Effects of fertilization and climatic factors in a long-term experiment with blackcurrant (Ribes nigrum L.) cv. Ben Tron

Abstract

Two-year-old blackcurrant plants cv. Ben Tron were planted in 1991 to investigate long-term effects of seven fertilization strategies. Broadcast fertilization was given in spring and autumn, fertigation from May until August, or a combination of fertigation and broadcast fertilization. Three fertilizer rates were used, and the amount was increased three times during the trial because of low mineral content in the leaves and insignificant yield response. Yield parameters, macro nutrient content in leaves and content of soluble solids in fruit juice were recorded over 11 years. The variation between years was significant for all parameters recorded, but the fertilization strategies had only a minor effect over time. Content of macro nutrients in leaves was low compared to recommended values, and did not respond significantly to increasing fertilizer amounts. Yield decreased with plant age. Frost reduced yield in at least two years, but few significant correlations with precipitation and temperature were found. A good water access seems to be important for a stable fruit yield, while precipitation during flowering is likely to reduce yield because of fruit drop.

Keywords:

• Fertigation
• fruit drop
• plant analysis
• soluble solids
• yield

View correction statement:

CORRIGENDUM

Introduction

In an open field crop, various abiotic factors such as temperature and precipitation influence fruit yield and quality. Yield variation between years is therefore common in commercial blackcurrant production, but long-term studies of the relations between climate and yield have not been reported. Basic information on growth and development in blackcurrant plants (Bould, 1955; Fernqvist, 1961; Atkinson, 1972; Rhodes, 1986) are useful in understanding the biology, but does not discuss long-term external influence.

Several factors affect yield components in blackcurrant (Nes, 1978), and one important factor is water access during fruit development. Several experiments have demonstrated how soil moisture and irrigation increase shoot growth (Kongsrud, 1969; Goode & Hyrycz, 1970; Kongsrud, 1970; Niskanen et al., 1993), which is essential to obtain a high yield (Rhodes, 1986; Hoppula & Salo, 2005). Precipitation during the blossoming period is on the other hand negative, and may increase fruit drop (Nes, 1978). Factors occurring in one year also influence yield the following year (Wilson & Jones, 1980; Heiberg, 1986b), making yield prediction complex.

Plant analyses are widely used in fruit crops. Leaf nutrient composition in blackcurrants has been investigated, and Niskanen (2001) found a close relationship between leaf values and nutrient content in the soil. Others found no differences in leaf mineral content after different fertilization treatments (Kongsrud, 1986; Nes, 1990; Aaltonen & Dalman, 1993). Although there is a well-documented link between growth and leaf mineral content (Ljones, 1966b; Niskanen, 2001), the use of leaf analysis in fertilizer recommendations may be questioned (Niskanen, 2002).

The present experiment was designed to evaluate the long-term effects of fertigation and broadcast fertilization, time of application and amounts of fertilizer in a blackcurrant field. As the fertilization strategies used in this trial only showed a minor effect on yield parameters, climatic factors have been related to the results.

Material and methods

The experiment was carried out at Apelsvoll Research Centre Division Kise, Norway (60° 40′N, 10°11′E) during the period 1991–2003. Mean temperatures (1961–1990) at the experimental site during the growing period (May–September) are 12.5°C, and −1.2°C in October–April. Minimum temperature during the trial was −31.4°C (February 1996). Precipitation (Table I) and temperature were recorded continually, and used for evaluating the biological parameters recorded.

Table I. Monthly precipitation (mm) at the experimental site during the trial.

