Introduction
Organic farming and retail sales of organic food products in North America have been characterized by strong annual growth rates over the past 20 yearsReference Reganold, Glover, Preston and Hinman1. In Canada, the market for certified organic potatoes (Solanum tuberosum L.) is growing rapidly with insufficient supply to meet domestic demand2. As production expands, there will be a need to evaluate the effectiveness of current management practices for organic potato production. Organic potato production is typically characterized by extended rotations involving leguminous crop green manures sometimes combined with organic amendments. Losses due to insects and pest damage and overall low quality, however, typically reduce organic potato yields to an average of 50–75% that of conventional productionReference Pimentel3–Reference Mattsson and Wallen5. Yield uncertainty in organic potato production may also promote post-harvest N losses through leachingReference De Neve, Dietjens, Moreels and Hofman6. A life cycle assessment of the potential environmental impact of organic potato production in Sweden suggested that environmental benefits would be increased through improving yields and tuber quality and reducing nitrous oxide emissions associated with organic potato productionReference Mattsson and Wallen5. Little is known, however, about the productivity and efficiency of nitrogen use in commercial organic potato production in Canada and how these production systems impact on potato tuber yield and quality.
Nitrogen is critically important for canopy development, tuber initiation and yield of potatoReference Porter and Sisson7, Reference Belanger, Walsh, Richards, Milburn and Ziadi8. The increasing availability of commercially produced organic amendments approved for use in organic production, such as composts and dehydrated pelletized manures, may change the intensity of N use in organic potato production systems. The efficient use of all such sources of N in potato production is required to minimize negative impacts on water and air quality and increase the economic return on the cropReference Klienhenz and Cardina9, Reference Belanger, Ziadi, Walsh, Richards and Milburn10 but few comparative studies exist on the impacts of contrasting organic amendments on potato tuber yield and quality. The slow release of N from organic amendments may reduce the risk of nitrate leaching early in the growing season. Moreover, studies in other systems have noted the ‘non-nutrient’ benefits of organic nutrient sources in improving soil structure and organic matterReference Foley and Cooperband11, Reference Carter, Sanderson and MacLeod12, water-holding capacity and root developmentReference Opena and Porter13 and soil microbial activityReference Lalande, Gagnon and Simard14. Evidence that soil organic amendments may be used to enhance potato yields has been presented as early as the 1970sReference Black and White15, when it was noted that manure increased potato yields and that the effect was probably due to improved cation-exchange capacity, organic matter content and water-holding capacity, rather than effects on nutrient supply. Little is known about the N use efficiency, soil mineral N dynamics and the environmental risks associated with organic potato production, and no research has been conducted on N cycling in organically managed potato soils in Atlantic Canada. The objective of this study was to evaluate processing potato yields and quality, and soil mineral N dynamics, under nonirrigated commercial organic production in Atlantic Canada, as affected by contrasting organic amendments.
Materials and Methods
A 2-year (2003–2004) field study was established on a commercial organic farm located in Winslow, Prince Edward Island, and on the Brookside organic agriculture experimental site at the Nova Scotia Agricultural College (NSAC), Truro, Nova Scotia. Soils at the Winslow site were a Charlottetown sandy loam (Orthic Humo-Ferric Podzol) and at the Brookside site were a Pugwash sandy loam (Orthic Humo-Ferric Podzol)Reference MacDougall, Veer and Wilson16, Reference Webb, Tompson, Beke and Nowland17. At the Winslow site, the potato plots in each year followed a preceding wheat crop, within the typical 4-year rotation sequence (2 years grass/clover perennial forage followed by grain and potatoes) maintained by the farm operator. Limited rotation options were available at the newly established Brookside site and preceding crops were long-term (~20 year) pasture in 2003, and grain in 2004. At the Brookside site, losses due to wireworm damage reduced plant stands by up to 30% in 2003, and the results for that year are not presented. Daily air temperature and precipitation was recorded at each site by an automated data logger with air temperature sensor and tipping bucket (Campbell Scientific Corp., Edmonton, Alberta, Cananda). The physical and chemical properties of the test soils are presented in Table 1.
