Climate and kiwifruit cv. Hayward 2. Regions in New Zealand suited for production

Transcription

1 New Zealand Journal of Crop and Horticultural Science ISSN: (Print) (Online) Journal homepage: Climate and kiwifruit cv. Hayward 2. Regions in New Zealand suited for production M. J. Salinger & G. J. Kenny To cite this article: M. J. Salinger & G. J. Kenny (1995) Climate and kiwifruit cv. Hayward 2. Regions in New Zealand suited for production, New Zealand Journal of Crop and Horticultural Science, 23:2, , DOI: / To link to this article: Published online: 22 Mar Submit your article to this journal Article views: 709 View related articles Citing articles: 5 View citing articles Full Terms & Conditions of access and use can be found at

2 New Zealand Journal of Crop and Horticultural Science, 1995, Vol. 23: /95/ $2.50/0 The Royal Society of New Zealand Climate and kiwifruit cv. 'Hayward' 2. Regions in New Zealand suited for production M. J. SALINGER National Institute of Water and Atmospheric Research (NIWA) P.O. Box Auckland, New Zealand G. J. KENNY Centre for Environmental and Resource Studies (CEARS) University of Waikato Private Bag 3105 Hamilton, New Zealand Abstract Matching crops to climate is an important activity for planning production. Three important climatic factors were identified as being important determinants of kiwifruit (Actinidia deliciosa 'Hayward' (A. Chev) C.F. Liang et A.R. Ferguson) distribution: winter chilling; growing season thermal time; and annual rainfall. Indices for each of these factors were developed to enable mapping of the most suitable areas for kiwifruit production. These were May-July temperatures of 11 C or less as the optimal winter chilling requirement, a thermal time accumulation of 1100 degree-days above 10 C from October to April, and an annual rainfall of 1250 mm or more. Apart from Northland, all the traditional areas of kiwifruit production have a suitable climate. However, there are substantial areas of inland Bay of Plenty to Rotorua, the Waikato, north Taranaki, and northern Hawke's Bay where the climatic requirements are also satisfied. A high frequency of extreme winds may be an additional limiting factor in some of these regions, such as Taranaki. The total area of suitability is enlarged if irrigation is available, particularly in the southern North Island and central H94070 Received 23 November 1994; accepted 9 January 1995 Marlborough. Within this climatic range the estimated dates for the average end of dormancy and 50% flowering are spread over 4 weeks, and estimated budburst dates over 11 days. Although use of average climate data to describe kiwifruit distribution provides a valuable first-order assessment, it would be desirable to incorporate analyses of climate variability in future studies of this kind. Keywords kiwifruit; climate; phenology; Actinidia deliciosa; climate requirements; agroclimatic mapping; spatial distribution INTRODUCTION The mapping of agroclimatic zones is a well established method for delineating areas suitable for particular land uses or specific crop types. At the international scale, the broad relations of climate and land use were an integral component of the Agro-ecological Zones Project (AEZ) of the FAO (FAO ) which has formed the basis for more recent detailed country level studies (FAO 1993). There have also been a number of more specific studies on matching crops and climate (Oldeman & Frere 1982; Biswas 1988). At the country scale this has involved mapping of crop suitability using climatic indices (see for example, Mikkelsen & Olesen 1984). Recently this sort of approach has been used in combination with geographical information systems (GIS) (Winkler et al. 1990; Kenny & Harrison 1992). The mapping of such indices can best be validated by comparison with the known distribution of the crops being examined. However, it must be realised that other physical factors such as soils, and socio-economic factors will have a significant influence on actual (as against potential) crop distribution. Another recent approach is the mapping of land suitability based on simplified crop-simulation models (Brisson et al. 1992). This approach requires detailed