Month	1993	1994	1995	1996	1997	1998	1999	2000	2001	2002	2003	Average 1961–90
May	68	23	58	82	106	6	22	43	41	79	69	38
June	28	44	79	59	62	118	124	71	47	30	65	63
July	134	4	79	52	50	54	48	85	37	77	49	82
August	99	116	17	65	41	47	20	68	91	21	80	70
September	32	83	42	47	49	71	104	26	60	21	96	64
Total	361	270	275	305	307	297	319	294	275	228	359	317

The trial was established with two-year-old plants of the Scottish cultivar Ben Tron, arranged in a randomized block design. Each treatment of eight bushes was replicated three times. The field was planted in the spring of 1991 in single rows 4.0 m apart and at a plant distance of 1.5 m in the row (2500 m row/ha). The soil of the experimental field was a morainic loam with 6–8% humus. Plant available P and K at planting was 90 and 250 mg kg⁻¹, respectively, which is in the upper recommended range. Before planting, 500 kg per ha of 15:4:12 NPK compound fertilizer was applied. The whole experimental field was fitted with a pressure compensated trickle irrigation system with an emitter spacing of 50 cm and a capacity of 1.6 l h⁻¹. Until 1997 the field was irrigated when the soil had a soil water deficit of 10 mm. Timing and amount of water supplied was equal for all treatments. From 1998 irrigation was performed once a week. Short grass covered the alleyways combined with a 1.0 m wide herbicide strip in the rows. All fertilizer was applied to the herbicide strip only, thus only 25% of the area was fertilized. Plots were fertilized either with dry compound fertilizer dispensed in the herbicide strip, fertigated through the trickle irrigation system or a combination of the two methods (Table II). When the application methods were combined, half of the total nutrient quantity was broadcast and the other half given as nutrient solution. Plots with fertigation were treated weekly from the last week of May until the first week of August. Plots without fertigation were irrigated at the same interval as fertigated plots. Broadcast fertilizer was a 15:4:12 NPK compound fertilizer, while the compound fertilizer Superba™ Gul (3:3:19 NPK) was used alternatively with Kalksalpeter™ (Ca(NO₃)₂) containing 15.5% N, in the fertigated plots. All fertilizers used were from Norsk Hydro AS (now Yara). Broadcasting in spring was performed in early May, and broadcasting in late summer or autumn about 20 August (after harvest). The amount of fertilizer applied was adjusted during the experimental period (Table II) because of low leaf mineral nutrient content. Fungicides were applied in 1993–1996. Because of no detectable effects, no fungicides or insecticides were applied later. The bushes were lightly pruned. In 1997 and 2002 winter damage reduced the number of yielding shoots significantly.

Table II. Methods, timing and amounts of nitrogen (kg) applied per 1000 m row (0.4 ha) in different periods of the trial.

Methods of fertilization	Timing	Treatment	1991–1995	1996–2000	2001–2003
Broadcast	Spring	B(S)	5	6	8
Fertigation	May–August	F	5	6	8
Broadcast	Autumn	B(A)	5	6	8
Broadcast + Fertigation	Spring + May–August	B(S) + F	10	12	16
Broadcast	Spring + Autumn	B(S) + B(A)	10	12	16
Fertigation + Broadcast	May–August + Autumn	F + B(A)	10	12	16
Broadcast + Fertigation + Broadcast	Spring + May–August + Autumn	B(S) + F+B(A)	15	18	24

The plants were harvested manually, and records of total yield, berry and cluster size were taken at harvest all years. Samples of leaves for analysis were collected at the end of August or the beginning of September every year except 1998, using leaves positioned on the upper third part of the shoot. Leaves were sampled from a minimum of 20 young shoots, randomly selected from the whole plot. Plant macro nutrient contents were analysed after sulphuric/hypochloric acid digestion, using a Skalar autoanalyser for N and P, flame photometry for K and atomic spectrophotometry for calcium and magnesium. Content of soluble solids was measured every year at harvest in berries sampled randomly from each plot, using a digital hand refractometer (ATAGO PR-1). All data were subjected to statistical analysis using MiniTab® statistical software. Two-way analysis of variance and the least significant difference (LSD) between means were determined at p<0.05. For not normal distributed data (phosphorus content in leaves and weight of 100 berries), the Friedman test was used. For correlations with climatic factors, the Pearson product moment correlation coefficient (r) was calculated.