Table 1. Initial physical and chemical properties of soils at the Winslow and Brookside sites in 2003 and 2004.
1 pH is measured in 0.01 M CaCl2 with a soil/solution ratio of 1:2.
2 Values are means (n=4)±standard deviation, of composite (1 per replicate) soil (0–30 cm) samples collected 10–14 days prior to planting at each site.
3 Particle-size distribution is measured by the hydrometer methodReference Sheldrick, Wang and Carter40.
4 Organic matter is determined by wet oxidationReference Tiessen, Moir and Carter41.
5 Soil P, K, Ca and Mg contents are measured by Mehlich III extractionReference Mehlich42.
The experiment used a randomized complete block design (RCBD) with 5 treatments and 4 replicates. Treatments included an un-amended control and two contrasting organic amendments approved for use in certified organic production: a commercial hog manure-sawdust compost (CP; Atlantic Country Composting, Antigonish, Nova Scotia, Canada) and a pelletized dehydrated poultry manure [Nutriwave 4-1-2 (NW); Envirem Technologies, Fredericton, New Brunswick, Canada], each applied at 300 and 600 kg total N ha−1 (CP300, CP600, NW300 and NW600). The compositions of the two organic amendments are provided in Table 2. We estimated that 30% of the total N applied as Nutriwave would be plant available in the year of application, i.e. 100 and 200 kg N ha−1 for treatments NW300 and NW600, respectively. A secondary objective of the research (data to be presented elsewhere) was to test a novel technique (in situ anion/cation-exchange membranes) for predicting plant available N (PAN) from soil and organic amendments. The latter treatment, NW600, was expected to provide available N in excess, but was included for this purpose. The compost, in contrast, because of the relatively high C/N ratio, was expected to supply much less PAN than the Nutriwave. Phosphorus (P) and potassium (K) were supplied as Sulpomag (0-0-22) and Calphos rock phosphate (0-3-0) as required by soil test. Application rates of Calphos were 2.3 t ha−1 at Brookside in 2004, and 3.5 and 5.8 t ha−1 at Winslow in 2003 and 2004, respectively. Sulpomag was applied at 0.490 t ha−1 to supply 108 kg K ha−1 at Brookside in 2004 and 0.743 t ha−1 to supply 165 kg K ha−1 at Winslow in 2003 and 2004. All amendments and supplemental P and K fertilizers were broadcast and incorporated by disking 1 day prior to planting by hand. Each plot consisted of 10 m of 4 rows of potatoes with 0.91 m between-row and 0.3 m within-row spacing. Trials were planted with hand-cut generation E1 seed of cultivar Shepody (McCain Produce Inc., Florenceville, New Brunswick, Canada). Regionally, optimum inorganic fertilizer N application rates for this cultivar following small grains are in the order of 200 kg N ha−1.Reference Porter and Sisson7, Reference Belanger, Walsh, Richards, Milburn and Ziadi18
Table 2. Composition of organic amendments used in 2003 and 2004.
1 With the exception of DM, all analyses are expressed on a dry matter basis.
2 Total C and N content determined by dry combustion (LECO CNS-1000)Reference Lynch, Voroney and Warman43.
3 Mineral N content of water extracts determined by a segmented flow autoanalyzerReference Lynch, Voroney and Warman43.
4 Total P and K content determined by inductively coupled plasma (ICP) following nitric acid digestion of dried samplesReference Lynch, Voroney and Warman43.