3 174 New Zealand Journal of Crop and Horticultural Science, 1995, Vol. 23 spatial and temporal data for both soils and climate, but allows the possibility of more comprehensive model validation and can provide more detailed crop information. In New Zealand there have been few attempts to map suitable climatic areas for specific cropping activities. Exceptions are Kerr et al. (1981) and Hurnard (1982). Kerr et al. (1981) used data from climate stations in the lower North Island to match horticultural crops and climate. Hurnard (1982) used the combined criteria of thermal time, frosts, and rainfall to identify areas at the mesoscale ( km) that were considered suitable for 'Riesling' grapes. A considerable research and knowledge base has been developed for a range of crops of economic importance to New Zealand, but despite the above initiatives there have, until recently, been no serious attempts to use this information in an integrated manner to map the climate and land suitability of crops. This paper provides a first step towards such an assessment for kiwifruit (Actinidia deliciosa 'Hayward' (A. Chev) C.F. Liang et A.R. Ferguson). Kiwifruit is a crop of fairly recent economic importance for which a considerable knowledge base exists. Because commercial production is presently based on a single cultivar the climatic limits can be defined, and mapped, in a fairly rigorous manner, as against most other temperate fruit crops where there is a wider genetic diversity and hence a range of tolerances to different climatic conditions. Kiwifruit growing has developed in many areas in New Zealand. Historically, it developed in the deep soils of the Bay of Plenty area where now c. 50% (Martin et al. 1990) of the production occurs. Subsequently kiwifruit production spread in Northland, Waikato, Poverty Bay, Hawke's Bay, Horowhenua, and Nelson, with small plantings in other areas, such as Taranaki and Wanganui. Some of these regions provide less than ideal conditions for consistent production of good quality kiwifruit. For example, it is now recognised that in Northland most winters provide insufficient winter chilling which has led to regular use of Hi-cane (hydrogen cyanimide) by growers to artificially induce the breaking of dormancy. Although there is a strong understanding of certain key phases of kiwifruit development and growth (McPherson et al. 1992; Salinger et al. 1993) there is incomplete understanding about others. There is a well established, positive linear relationship between temperature and the rate of development from budburst to flowering (Morley-Bunker & Salinger 1987 ; McPherson et al. 1992; Salinger et al. 1993). On the other hand, although there is partial knowledge about the effects of winter chilling on crop development (Salinger et al. 1993) there is still incomplete understanding of the biological response (McPherson pers. comm.). This paper aims to identify those climatic factors which are most likely to be dominant in determining kiwifruit distribution in New Zealand and to evaluate their effectiveness in mapping this distribution. This is a first-order assessment, aimed at determining the climatic envelope in which conditions are, on average, optimal for kiwifruit production. Within this envelope it will then be possible to examine in more detail (e.g., on a site-by-site basis) some of the more functional crop-climate relationships whichhave been previously described by Salinger et al. (1993). In addition it will provide the basis for future analyses incorporating, for example, land use capability data and enable assessment of the possible effects of climate change. KIWIFRUIT-CLIMATE RELATIONSHIPS Effect of temperature on vine and fruit development A review and analysis of the effects of temperature on kiwifruit development has been given by Salinger etal. (1993). Abrief summary of the most important relationships is given here. Temperature is the predominant climatic factor affecting kiwifruit vine and fruit development, from breaking of dormancy in late winter to fruit maturity. Winter temperature conditions are important for winter chilling of vines, which influences the breaking of dormancy, as first observed by Brundell ( 1976). Following the breaking of dormancy, warm spring and early summer conditions are important for subsequent development (Morley-Bunker & Salinger 1987; McPherson et al. 1992; Salinger et al. 1993). These studies have shown a significant, positive linear relationship, between temperature and the rate of development from budburst to flowering. No clear relationships have been established over the summer period, although it is likely that summer warmth is an important factor in fruit development. The effect of cool summers was observed in field surveys by Salinger et al. (1993) who found that fruit failed to form properly or reach maturity in 2 of 4 years surveyed at Rangiora (near Christchurch). This indicated that commercial cultivation was not viable at this cool margin. In the

