linear (MaxT) Linear (MinT)
Figure 6. Minimum (blue) and maximum (red) February temperature per decade at Nietvoorbij weather station in the Stellenbosch Wine District (1967-2005 period). Réchauffement climatique, quels impacts probables sur les vignobles ? 7 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
The Winkler index was calculated for the different wine-producing regions (Fig. 7). There was no trend for the Winkler index series of Lourensford over the 1942-1965 period (Fig. 7a) as there was no significant increase in summer temperature during this period as shown previously (Fig. 3). It had a mean Winkler index that falls in Region II and a regular inter-annual variability through the years. Nietvoorbij in the Stellenbosch district had a mean index that fell into the lower level of Region IV for the 1967-2005 period (Fig. 7b). The years with lower values (Region III) occurred at the beginning of the time series, became less frequent after 1985. From the mid 1980’s onwards the index was regularly higher. The mean Winkler index for Constantia for the 1967-2000 period was calculated as being in the lower level of Region III. Up to the 1980’s, the index varied between the upper level of Region II and the lower level of Region III. In the 1990’s, the index reached the upper level of Region III (Fig. 7c). In the Walker Bay district (Hermanus), the mean Winkler index was calculated as being at the lower threshold of Region III for the 1977-1991 period, while the index fell mainly in Region II before 1984-85 (Fig. 7d). In the other districts, namely Paarl (Fig. 7e) and Worcester (Fig. 7f), as well as at Klawer in the Olifant River region (Fig. 7g), the same patterns (a higher index after the mid 1980’s) were observed; while there was no trend observed over the 1982-2001 period at Swellendam (Fig. 7h) situated past the eastern margin of the area of winter rainfall, downstream of the Breede River towards the South Coast of the Western Cape. The breakpoint in the mid 1980’s corroborates the trends observed in regional climate of Southern Africa (WWF, 1996).
7a) Lourensford
1900
1800
1700
1600
R² = 0,02 1500 Winkler index 1400
1300
1200 42/43 43/44 44/45 45/46 46/47 47/48 48/49 49/50 50/51 51/52 52/53 53/54 54/55 55/56 56/57 57/58 58/59 60/61 61/62 62/63 63/64 64/65 Seasons
Mean index Upper level Region I Upper level Region II Upper level Region III Winkler index Linear (Winkler index)
Réchauffement climatique, quels impacts probables sur les vignobles ? 8 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
7b) Nietvoorbij
2300
2200 R² = 0,12 p < 0,4 2100
2000
1900 Winkler index 1800
1700
1600 67/68 69/70 71/72 73/74 75/76 77/78 79/80 81/82 83/84 85/86 87/88 89/90 91/92 93/94 95/96 97/98 99/00 2001/02 2003/04 2005/06 Seasons
Mean index Upper level Region III Upper level Region II Upper level Region IV Winkler index Linear (Winkler index)
7c) Constantia
2100 R² = 0,35 p < 0,05 2000
1900
1800
1700 Winkler index 1600
1500
1400 67/68 69/70 71/72 73/74 75/76 77/78 79/80 81/82 83/84 85/86 87/88 89/90 91/92 93/94 95/96 97/98 Seasons
Mean index Upper level Region II Upper level Region III Winkler index Linear (Winkler index)
Réchauffement climatique, quels impacts probables sur les vignobles ? 9 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
7d) Hermanus
2000
1900 R2 = 0,34 p < 0,25 1800
1700
1600 Winkler index 1500
1400
1300 77/78 78/79 79/80 80/81 81/82 82/83 83/84 84/85 85/86 86/87 87/88 88/89 89/90 Seasons
Upper limit Region III Upper limit Region II Upper limit Region I Mean index Winkler index Linear (Winkler index)
7e) Nederburg
2700
2600
2500
R² = 0,25 2400 p < 0,3
2300 Winkler index
2200
2100
2000 70/71 72/73 74/75 76/77 78/79 80/81 82/83 84/85 86/87 88/89 90/91 92/93 94/95 96/97 98/99 00/01 Seasons
Mean index Upper level Region IV Winkler index Linear (Winkler index)
Réchauffement climatique, quels impacts probables sur les vignobles ? 10 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