5 Values are means (n=4)±standard deviation.
Potatoes were planted on 9 June 2003 and 2 June 2004 at the Winslow site, and on 27 May 2004 at the Brookside site. Hills were re-shaped mechanically at approximately 30 days after planting (DAP). Weeds were controlled until hilling by hand weeding. Late blight and Colorado potato beetle were controlled by using copper hydroxide (Parasol) and Bt (Novador), respectively, as required. Total copper applied as Parasol at each site ranged between 2.8 and 5.6 kg Cu ha−1. Vines were killed by mowing and plots were machine harvested 2 weeks later using a one-row digger. Biomass and N uptake of shoot, root and tuber were determined just prior to topkilling through recovery of 4 plants per plotReference Zebarth and Milburn19. Potatoes were harvested on 7 October 2003 and 22 September 2004 at the Winslow site and on 15 September 2004 at the Brookside site. Fresh weight yields were measured from 8 m of row taken from the center two rows in each plot. A 25 kg sample of tubers was retained and stored at 9°C until graded for size, defects and then sub-sampled for determination of dry matter and specific gravity. Tubers were sized in accordance with the Canadian Agricultural Standard Act for processing potatoes; small is 38–51 mm, Canada no. 1 size is 51–89 mm, large (‘Jumbo’) size is 89–114 mm and culls are <38 mm20.
The dynamics of soil nitrate (NO3−-N) in the 0–30 cm depth layer were gauged from soil samples (a composite of 10×2.5 cm diameter cores per replicate in spring, or per plot thereafter) collected repeatedly (i.e. approximately every 3 weeks) until topkilling and following tuber harvest. Samples were immediately frozen until analyzed. Samples were extracted with 2 M KClReference Zebarth and Milburn19. The NO3− and NH4+ concentrations of the extracts were determined by molecular absorption spectroscopy in a segmented flow system (Technicon Industrial Systems Corp., Tarrytown, New York). We assumed a soil bulk density of 1.0 mg m−3 to convert inorganic N concentrations measured at planting and post-harvest to units of kg N ha−1.Reference Zebarth, Leclerc, Moreau, Gareau and Milburn21 The apparent N recovery (ANR) was determined as the difference in whole plant N uptake between amended and un-amended plots expressed as a percentage of total N applied, i.e.
PAN from organic amendments was calculated as ANR, plus gain of soil residual mineral N at harvest relative to the control expressed as a percentage of total N applied, i.e. PAN (%)=[((N uptake for treatment−N uptake for unfertilized control)+(seasonal change in soil N mineral N for treatment))/N applied]×100.
Normality distribution of data was checked and logarithmic transformations were made when necessary. Because of partial changes in the composition of both organic amendments between years, analysis of variance (ANOVA) was performed separately for each site-year using the GLM procedure22. The basic difficulty here was that there were only 2 (source)×2 (rate)+1 (control) unique treatments and these did not comprise a complete 2×3 factorial design. A correct analysis of such experiments was previously demonstratedReference Addelman23, Reference Gates24. A 5×1 RCBD model was set to separate treatment means at first, followed by a 2×2 factorial RCBD model excluding the control to further explore the effects of source and rate of organic amendments. Two orthogonal contrasts involving control were created to differentiate the control and others. Next, a univariate repeated-measure ANOVA, after sphericity test, was carried out for the soil nitrate variable by integrating time effect in the model as an independent variable to assess the interaction of sampling time with treatment.
Results and Discussion
Seasonal temperature and rainfall
In 2003, 328 mm of rain fell at Winslow during the 5-month period of June to October, approximately 10% less than the 30-year average of 361 mm distributed well across all 5 months. Average monthly temperatures, with the exception of September (+3°C), closely followed (±1°C) the 30-year average for this location. In 2004, at the same site, 340 mm of precipitation fell during the same period. Temperatures were slightly cooler than average in June (13.7°C compared with 15.3°C) but followed closely the 30-year average for the remainder of the season. Annual temperatures at Winslow in 2003 and 2004 averaged 5.7 and 5.3°C, respectively, whereas total precipitation (snow plus rainfall) totaled 1024.2 and 986.8 mm, respectively. At Brookside, 294 mm of rain fell from June to October 2004, well below (−24%) the 30-year average of 385 mm. Much of this deficit (64 mm) occurred during the first 2 months (June and July) of the season. Monthly average temperatures during the season compared closely (±1°C) with 30-year averages for this site, with the exception of the month of June, which was notably cooler (12.7°C compared with a 30-year average of 15.1°C). The annual temperature at Brookside in 2004 averaged 5.6°C, whereas total precipitation totaled 1247.1 mm.