4 Salinger & Kenny Climate and kiwifruit 2. Regions in NZ 175 autumn months it is evident that final fruit maturity is enhanced by lower mean temperatures (Seager et al. 1991; Salinger et al. 1993). Interannual variability in thermal time is also important. The standard deviation between growing seasons (October-April) amounts to degree-days (base temperature of 10 C) over New Zealand. Cool seasons, where insufficient thermal time is accumulated, will occur one in every 3 years at the southern (in Canterbury) and attitudinal (around Taupo on the volcanic plateau) extremities. At these locations, which represent the climatic margin, the frequent reduction in potential yield from adverse weather makes commercial production a highly marginal activity. Effect of extreme events on fruit development and yield Air frosts (when the temperature at 1.7 m falls below 0 C) damage or even kill the crop at sensitive growth stages. Frosts in spring are the most harmful because this is the period when there is new season's growth. The crop may be damaged if there is an insufficient frost free period (the period between the last frost in spring and the first frost in autumn). However, mature fruitcan often withstand an air frost in autumn because of the protection afforded by the leaf canopy. Active shoot growth commences after budburst in spring for kiwifruit vines. Salinger et al. (1993) found that budburst ranged from 15 September in Hawke's Bay to 8 October in Kerikeri. Other sites were intermediate. These dates fell much later than the 75 percentile dates of the last spring frost (Goulter 1981) at most sites except in Wairarapa, Nelson, central Marlborough, north Canterbury, and from Rotorua to Taupo on the volcanic plateau. Except around Taupo, the 75 percentile dates were within 3 weeks of budburst. However, in more extreme seasons, the 90 percentile dates show that air frosts pose a risk to crops in the above-mentioned areas. Growers in some of these localities already employ frost protection techniques to prevent damage from these one in 10-year events. Fruit maturity was attained before the 25 percentile date (one in 4 years) of first air frost in autumn at all sites used in the Salinger et al. (1993) study. In more extreme seasons 10 percentile dates (one in 10-year event) show air frosts occurring before fruit maturity (6.5% soluble solids (SS)) is attained in Manawatu, Horowhenua, Wairarapa, Nelson, central Marlborough, andnorth Canterbury. The first air frost usually damages the leaf canopy only, which affords protection to the fruit underneath. Economic considerations will determine the impact of a one in 10-year event of crop loss, at the expense of crop insurance or frost protection. Hail can devastate kiwifruit crops and in recent years has caused damage in Hawke's Bay and western Bay of Plenty (Steiner 1988). However, knowledge of the hail climate of New Zealand is limited. Severe hail information is limited to reports in newspapers, and Steiner (1988) concludes there is a requirement to gather more information on the occurrence of severe events. From the limited information, there was no apparent regional preference of damaging hail. Presently growers insure for this risk. Frequency of extreme wind conditions is also important. Shelter from wind is essential in most areas, and is the norm for kiwifruit orchards. Average wind conditions are not too different between the Bay of Plenty (average annual windspeed at Tauranga Airport is 17 km/h) and Taranaki and Manawatu-Wanganui (20 km/h at New Plymouth Airport and 17 km/h at Ohakea) (New Zealand Meteorological Service 1983). However the latter two sites have at least double the frequency of days with wind gusts greater than or equal to 63 km/h (of the order of 80 days/year compared to 34 days/year at Tauranga). These extreme conditions are potentially more damaging than the average, and a higher frequency of extreme winds may be considered a climatic limitation in regions that are otherwise suitable for kiwifruit. Effect of rainfall on vine and fruit growth The availability of water for vine growth is one of the main determinants for adequately sized fruit (Prendergast et al. 1987; Judd et al. 1989). Insufficient moisture leads to smaller fruit. Kerr et al. (1981) suggested a minimum annual rainfall of 1250 mm was required with 100 mm/month or more between December and March. The results of Judd etal. (1989)and Prendergast et al. (1987) confirmed these as the necessary moisture requirements. Ultimately the minimum rainfall required to match the evaporative demand is determined by the soil water storage and critical deficit for the soil type. When the deficit exceeds the critical deficit, growth ceases. However, in areas with insufficient rainfall, or where interannual variability is high, the deficit may be met through irrigation. This would be an added cost to production, and in severe drought years when demand for water is high there may be the situation of water rationing.