Results

Yield

The cultivar Ben Tron gave a good crop from the second year after establishment (Figure 1). The yield varied distinctly between years (p<0.01). The average yield during 11 years was close to 7 metric tons ha⁻¹ and varied between 2.9 tons ha⁻¹ in 2002 and 9.5 tons ha⁻¹ in 1994. In total for the experimental period, there was a negative trend in yield (r= − 0.572; p<0.001), independent of fertilization strategy.

Figure 1.  Annual yield (t ha⁻¹) in blackcurrant cv. Ben Tron. Mean of seven fertilization treatments. Arrows indicating fertilizer increase.

Severe winter damage in 1997 and 2002 resulted in very low yield in these years, and heavy pruning was necessary to remove damaged shoots and branches. Also in 2000, some branches were affected by winter damage, but the effect on yield is not known.

In an average of 11 years, broadcasting in early May gave a significantly lower yield than most of the other treatments (Table III). Fertigation and broadcasting in autumn with low fertilizer rate gave a high yield compared to the amount of fertilizer added. The increase in fertilizer rate in 1996 and 2001 did not seem to have any influence on yield in the subsequent years.

Table III. Yield (t ha⁻¹) in blackcurrant cv. ‘Ben Tron’ during three periods with different fertilization strategies and increasing fertilizer amounts.

Treatment	1993–1995	1996–2000	2001–2003	Mean 1993–2003
B(S)	7.13	5.76	3.60	5.55
F	9.17	7.17	4.63	7.02
B(A)	8.47	6.64	5.13	6.73
B(S) + F	8.27	7.18	4.22	6.67
B(S) + B(A)	8.94	7.18	4.39	9.90
F + B(A)	8.92	7.67	4.57	7.17
B(S) + F+B(A)	10.24	7.27	4.28	7.26
Mean	8.74	6.98	4.40	6.76
LSD	n.s.	n.s.	n.s.	1.30

Berry size

Broadcasting of fertilizer in spring gave the smallest berries in most years, and significantly smaller berries than most treatments in total for all years (data not shown). Berry size increased with increasing fertilizer rate, from 106 to 114 g per 100 berries (Figure 2), but the only statistically significant difference between treatments and amounts was for broadcasting in spring. Total precipitation from May to August significantly affected berry size in plots with fertilizer broadcasted in spring (r=0.621; p<0.05). The berry size was larger in the first years of the trial, and the variation between years was significant (p<0.001) (data for single years not shown).

Figure 2.  Berry size (g per 100 berries) and cluster size (number of berries per 100 cluster) in blackcurrant cv. Ben Tron with three fertilizer levels during three periods. Mean of different fertilization strategies.

Cluster size

Numbers of berries per 100 clusters were highest when the highest amount of fertilizer was given (Figure 2), but only in 1997 was the difference between treatments significant. The variation between years was significant and varied from 537 to 869 (data for individual years not shown). Broadcasting alone gave on average few berries per cluster, while fertigation affected cluster size positively in most years.

Fruit drop

Fruit drop was recorded only in the seasons from 2001 to 2003 (data not shown). No significant differences between the treatments were found. The variation between years was substantial: in 2003 the fruit drop on average was close to 43%, while in 2001 it was only 10%. A correlation coefficient of 0.993 indicates a positive connection between degree of fruit drop and days with precipitation in May, but the correlation was not significant.

Macro nutrient content in leaves

Leaf macronutrient content showed significant response to fertilization in the first years of the trial (Table IV), but from 1999 no significant differences between treatments were found. The recommended concentrations (%) of NPK in the dry matter of currant leaves are 2.6–3.0, 0.2–0.3 and 1.2–1.8, respectively. According to this, the mean concentration of nitrogen was too low in most years for all treatments (Figure 3). Broadcast fertilization in May had the lowest effect on leaf nitrogen content in most years (Table IV).

Figure 3.  Yearly variation in leaf NPK in blackcurrant cv. ‘Ben Tron’ as percent of dry matter. Average of seven fertilization treatments from 1993–1997 and 1998–2003. Arrows indicate recommended range.