Tuber yield
At the Brookside site, tuber yields (21.5 Mg ha−1, on average across treatments) were 30% lower than the average yield recorded at the Winslow site in either year. The lower yields at the Brookside site are most likely attributable to the early season moisture deficit in 2004. In MaineReference Porter, Opena, Bradbury, McBurnie and Sisson25 a 36–37% increase in yield (cv. Superior) was obtained through supplemental irrigation over 2 years when rainfall amounts fell to 323 and 217 mm from June to October. In an Atlantic Canada studyReference Belanger, Walsh, Richards, Milburn and Ziadi18, the response of Shepody to applied N was also greater with irrigation than without, and was reflected in both increased number of tubers per plant and average fresh tuber weight. In site years (Winslow 2003 and 2004) when seasonal soil moisture levels were non-limiting, the relatively high tuber yields (29.9±2.2 Mg ha−1) for the un-amended control plots were reflective of the long rotations (4-year+) characteristic of organic potato production. In station-based trialsReference Zebarth, Leclerc, Moreau, Sanderson, Arsenault, Botha and Wang-Pruski26 average tuber yields of 34.5 Mg ha−1 were reported for Shepody fertilized with P and K but not N.
The effects of organic amendments on tuber yields varied over the three site-years (Table 3). The hog manure–sawdust compost significantly decreased total tuber and marketable yields in 2003 at the Winslow site (Table 4), presumably attributable to a strong net immobilization of soil N promoted by the compost characterized by a high C/N ratio of 21.6 (Table 2). Solid manures with C/N above 15:1 have been reported to decrease N availability in the short termReference Qian and Schoenau27, Reference Beauchamp, Paul, Hansen and Henriken28, while negligible net N mineralization has been found for composts with C/N of greater than 22:1Reference Janssen29. In 2004 at Brookside, total tuber yields in the compost plots were significantly higher than the un-amended control (Tables 3 and 4) but, as discussed below, the increases in tuber yields were not accompanied by increases in tuber or plant N content at topkill (Table 5).
Table 3. Significant levelsFootnote 1 of ANOVA for planned variables over all three site-years.
1 *, ** and ***, significant at P<0.5, 0.01 and 0.001, respectively; ns, not significant at P<0.05; nd, not determined.
2 CT, un-amended control; CP, hog manure–sawdust compost; NW, pelletized poultry manure ‘Nutriwave’.
Table 4. Potato tuber yield, marketable yield, tuber size, mean tuber weight and specific gravity over the three site-years.
1 CT, un-amended control; CP300 and CP600, hog manure–sawdust compost applied, respectively, at 300 and 600 kg total N ha−1; NW300 and NW600, pelletized poultry manure ‘Nutriwave’ applied, respectively, at 300 and 600 kg total N ha−1; SEM, standard error of the mean.
2 Jumbo, 89–114 mm; Canada no. 1, 51–89 mm.
3 nd, not determined due to wireworm reduction in marketable yield.
Summary of statistical analyses of data is provided in Table 3.
Table 5. Tuber N concentration, tuber and whole plant N accumulations, ANR and PAN measured at topkill over the three site-years.
1 CT, un-amended control; CP300 and CP600, hog manure–sawdust compost applied, respectively, at 300 and 600 kg total N ha−1; NW300 and NW600, pelletized poultry manure ‘Nutriwave’ applied, respectively, at 300 and 600 kg total N ha−1.
Summary of statistical analyses of data is provided in Table 3.