5 176 New Zealand Journal of Crop and Horticultural Science, 1995, Vol. 23 INDICES FOR MAPPING KIWIFRUIT SUITABILITY Based on the available information it was decided to use three climatic factors for a first-order assessment of the spatial suitability of kiwifruit in New Zealand. These were: a winter chilling requirement, a growing season thermal-time requirement, and a minimum rainfall requirement for unirrigated production. Winter chilling Salinger et al. (1993) have shown that the end of dormancy is strongly influenced by winter temperature. There was a strong, negative relationship between May, June, and July mean temperature and the end of dormancy. In general it appears that sufficient winter chilling occurs when the mean temperature over the May-July period is of the order of 11 C or less. Dormancy will be broken with less chilling but field surveys (Morley-Bunker & Salinger 1987; Salinger & Morley-Bunker 1988) showed that budburst was late with poor flowering which was spread over a lengthy period. Thermal time Many crops have a minimal thermal requirement over the growing season to attain maturity. A seasonal (October-April) thermal-time requirement for kiwifruit of at least 1100 C days above a base of 10 C was suggested by Kerr et al. (1981). This has been verified by field observation (Salinger etal. 1993). Rainfall The above factors will determine the thermal "envelope" in which the crop will, on average, develop to maturity. Many factors will determine the size of the crop, not least of which is vine management. However, it is clear that sufficient moisture is a prerequisite for consistent yields of saleable fruit. An annual rainfall requirement of at least 1250 mm has been defined as the minimum (Kerr et al ; Judd et al. 1989) with the need for irrigation in drier areas and in drier than average years. Indices have been developed to represent the climatic and geographic limits of each of these critical factors, which form the basis for a first-order spatial representation of kiwifruit suitability. These indices are: (1) winter chilling May-July mean temperature < 11 C; (2) seasonal thermal time 1100 C, base 10 C over the October-April growing period; and (3) rainfall requirement annual rainfall of at least 1250 mm. REFERENCE CLIMATE AND SYSTEM FOR MAPPING New Zealand has a complex topography which results in significant local modification of climate. It is well known that these local-scale modifications of climate are important for crop production (Geiger 1965; Yoshino 1975; Skaar 1980). For spatial mapping of crop suitability in New Zealand, it was important that a climatology be developed that represented, as well as possible, the variation of climate with topography. This was achieved using methods originally developed by Hutchinson (1989) and applied to New Zealand by Mitchell (1991, 1994). A description of the climate data used (based on Normals) is given in more detail by Kenny et al. (1995), whereas the method by which these data were fitted to the topography of New Zealand is described in Mitchell (1991). These climate data were incorporated in the CLIMPACTS system (Warrick & Kenny 1994; Kenny et al. 1995). In brief, CLIMPACTS is an integrated system of models and data, within a user-interface, which has been developed to allow assessments of the effects of climate variability and change on the New Zealand environment. There are three key components which were of relevance to the assessment of kiwifruit suitability: (1) the fitted climatology (monthly rainfall and temperatures); (2) a model for kiwifruit, which incorporated the three indices described above; and (3) the IDRISI GIS (Eastman 1992), which allowed for spatial display and analyses of these data. FIRST-ORDER MAPPING OF KIWIFRUIT SUITABILITY Each of the three indices were mapped separately (Figs 1-3) to show the climatic margins based on each of these. These were then combined (Fig. 4) to identify those regions most climatically suited (optimal) for kiwifruit, those that are limited and those that are unsuitable. Sufficient summer warmth is crucial for development to maturity. Regions of the North Island that are warm enough include Northland, Auckland, north Waikato, and most coastal regions

6 Salinger & Kenny Climate and kiwifruit 2. Regions in NZ 177 Fig. 1 Areas in New Zealand suitable for kiwifruit (Actinidia deliciosa) based on a thermal time requirement of 1100 degree-days, base 10 C for the October -April growing period. Suitable Unsuitable Water/cities I: 1 of both the east and west coast. In the South Island, the only regions that have warm enough summers, on average, are coastal Nelson and Marlborough (Fig. 1). Kiwifruit is grown in most of the regions shown in Fig. 1. However in Northland, growers have resorted to application of hydrogen cyanamide to artificially break dormancy, because there is generally insufficient winter chilling. The winter chilling index, defined above, is useful for identifying those areas that have sufficient summer warmth but low winter chilling (Fig. 2), which are largely confined to Northland. Rainfall is below the optimum in some important, or potentially important, regions including part of the Thames Valley, Waikato region, the coastal plain near Gisborne, Hawkes Bay, Wanganui- Manawatu, and Wairarapa in the North Island and coastal Nelson and much of Marlborough in the South Island (Fig. 3). Combining this information helps identify optimal areas, and equally important those that may be limited by one or two climatic factors. Optimal areas are those where the three climatic factors are favourable for kiwifruit production. Those areas with low degree-days (i.e., cool summers) are classed as unsuitable, as the crop would not, on average, reach maturity and would not be economically viable. Three classes are shown where there is some climatic limitation but kiwifruit production is still