Table IV. Mean content of NPK (%) in leaf dry matter in blackcurrant cv. ‘Ben Tron’ with three fertilizer levels during three periods. Mean of seven fertilization treatments.

	1993–1995	1996–2000	2001–2003	Mean
N
Low	2.23	1.87	2.12	2.07
Medium	2.35	2.03	2.17	2.18
High	2.42	2.06	2.27	2.25
Mean	2.31	1.96	2.16	2.14
LSD	0.12	0.25	n.s.	0.15
P
Low	0.29	0.17	0.29	0.25
Medium	0.27	0.19	0.37	0.28
High	0.25	0.21	0.31	0.26
Mean	0.28	0.18	0.33	0.26
LSD	0.02	n.s.	n.s.	n.s.
K
Low	1.28	1.01	1.34	1.21
Medium	1.22	1.03	1.24	1.16
High	1.17	1.00	1.47	1.21
Mean	1.24	1.02	1.31	1.19
LSD	n.s.	n.s.	n.s.	0.03

The phosphorus leaf content was within or above the recommended range in most years, but from 1996 to 2000 the phosphorus content was low (Table IV), despite increased application. The second fertilizer increase in 2001 augmented the phosphorus level, and the leaf content was adequate the following years (Figure 3).

The potassium content was low until the second fertilizer increase in 2001 (Figure 3). There was a considerable variation between years, and no strategy had a significantly different effect on leaf potassium content (Table IV). Nevertheless, in all years except 2000, fertigation combined with broadcasting in spring gave lower potassium content than when only half the amount of fertilizer was broadcast in spring. Broadcasting in spring with the lowest fertilizer rate improved the potassium accumulation in leaves.

Content of calcium and magnesium in leaf dry matter was registered only in the last four years of the trial, and no significant differences between treatments were found (data not shown). Calcium content was on average 1.37% and varied between 1.24 (broadcast, spring) and 1.46%. Mean magnesium content varied only from 0.24 to 0.27% between treatments.

Macro nutrient content in berries

There were only small and insignificant differences between treatments in accumulations of any of the macronutrients analysed (data not shown). The content of nitrogen was significantly lower than in leaves, and varied from 0.70 to 0.91% of dry matter. Plots with broadcast fertilizer in spring had lower nitrogen content compared to plots fertilized in autumn. Phosphorus and potassium content was approximately at the same level as in leaves; 0.21 and 1.22% in average of all treatments. Also in berries there was a tendency that broadcasting of fertilizer in spring resulted in the lowest nitrogen content, while low potassium content was found in plots fertilized in the autumn. The highest fertilizer amount showed the lowest calcium accumulation in berries (0.193%), while broadcasting in spring together with fertigation resulted in the lowest magnesium content (0.077%). The differences were, however, small and not significant. No correlation between macro nutrient content in berries and any of the yield components was found.

Soluble solids

The various fertilization strategies caused no significant differences in content of soluble solids (SSC) in the berries in any single year (data not shown). On average for all years, broadcasting of fertilizer in spring with low fertilizer rate gave a significantly higher content of soluble solids than the other treatments (Table V). The variation between years was from 12.9 to 15.9%. No correlation between SSC and yield was found, nor between SSC in berries and accumulated macro nutrients in leaves. Winter damage in 1997 and 2002 severely affected yield but had various effects on SSC. The SSC in berries was high in the year with lowest yield (2002).

Table V. Soluble solids (Brix) as mean of 11 years after seven fertilization treatments.

Treatment	Soluble solids (Brix)
B(S)	15.1
F	14.2
B(A)	14.5
B(S) + F	14.6
B(S) + B(A)	14.6
F + B(A)	14.2
B(S) + F+B(A)	14.1
Mean	14.5
LSD	0.81

Discussion

Yield components

High fertilizer rate gave the highest yield, although the effect was relatively small, considering the difference in fertilizer consumption.