Application of the pelletized poultry manure at the rate of 300 kg total N ha−1 (NW300) significantly and consistently increased total tuber and marketable yields at the Winslow site, with average increases over 2 years of 5.8 and 7.0 Mg ha−1 in tuber total and marketable yields compared with the un-amended control (Tables 3 and 4). However, the yield response was not observed at the higher rate application (600 kg N ha−1) of Nutriwave (NW600). While tuber number differed little between Nutriwave treatments (187 (NW300) versus 203 (NW600); data not shown), NW600 resulted in mean tuber weights much smaller than the unfertilized control probably due to delayed plant maturity and tuber fill as a result of excessive N supply (Table 4)Reference Harris and Harris30.
Reductions in marketable yields from total tubers at the Winslow site varied substantially between years, with losses more apparent in 2004 than 2003 (Table 4). These were primarily due to scab (average ~30% of tubers) with culls and other defects accounting for less than 5% of tubers. At 7.0, soil pH of un-amended soil at this site in 2004 was notably high (Table 1), a factor which undoubtedly contributed to the greater incidence of scab recorded. At the Brookside site in 2004, most tubers, irrespective of treatment, showed some incidence of wireworm damage and this was reflected in greatly reduced marketable yields (Table 4). Nutriwave, but not compost, significantly decreased tuber specific gravity in all three site-years. The tuber size distribution was variable in the compost plots, with Canada no. 1 and Jumbo size tubers significantly lower at the Winslow site in 2003 but greater at Brookside in 2004, as compared with the un-amended control (Table 4).
With conservative estimated gross wholesale returns of can$400 per Mg, a 20% yield gain (in the order of 5.0 Mg ha−1) in response to compost and the Nutriwave amendments would produce added returns of approximately can$2,000 ha−1 before culling losses. At the application rates used in this study and costs in the range of can$100 Mg−1 for the Nutriwave and can$30 Mg−1 for compost, amendment costs would be in the range of can$700 ha−1. As noted below, however, we underestimated PAN from Nutriwave by up to 9%. Given current premiums for certified organic potatoes, enhancing yields through application of organic amendments which supply N or organic matter appears warranted. It can be argued that such imports constitute a ‘substitution’ for synthetic sources by organic systemsReference McRae, Hill, Mehuys and Henning31. A central principle of organic systems is a commitment to regional recycling of nutrients, however, and given the benefits of organic amendments to soil physical, chemical and biological properties, the use of such materials can be considered consistent with the overall objectives of production system ‘redesign’ espoused within certified organic industry standards, as noted elsewhereReference Gallandt, Mallory, Alford, Drummond, Groden, Liebman, Marra, McBurnie and Porter32. Organic production systems are criticized for potentially mining reserves of phosphorus and potassiumReference Gosling and Shephard33, and an over-reliance on imported phosphate rock as the primary soil P amendmentReference Kirchmann and Thorvaldsson34. Consideration of the use of regionally available organic amendments in organic production systems must also be made within that context.
Crop N uptake and ANR
Crop N uptake measured at topkilling in un-amended soils averaged 112 kg N ha−1 (Table 5). This is significantly higher than an average of 85 kg N ha−1 uptake by potatoes receiving no N fertilizer on 56 farm sites in Atlantic Canada, and slightly greater than the 105 kg N ha−1 uptake reported specifically for Shepody following red cloverReference Zebarth, Leclerc, Moreau, Sanderson, Arsenault, Botha and Wang-Pruski26. In un-amended control plots, tuber N represented 62% of total plant N uptake, well within the range of 53–78% N harvest index reported for Shepody fertilized, in station-based trials, with P and K but not NReference Zebarth, Leclerc, Moreau, Sanderson, Arsenault, Botha and Wang-Pruski26.