7 178 New Zealand Journal of Crop and Horticultural Science, 1995, Vol. 23 Sufficient Insufficient Water/cities L 1 Fig. 2 Areas in New Zealand identified as having sufficient or insufficient winter chilling for successful kiwifruit (Actinidia deliciosa) production, based on a May-July mean temperature of viable. The least viable of these are those that are limited by low winter chilling. This is the factor which would probably be the most important in determining future viability of regions as a consequence of global warming. In low rainfall areas with sufficient chilling the long-term availability of water for irrigation would be the crucial factor in determining viability. Such areas are the least limited of those with some climatic limitation. In the North Island those areas that are optimal include the Waikato valley, a large area of the Bay of Plenty inland to Rotorua, Gisborne, northern Hawke's Bay, and north Taranaki (Fig. 4). Assuming that irrigation can be provided other favourable areas include the Thames Valley, central and southern Hawke's Bay, Wanganui, Manawatu, and Wairarapa. Suitable South Island areas are much more limited (Fig. 4). Most suitable areas require irrigation, including the Waimea plain, and the lower Wairau and Awatere Valleys in Marlborough. The spatial resolution of the data and the relatively arbitrary thresholds used for each of the three indices may have excluded some otherwise suitable areas. In the South Island, this includes Takaka. Generally the results confirm that the present distribution of kiwifruit is largely within areas with optimum climatic conditions. The Waikato region, although accounting for a small percentage of the total area of vines planted, is well suited to kiwifruit

8 Salinger & Kenny Climate and kiwifruit 2. Regions in NZ 179 Fig. 3 Areas in New Zealand with optimal, or less than optimal rainfall for kiwifruit (Actinidia deliciosa), based on an annual requirement of at least 1250 mm. ruit(actin Low rainfall Water/cities and most likely to other fruit crops. This region has traditionally been the heartland of dairy production in New Zealand, with very high land prices. Economics will be the main factor that will determine whether or not kiwifruit production expands in this, and other potentially suitable regions. In other areas that may be suitable, such as northern Hawkes Bay, around Wairoa the present optimism of the wine industry in relation to low returns for kiwifruit is leading to establishment of vineyards. SITE-BY-SITE ANALYSES To further the understanding of the effects of climate on kiwifruit, crop/climate relationships from Salinger et al. (1993) were used to estimate average dates of key developmental events for locations in climatically optimal areas. Using longer-term climate averages the timing of key phenological events were determined, based on Equations 1,3,6, and 8 from Salinger et al. (1993). These events included average dates of end of dormancy, budburst, 50% flowering, and 6.5% SS (fruit maturity). The dates of key development events from selected locations within the optimal areas are shown in Table 1. Average dates for end of dormancy span c. 4 weeks for the crop range of kiwifruit in New Zealand, and show a south to north progression. Dormancy ends in early August in North Canterbury

9 180 New Zealand Journal of Crop and Horticultural Science, 1995, Vol. 23 Optimal Limited Marginal Unsuitable Water/cities Fig. 4 Suitability for kiwifruit (Actinidia deliciosa) production in New Zealand, based on indices for winter chilling, summer warmth, and annual rainfall. ff (a highly marginal area), Nelson, and the volcanic plateau (marginal). In contrast, it occurs at the end of August in the Thames Valley, Bay of Plenty, and north Taranaki. Average budburst dates only span 11 days. Earliest dates are in the South Island districts and latest dates in north Taranaki. However, average 50% flowering dates are spread over 3 weeks. Hawke's Bay and Thames Valley are the earliest and north Canterbury and the volcanic plateau the latest. Maturation dates (6.5% SS) are also spread over 3 weeks. Based on the estimations of maturation dates northern locations were the latest to mature in early May, with earliest maturation being achieved in Canterbury and central Marlborough. This range of dates is consistent with Snelgar et al. ( 1993), however it was cautioned that the equations for calculation of maturation, derived by Salinger et al. (1993) and Snelgar et al. (1993), do show some inconsistencies. These results are interesting, as the areas previously identified as optimal are those with later maturation times. This is not an undesirable outcome as a longer maturation time ensures a longer period for both plant and fruit growth. The main risk would be from early autumn frosts, although this is generally not a problem in the optimal areas of the North Island, particularly in the coastal Bay of Plenty.