7f) Veldreserwe
2600 R2 = 0,15 p < 0,4 2500
2400
2300
2200 Winkler index 2100
2000
1900 67/68 69/70 71/72 73/74 75/76 77/78 79/80 81/82 83/84 85/86 87/88 89/90 91/92 93/94 95/96 97/98 99/00 2001/02 Seasons
Mean index Upper level Region III Upper level Region IV Winkler index Linear (Winkler index)
7g) Klawer
3000
2900 R2 = 0,24 p < 0,4 2800
2700
2600 Winkler index 2500
2400
2300 73/74 75/76 77/78 79/80 81/82 83/84 85/86 87/88 89/90 91/92 93/94 95/96 97/98 99/00 2001/02 Seasons
Mean index Winkler index Linear (Winkler index)
Réchauffement climatique, quels impacts probables sur les vignobles ? 11 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
7h) Swellendam
2400
2300
2200
2100 R2 = 0,02 2000 Winkler index 1900
1800
1700 82/83 83/84 84/85 85/86 86/87 87/88 88/89 89/90 90/91 91/92 92/93 93/94 94/95 95/96 96/97 97/98 98/99 99/00 00/01 Seasons
Mean index Upper level Region III Upper level Region IV Winkler index Linear (Winkler index)
Figure 7. Winkler index for a) Lourensford (1942-1965) and b) Nietvoorbij (1967-2005) in the Stellenbosch district; c) Hermanus (1977-1990) in the Overberg district; d) Constantia (1967-1999) in the Constantia Ward; e) Nederburg (1970-2002) in the Paarl district; f) Veldreserwe (1962-2002) in Worcester district; g) Klawer (1973-2002) in the Olifant river region and h) Swellendam (1982-2002) in the Swellendam district. Markers in dark blue represent seasons falling into Region I; light blue into Region II; yellow into Region III; red into Region IV and purple into Region V of Winkler.
The Winkler index was then calculated assuming an even 0.5°C temperature increase, accepted value for global climate as well as for Southern Africa since the nineteenth century (WWF, 1996), over the growing season (Fig. 8). Knowing that a 0.5°C temperature increase is an optimistic scenario of future climate changes, the associated shift to the upper level of current climatic regions or the shift to warmer climatic regions could represent causes for alarm for some of the warmer wine producing regions and definitely suggests a need for change or adaptation of cultivars in other regions. Viticulture could, however develop in non-traditional regions.
With respect to rainfall (results not shown), the results were not always highly significant and were disparate and interpretation of these results should be made with great care. A 50 to 100 mm increase in the annual total over the past 3 decades could be observed at Nietvoorbij. There appears to be a shift in rainfall occurrence with a decrease in rainfall in autumn and an increase in rainfall during winter due to an increase in rainfall amount in August. As a result, there appears to be a delay in seasonal rainfall towards the end of winter and beginning of spring (July, August and September months).
Réchauffement climatique, quels impacts probables sur les vignobles ? 12 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007
2800
2600
2400
2200
2000
1800 Winkler index 1600
1400
1200 97) Ceres Klawer (1977-90) (1942-97) (1967-99) Wolseley Hermanus (1942-65) Biendonne Nederburg Nietvoorbij Constancia (1962-2002) (1970-2002) (1973-2002) (1978-2002) (1967-2005) Elgin (1963- (1982-2002) (1962-2002) Swellendam Lourensford Veldreserwe Weather stations
Winkler index with a Temp >0,5°C Winkler index Upper level Region I Upper level Region II Upper level Region III Upper level Region IV
Figure 8. Winkler index calculated for 12 weather stations in different wine producing areas of South Africa (Elgin and Hermanus in the Overberg district; Lourensford and Nietvoorbij in the Stellenbosch district; Bien Donné and Nederburg in the Paarl district; Constantia in the Constantia Ward; Veldreserwe in Worcester district; Wolseley in Tulbagh district; Swellendam in Swellendam district; Ceres in the Ceres Ward and Klawer in the Olifant river region) and also calculated using an even 0.5°C temperature increase over the growing season.