Although the tuber N accumulation and total plant N uptake were complicated by a source by rate interaction in the Winslow site in 2003 (Table 3), the tuber N and total plant N uptake in the pelletized poultry manure treatments were generally considerably higher than the un-amended control and the compost plots in all three site-years, except for tuber N at Brookside in 2004 where the effects were depressed by yield suppression caused by a moisture deficit (Table 4). The Nutriwave effects were also reflected in elevated tuber N concentrations (Tables 3 and 5), most evident under the high application rate (NW600). Tuber nitrate has been found to increase substantially for Shepody at tuber N contents greater than 1.6%Reference Belanger, Walsh, Richards, Milburn and Ziadi18 although the relationship between them appears very field- and year-specificReference Zebarth, Chabot, Coulombe, Simard, Douheret and Tremblay35. In Germany, increased tuber nitrate in organically managed potatoes was obtained when farmyard manure was supplied at 160 and 240 kg N ha−1.Reference Neuhoff and Kopke36
Typically, where mineral fertilizer N is applied to potatoes, approximately 45% of this N is recovered in potato tubersReference Greenwood, Neeteson and Draycott37, Reference Tran and Giroux38. With increasing rate of the Nutriwave, 52 and 49% of plant N, respectively, was recovered in tubers (Table 5). Over all site-years, the ANR of amendments averaged from 1.5 to −2.3% for CP300 and CP600 and 26 to 17% for NW300 and NW600, respectively (Table 5). The negligible N benefit of the compost treatments even when promoting a yield response at Brookside in 2004 was notable, suggesting that the compost acted as soil amendment rather than as a nutrient source. A growing body of reportsReference Foley and Cooperband11–Reference Lalande, Gagnon and Simard14 suggests that for composts in particular, the ‘non-nutrient’ benefits to soil physical, chemical and biological properties, such as improvements in soil structure and organic matter, water-holding capacity and microbial activities, are primarily responsible for observed increases in potato yields. Potatoes are a shallow rooted crop, and even small improvements in soil moisture holding capacity can significantly improve yields, as shown when composts were applied in potato rotationsReference Foley and Cooperband11, Reference Carter, Sanderson and MacLeod12, Reference Lalande, Gagnon and Simard14. In a study in MaineReference Opena and Porter13 the benefits of compost applied at 45 Mg ha−1 fresh weight on root length density and tuber yields were attributed at least partially to improved soil aggregation. In the present study, compost significantly increased gravimetric soil moisture at 0–30 cm by up to 3% depending on sampling date (data not shown).
The PAN recorded for the Nutriwave treatments ranged from 33 to 39% (on average across all site-years), somewhat greater than our initial estimate of 30% for this product at pre-planting. An earlier studyReference Zebarth, Chabot, Coulombe, Simard, Douheret and Tremblay35 obtained ANR rates of 26–60% for a pelletized organo-mineral fertilizer product which combined poultry, dairy and hog manure and was of higher N content (7%). Assuming minimal in-season N losses, treatments NW300 and NW600 supplied an estimated 116 and 194 kg PAN ha−1, respectively. As noted above, although small grains were the preceding crop, yield and tuber size responded to the lower of these N input rates alone. In a previous study, Shepody yield and tuber size were found to rarely respond to fertilizer N supplied at greater than 135 kg N ha−1 when preceded by red cloverReference Porter and Sisson7. An N credit in the order of 75 kg N ha−1 was recommended where red clover was the preceding crop. These results at the Winslow site, given that a small grain was the preceding crop and that inorganic N levels recorded at planting were not unusually high, suggest that a relatively large mineralizable soil N pool accrues under the extended rotations practiced at this commercial organic farm. Supplementing a clover rotation crop with dairy manure compost at the much lower rate of 6.5 Mg ha−1 (approx. 62 kg N ha−1) was found in one study to significantly increase marketable yields of organic potatoesReference Klienhenz and Cardina9. As the site had been previously managed organically for 4 years, the authors argue that even in post-transitional organic systems, organic amendments are likely required for high-N-demanding crops to supplement N supplied from leguminous green manures.