10 Salinger & Kenny Climate and kiwifruit 2. Regions in NZ 181 DISCUSSION A first-order mapping of kiwifruit suitability in New Zealand has been successfully achieved. This is an important advancement, given the topographical and climatic variation within the country and the generally discrete distribution of areas suitable for fruit production. The temperature margins for kiwifruit production in New Zealand were identified, based on low winter chilling in warmer northern regions (notably Northland) and summer thermal time in inland, high attitudinal regions of the North Island and coastal Marlborough and North Canterbury in the South Island. Within these margins there were large areas with sufficient rainfall, which were considered optimal for kiwifruit production. Kiwifruit is grown, or could potentially be grown, in areas thermally suited but limited by low rainfall, providing water is available for irrigation. In general, the climatic areas delineated for kiwifruit production compare well with present orchard distribution. The Bay of Plenty and Waikato is the largest area where all three criteria are met. The former is presently the greatest single area of kiwifruit production in New Zealand (Martin et al. 1990). Other important areas include Gisborne, Hawke ' s B ay, Nelson and Golden B ay, with smaller plantings in Wanganui and Horowhenua. However, there are still significant areas of New Zealand where the climate is suitable for kiwifruit production, especially if there is provision for irrigation. The most suitable areas which do not have large areas in production include the Waikato, Rotorua, northern Hawke's Bay, and north Taranaki. All these areas have the benefit that irrigation requirements are low. The provision of irrigation enables large areas to be utilised for this activity especially in Wanganui, Manawatu, Gisborne, Hawke's Bay, Wairarapa, Nelson, and central Marlborough. Table 1 Estimated average dates of dormancy end, budburst, 50% flowering, and 6.5% soluble solids (SS) (fruit maturity) for locations with a climate suitable for kiwifruit (Actinidia deliciosa) production. Location Dormancy end Budburst end 50% flower 6.5% SS Ardmore Ruakura Paeroa Te Aroha Tauranga Te Puke Whakatane Rotorua Taupo New Plymouth Mokau Patea Wanganui Ohakea Kairanga Waiterere Paraparaumu Ruatoria Gisborne Wairoa Taradale Masterton Tauherenikau Takaka Riwaka Nelson Blenheim Grassmere Waipara Christchurch Lincoln 29 Aug 21 Aug 25 Aug 29 Aug 28 Aug 26 Aug 20 Aug 18 Aug 10 Aug 26 Aug 30 Aug 27 Aug 25 Aug 21 Aug 16 Aug 18 Aug 21 Aug 29 Aug 26 Aug 25 Aug 26 Aug 13 Aug 16 Aug 16 Aug 12 Aug 10 Aug 14 Aug 16 Aug 6 Aug 3 Aug 3 Aug 24Sep 19Sep 21 Sep 23Sep 22 Sep 22 Sep 19 Sep 21 Sep 19 Sep 23 Sep 25 Sep 23 Sep 22 Sep 20 Sep 18 Sep 20 Sep 21 Sep 23 Sep 22 Sep 21 Sep 22 Sep 18 Sep 19 Sep 18 Sep 17 Sep 20 Sep 17 Sep 18 Sep 16 Sep 16 Sep 14 Sep 23Nov 23Nov 18Nov 16Nov 17Nov 23Nov 23Nov 30Nov 5 Dec 28Nov 26Nov 28Nov 23Nov 27Nov 28Nov 1 Dec 30Nov 20Nov 20Nov 19Nov 17Nov 1 Dec 29Nov 29Nov 28Nov 1 Dec 26Nov 25Nov 3 Dec 3 Dec 5 Dec 5 May 3 May 4 May 6 May 6 May 6 May 4 May 2 May 29 Apr 5 May 6 May 2 May 2 May 29 Apr 26 Apr 26 Apr 1 May 7 May 5 May 4 May 2 May 27 Apr 29 Apr 31 Apr 2 May 30 Apr 17 Apr 14 Apr 17 Apr 17 Apr 23 Apr

11 182 New Zealand Journal of Crop and Horticultural Science, 1995, Vol. 23 The margins of production were identified using an average climatology. Although this was generally sufficient for the purposes of this study, the viability of kiwifruit in any particular area will depend much more on the interannual variability of climate than the long-term average climate. Ideally this study should be extended, by developing the capacity to map areas that are at risk from either a high frequency of adverse years, or years with a high frequency of extreme events. This would help identify areas, particularly at the climatic margins, which are marginal for production. Factors that could be considered, given availability of data, are frequency of warm winters, frequency and timing of frost, frequency of cool summers, and frequency of low rainfall years. This type of analysis has previously been used by Winkler et al. (1990) and Kenny & Harrison (1992) to identify marginal and suitable areas for peaches in Michigan, United States and wine grapes in Europe respectively. Another factor which should also be considered is windiness. Shelter is an important requirement for protection against wind damage, most particularly in south-western regions of the North Island (Taranaki, Manawatu-Wanganui) where a greater frequency of extreme wind conditions is experienced. Within the climatic margins