The extension growth is important for the cropping potential, and number of clusters is closely relateds to the total extension growth (Hansen, 1986; Rhodes, 1986; Nes, 1990), as blackcurrant bear fruits mostly on young wood (Ljones, 1966a; Rhodes, 1986). Competition within the plant may give decreased vegetative growth when the crop is heavy, but some compensation may take place due to prolonged extension growth after harvest (Hansen, 1986). A good nutrient status in the plant after harvest would therefore be positive, and the present trial indicated that an adequate nutrient access throughout the season combined with an additional fertilization after harvest was positive.

Low precipitation in June decreased yield in the last years of the trial, when irrigation was not adjusted according to evaporation and precipitation (r = 0.830; p<0.05). This indicates that the water demand in the trial was underestimated, as irrigation increases shoot growth in Ribes (Kongsrud, 1969; Goode & Hyrycz, 1970; Kongsrud, 1970; McCarthy & Stoker, 1988; Ostermann & Hansen, 1988; Niskanen et al., 1993; Dencker & Hansen, 1995; Hofman, 1995), as well as berry size (McCarthy & Stoker, 1988) and number of berries per cluster (Ostermann & Hansen, 1988).

Fluctuations in berry yield between years are common (Niskanen, 2001). This was also true in the present trial, and none of the fertilization strategies prevented these variations, indicating that the variation was due to other factors. Low yield in three of the years was due to winter damage. In the spring of 1997 the temperatures were relatively high for a period, followed by lower temperatures later. It is well known that such conditions may result in decreased hardiness in plants. Winter damage in Ben Tron has been reported earlier, even when other cultivars were not affected (Heiberg & Måge, 1991). Nevertheless, in February 1996 the minimum temperature was –31.4°C, and no winter damage was registered in any plots. It seems therefore that Ben Tron has an acceptable tolerance for low temperatures during the dormant period. Frost damage during flowering was not found in any years.

Ben Tron has large berries and medium sized clusters (Heiberg & Måge, 1991; Nes & Meland, 1992). In the first years of the trial, the cluster size was acceptable for all treatments (Figure 2). In this experiment fertilizer treatments had some effects on berry size, but the main variation was between years. Heiberg (1986a) suggested that precipitation in July could be a reason for variable berry size. In this trial, total precipitation from May to August significantly affected berry size in plots where fertilizer were broadcasted in spring. The low berry size in 1998 might therefore be related to limited water access in spring, with only 6 mm precipitation in May. When using broadcast fertilizer, sprinkle irrigation is necessary to make the nutrients available to the plants in years with low precipitation. The amount of water supplied through the drip irrigation system was probably inadequate to secure a good berry size when precipitation was low.

Fertilizer treatments demonstrated a minor effect on cluster size. Cluster size in the last years of the trial, after increasing the fertilizer amount twice, was the lowest recorded. Cluster size is affected by the degree of fruit drop, which was recorded in too few years to test any long-term correlations. Fruit drop is a common and well-documented reason for yield variation in currants (Fernqvist, 1961; Nes, 1978; Wilson & Jones, 1980; Heiberg, 1986b; Kühn, 1988; Toldam-Andersen & Hansen, 1993). Fruit drop or reduced fruit set have been connected with insufficient pollination (Fernqvist, 1961 with references) or low temperatures (Dale, 1984; Kühn, 1988), but also inadequate water access has been found to decrease fruit set and development (Kongsrud, 1969). In the present trial, fruit drop was largest in 2003, a year with high yield compared to the two previous years. Precipitation was moderate in June 2001 and 2002, while in 2003 water access was good. In the present trial, number of days with precipitation in May corresponded with the degree of fruit drop, as reported by Nes (1978). Number of days with precipitation in May 2001 and 2002 there were nine days with precipitation, while in 2003 precipitation was recorded on 18 days, resulting in an increased risk of Botrytis infections.