II.2. Impact for viticulture and a suggested response strategy
The increase in annual temperature implies increased evaporation and reduced water availability, which in turn would be expected to interfere with general grape physiology as well as reducing the duration of conditions for optimum physiology. It may cause some regions to become unsuitable for certain cultivars. The temperature increase during the ripening period (February) may hasten sugar accumulation in the grapes. A change or a loss of aroma in certain cultivar may be possible. Warmer temperatures are also associated with reduced acidity and higher pH values (Coombe, 1987).
With a significant temperature increase from April to July (results not shown), the winter season has shortened. As a result of a warmer late autumn, the leaf fall and the lack of sufficient cold during late autumn and beginning of winter may interfere with release from bud dormancy, which in turn may lead to irregular budburst, resulting in uneven growth, berry set, fruit ripening and reduced crop quality or crop loss. Increased use of chemical rest breaking agents may become necessary.
The delay in seasonal rainfall and increased late winter to early spring rainfall may result in increased disease pressure during the early season.
Assuming an even 0.5°C temperature increase over the year, the Winkler index for viticulture calculated for the different wine regions revealed that a shift towards more heat Réchauffement climatique, quels impacts probables sur les vignobles ? 13 Global warming, which potential impacts on the vineyards? 28-30 mars 2007 / March 28-30, 2007 tolerant varieties will be necessary. Changes in the areas of suitability for certain cultivars could be of major importance for the regional economy.
Adaptation, change in production practices and development of new wine regions are the keys to surviving climate change. The search for cooler locations towards the coast, where the moderating effect of the ocean is expected to temper increasing summer temperatures or at altitude, as well as the development of vineyards towards the eastern part of the Western Cape Province with an evenly distributed rainfall are an existing response to climate change in South Africa.
III. Conclusion
Significant climatic trends, similar across the different wine regions of South Africa were observed. The first signs of warming were visible in the maximum winter temperature during the late 1960’s and 1970’s. The significant breakpoint seems to occur in the mid 1980’s similarly to what is found in literature. The temperature increase has accelerated rapidly since 2000. The warming trend has reached values of on average 1°C for the annual temperature and 2°C for the February maximum temperature over the past three decades. The delay in winter rainfall must still be confirmed with longer time series of data and within the different regions. Further statistical analyses are under way to confirm the break point in the time series. Relationships between climatic data (including relative humidity, wind and evaporation) and vine phenology are also investigated in specific districts in order to study the potential impact of climate change on viticulture.
References
Coombe, B.G. (1987), Influence of temperature on composition and quality of grapes. Acta Hort. 206: 23-33. IPCC (2001), Third Assessment Report: Climate Change 2001 Synthesis Report, Stand-alone edition, Watson, R.T. and the Core Writing Team (Eds.) IPCC, Geneva, Switzerland, 184. Nhemachena C., Hassan R. & Benhin J. (2006), Report on Assessment of farmers’ adaptation strategies to climate change in Southern Africa, CEEPA, University of Pretoria, South Africa. SAWIS (2006), South African Wine Industry Statistics, Annual booklet n°30. WWF International (1996), Climate change and Southern Africa: an exploration of some potential impacts and implications in the SADC region, edited by Mike Hulme and published by Climatic Research Unit, School of Environmental Sciences, University of East Anglia, Norwich, UK and WWF International, Gland, Switzerland. ISBN: 2-88085-193-9.