of production there is a wide diversity of land use and soil types. Future analyses should account for at least some of these factors, for example by using the New Zealand land use capability (LUC) classification (Water and Soil Division 1979) to identify areas of arable and non-arable land. This would allow for a more accurate representation of land areas either presently suitable or potentially suitable for kiwifruit. Site-by-site analysis of kiwifruit-climate relationships provided estimates of average dates of vine developmental events in both the suitable and marginal locations. The average climatological dates give the typical expectation of budburst, flowering, and fruit maturity in these locations, based on a reference period of recent climate. The dates of flowering and fruit maturity were spread over 3 weeks for the climatic range of kiwifruit in New Zealand. Dormancy dates reflect exposure to low temperatures whereas 50% flowering depends on spring warmth. Budburst dates do not show a clear pattern. The pattern of 6.5% SS fruit maturation depends on exposure to lower mean daily temperatures in autumn. The linking of these type of relationships with the spatial analysis would provide more detailed spatial information. Agroclimatic mapping is a very useful method by which to delineate areas suitable for crop production. Such methods are being applied in an integrated way, within the context of examining the possible effects of climate variability and change, with the development of the CLIMPACTS system for New Zealand (Kenny et al. 1994; Warrick & Kenny 1994). By this method areas in New Zealand where the climate is suitable for different sectors of the economy, including kiwifruit production, can be identified. However, other environmental factors, especially land use capability and soil type, are very important for successful production. Only average conditions have been used for the present study, but future studies should aim to incorporate spatial mapping of climatic risk. Finally, the phenological relationships could also be extended to the spatial analysis. This would enhance understanding of the nature and variability of kiwifruit production in New Zealand, and help identify further gaps in the knowledge base and areas of existing knowledge which require further refinement. These analyses could be coupled with a more detailed study of specific climate districts, in particular those at the climatic margins. ACKNOWLEDGMENTS This research was funded by the New Zealand Foundati on for Research, Science and Technology projects BIO91387, UOW404, and CO1318. This work would not have been possible without the intellectual input of Richard Warrick in development of the CLIMPACTS system and of Neil Mitchell in development of the spatial climatology. The contribution of Graham Sims to development of the CLIMPACTS system, which facilitated the spatial analysis, is also acknowledged. Finally, thanks to Jenny Alder for her contribution to the preliminary mapping work. REFERENCES Biswas, B. C. 1988: Agroclimatology of the sugarcane crop. World Meteorological Organisation technical note p. Brisson, N.; King, D.; Nicoullaud, B.; Ruget, F.; Ripoche, D.; Darlhout, R. 1992: A crop model for land suitability evaluation: a case study of the maize crop in France. European journal of agronomy 1(3): Brundell, D. J. 1976: The effect of chilling on the termination of rest and flower bud development of the Chinese gooseberry. Scientia horticulturae 4:

12 Salinger & Kenny Climate and kiwifruit 2. Regions in NZ 183 Eastman, J. R. 1992: IDRISI, Version 4: User's Guide. Clark University, Graduate School of Geography. Geiger, R : The climate near the ground. Cambridge, Massachusetts, Harvard University Press. FAO : Report on the agro-ecological zones project. Volume 1, Methodology and results for Africa; Volume 2, Results for Southwest Asia; Volume 3, Methodology and results for South and Central America; Volume 4, Results for Southeast Asia. FAO World Soil Resources report, FAO, Rome. FAO 1993: Agro-ecological assessments for national planning: the example of Kenya. FAO soils bulletin 67. FAO, Rome. Goulter, S. W. 1981: An air frost chronology for New Zealand. New Zealand Meteorological Service miscellaneous publication p. Hurnard, S. M. 1982: Climatological aspects of grape growing in New Zealand. New Zealand Meteorological Service technical note p. Hutchinson, M. F. 1989: A new objective method for spatial interpolation of meteorological variables from irregular networks applied to the estimation of monthly mean solar radiation, temperature, precipitation and windrun. CSIRO Division of