Leaf content of macro nutrients

The plants in the trial grew well and showed no deficiency symptoms, although the nitrogen status was suboptimal according to the recommended range. The recommendations are mainly based on investigations on cultivars used in the 1960s (Ljones, 1966b). Later investigations have indicated that the optimal range may be quite high for newer cultivars (Nes, 1978; Nes, 1990). Blackcurrant cultivars have considerably different N requirements (Ljones, 1966b), and it could be assumed that the present recommendations are imprecise for ‘Ben Tron’. Both nitrogen and potassium content were below recommended values at all fertilizer levels.

Yield and berry size have been negatively correlated with leaf N and Mg and positively with leaf K (Niskanen, 2001), but also a positive correlation between yield and leaf N content has been reported for some cultivars (Nes, 1990). In the present investigation there were no correlations between leaf nutrient content expressed as percent of dry matter, and any of the quality parameters recorded.

The phosphorus content in leaf dry matter was lower than reported by Niskanen (2002) and Kondakov (1993). In the present trial, leaf content was moderate or low for both nitrogen and phosphorus despite the triple increase in nitrogen and phosphorus fertilizer rate. The latest recommendations on phosphorus content in leaf dry matter for Scandinavian conditions have been developed in Finland, and are 3–8 mg kg⁻¹ (Aaltonen & Dalman, 1993). This is higher than in our experiment with Ben Tron.

Content of potassium in leaf dry matter was below recommendations until the last fertilizer increase in 2001. (Ljones, 1966b) calculated a negative NK interaction both in soil and in the ash content of the leaves. In our trial a significant correlation between leaf N and K was found (r = 0.743; p<0.05). The peaks in NPK content in leaves in 1995 and 2001 cannot be connected to precipitation, evapotranspiration or temperature.

Calcium and magnesium accumulation both varied significantly between years, but no connections with climatic factors were found. There was a positive correlation between calcium and magnesium accumulation in the leaves (r=0.669; p<0.01). Both calcium and magnesium were within the optimal range of 1.0–1.5 and 0.2–0.3%, respectively.

Ljones (1966b) also found that leaf nitrogen was not significantly affected by fertilizer treatment. The weak correlation between leaf macro nutrient content in leaf dry matter, yield and quality parameters, should stimulate development of other methods of recording nutrient status of the plants.

Soluble solids

Soluble solids content (SSC) in the fruits in relation to fertilization regime has recently been investigated in Ribes (Hoppula & Salo, 2005), where fertigation was found to be superior to broadcast fertilization. In apple, highest SSC is found when yields are low, regardless of nitrogen treatment (Drake et al., 2002). We found no significant correlations between yield and soluble solids in any single year, but low fertilizer rate gave low yield and high SSC in total for all years. The small variation in SSC between treatments within the same year corresponds to previous results in berries (Alleyne & Clark, 1997; Jeppson, 2000). High nitrogen fertilizer rate is known to delay maturity (Drake et al., 2002), and this may have an influence in the variation in SSC between treatments, as quality components differ in relation to fruit development and maturity (Heiberg, 1986a; Toldam-Andersen & Hansen, 1997; Lee & Kader, 2000). The SSC in our trial was significantly higher than reported by McCarthy and Stoker (1988), who found that preharvest irrigation reduced SSC compared to unirrigated treatments. In our trial, high precipitation in June led to low SSC. This correlation was not statistically significant, whilst precipitation in June was positively correlated to yield. Content of soluble solids was negatively correlated with leaf N, P and K, but none of the correlations was significant. Temperature in the growing season has been found to affect berry quality (Redalen, 1993), but in this trial no such correlations were revealed for any of the parameters recorded.

In conclusion, a triple increase in fertilizer level, two application methods and fertilization at different times had a limited effect on yield, leaf mineral content and fruit quality. Fertilization treatments generally affected the various parameters in the first years of the trial, but thereafter the effects became insignificant. Good water access seems to be essential for yield amount and berry size, while precipitation during flowering is likely to reduce yield because of fruit drop. Winter damage caused by relatively high temperatures in early spring followed by low temperature, gave a severe yield reduction in at least two years of the trial.