Water Resources technical memorandum 89/5: Judd, M. J.; McAneney, K. J.; Wilson, K.S. 1989: Influence of water stress on kiwifruit growth. Irrigation science 10: Kenny, G. J.; Harrison, P. A. 1992: The effects of climate variability and change on grape suitability in Europe. Journal ofvane research 3(3): Kenny, G. J.; Warrick, R. A.; Mitchell, N. D.; Mullan, A. B.; Salinger, M. J. 1995: COMPACTS: an integrated model for assessment of the effects of climate change on the New Zealand environment. Global ecology andbiogeography letters. In press. Kerr, J. P.; Bussell, W.; Hurnard, S.; Sale, P.; Todd, J.; Wilton, J.; Wood, R : Matching horticultural crops and the climates of the lower North Island. DSIR Plant Physiology Division technical report p. Martin, R. J.; Salinger, M. J.; Williams, W. M. 1990: Agricultural industries. Pp in: MfE (Ministry for the Environment). Climatic change: impacts on New Zealand. Wellington, Ministry for the Environment. McPherson, H. G.; Hall, A. J.; Stanley, C. J. 1992: The influence of current temperature on the time from bud break to flowering in kiwifruit (Actinidia deliciosa). Journal of horticultural science 67(4): Mikkelsen, S. A.; Olesen, J. E. 1984: Computer-aided mapping of growing degree-days for Denmark, calculated from monthly temperature normals. Ada Agriculturae Scandinavia 34: Mitchell, N. D. 1991: The derivation of climate surfaces for New Zealand, and their application to the bioclimatic analysis of the distribution of kauri (Agathis australis). Journal of the Royal Society of New Zealand 21(1): Mitchell, N.D. 1994: The development of a New Zealand climatology for the CLIMPACTS project. In: Towards an integrated approach to climate change impact assessment. Report of a workshop held on 26 April Pittock, A. B.; Mitchell, C. D. ed. Australia, CSIRO Division of Atmospheric Research. Monteith, J. L. 1981: Climatic variation and the growth of crops. Quarterly journal of the Royal Meteorological Society 107: Morley-Bunker, M. J.; Salinger, M. J. 1987: Kiwifruit and development: the effect of temperature on budburst and flowering. Weather and climate 7: New Zealand Meteorological Service. 1983: Summaries of climatological observations to New Zealand Meteorological Service miscellaneous publication 177. Oldeman, L. R.; Frere, M. 1982: A study of the agroclimatology of the humid tropics of South-East Asia. World Meteorological Organisation technical note p. Prendergast, P.; McAneney, K. J.; Astill, M. S.; Wilson, A. D.; Barber, R. F. 1987: Water extraction and fruit expansion by kiwifruit. New Zealand journal of experimental agriculture 15: Salinger, M. J.; Kenny, G. J.; Morley-Bunker, M. J. 1993: Climate and kiwifruit cv. Hayward 1. Influences on development and growth. New Zealand journal of crop and horticultural science 21: Salinger, M. J.; Morley-Bunker, M. J. 1988: Theeffect of climate on kiwifruit flowering and maturity. New Zealand kiwifruit special publication 2: Ab-M. Seager, N. G.; Hewett, E. W.; Warrington, I. J.; MacRae, E. A : The effect of temperature on the rate of kiwifruit maturation using controlled environments. Ada horticulturae 297: Skaar, E. 1980: Application of meteorological data to agroclimatic mapping. International journal of biometeorology 23(1): Snelgar, W. P.; Hopkirk, G.; McPherson, H. G. 1993: Predicting harvest date for kiwifruit: variation of soluble solids concentration with mean temperature. New Zealand journal of crop and horticultural science 21: Snelgar, W. P.; Manson, P. J.; Hopkirk, G Effect of overhead shading on fruit size and yield potential of kiwifruit {Actinidia deliciosa). Journal of horticultural science, 66(3):

13 184 New Zealand Journal of Crop and Horticultural Science, 1995, Vol. 23 Steiner, J. T. 1988: The hail problem. The orchardist of New Zealand: Warrick, R. A.; Kenny, G. J. 1994: CLIMPACTS: the conceptual framework and preliminary development. In: Towards an integrated approach to climate change impact assessment. Report of a workshop held on 26 April Pittock, A. B.; Mitchell, C. D. ed. Australia, CSIRO Division of Atmospheric Research. Warrington, I : Kiwifruit the latest on fruit yield, quality. Horticultural news, December: Water and Soil Division 1979: Our land resources. A bulletin to accompany New Zealand land resource worksheets. Wellington, New Zealand, Ministry of Works and Development. Winkler, J. A.; Burnett, A. W.; Skipper, B. J.; Moore, J. B.; Mulugeta, G.; Olson, J. M. 1990: Agroclimatic resource assessment: an example for peach cultivation in the lower peninsula of Michigan. Physical geography 11: Yoshino, M. M Climate in a small area. Tokyo, University of Tokyo Press.