Seasonal change in soil nitrate
There were treatment-by-measuring date interactions on soil NO3−-N at all three site-years, indicating distinctive seasonal changes in soil NO3−-N amongst the treatments. Soil NO3−-N contents increased at early growing season until hilling (30–35 DAP) and decreased thereafter. The seasonal fluctuation was more evident in the Nutriwave plots than in the other plots, especially at the Brookside site (Fig. 1). Soil NO3−-N remained considerably higher in the pelletized poultry manure plots than in the other plots throughout the growing season (Fig. 1). The effects of organic amendments on soil NO3−-N were also complicated by an interaction of source by source. The treatment NW600 presented a higher soil NO3−-N than the treatment NW300, while the difference was diminished between the two compost treatments. A gradual late season (August to September) decline in soil NO3−-N levels for the treatment NW600 was observed at Winslow and Brookside in 2004, but less so at Winslow in 2003 (Figs 1 and 2). This may be attributed to the later planting, topkilling and harvest date at Winslow in 2003 compared with 2004, combined with the slightly reduced rainfall (71 mm) received at this site during the month of September when compared with that received (82 and 81 mm) in September 2004 at Winslow and Brookside, respectively. A lower soil NO3−-N in the compost plots than in the un-amended control observed during most of the growing season either at the Winslow or at the Brookside site, once again, suggested net immobilization of soil N in the plots receiving high C/N compost. In un-amended plots, soil mineral N (NO3−-N+NH4+-N) levels to 30 cm depth at pre-planting ranged from 35 to 75 kg N ha−1 (Table 1) in all three site-years, within the range reported regionally for potatoes at pre-planting following small grain cereals. In a survey of 228 conventionally managed potato fields in New Brunswick, Zebarth et al.Reference Zebarth, Leclerc, Moreau, Gareau and Milburn21 found that inorganic N levels to 30 cm ranged from 17 to 69 kg N ha−1 where a cereal crop preceded potatoes, and 7 to 107 kg N ha−1 with red clover or grassland as the preceding crop.
Figure 1. Seasonal changes in soil (0–30 cm) NO3−-N as affected by organic amendments over the three site-years (CT, un-amended control; CP300 and CP600, hog manure–sawdust compost applied, respectively, at 300 and 600 kg total N ha−1; NW300 and NW600, pelletized poultry manure ‘Nutriwave’ applied, respectively, at 300 and 600 kg total N ha−1).
Soil residual NO3-N to 30 cm depth after harvest was low for un-amended plots and compost plots (Fig. 2). Residual soil nitrate for these treatments, averaging less than 25 kg N ha−1, was very low when compared with regional averages of approximately 60 kg N ha−1 soil nitrate at 0–30 cm depth following potatoes which received inorganic fertilizerReference Zebarth, Leclerc, Moreau, Gareau and Milburn21. Other studies also found residual soil nitrate to be low, in the range of 14–19 kg N ha−1, for Shepody following barley or red clover but receiving no fertilizer NReference Zebarth, Leclerc, Moreau, Sanderson, Arsenault, Botha and Wang-Pruski26. In contrast, the average residual soil nitrate for NW300 and NW600 was 61 and 141 kg N ha−1, respectively, when averaged over all three site-years. While it would have been preferable to have data on residual NO3-N to 60 cm depth, in previous Atlantic Canada studies, residual soil NO3-N at 0–30 cm after fertilized potato harvest was highly related (R 2=0.94) to, and accounted for an average of 60%, of NO3-N in the entire soil profile (0–90 cm)Reference Belanger, Ziadi, Walsh, Richards and Milburn10. Very little leaching of nitrate to below the root zone appears to occur during the potato growing season, with most leaching losses occurring between seasons, although these losses can be highly variable between yearsReference Zebarth, Leclerc, Moreau, Gareau and Milburn21. Residual soil (0–90 cm) NO3-N levels of up to 70 kg N ha−1 were considered a ‘reasonable level’ achievable only when N fertilization is based on an economically optimum N application rate. Levels of inorganic N remaining in soil as high as 140–250 kg N ha−1 derived from fertilizer N unharvested in the potato crop have been recorded in E. Canada, but with 50% of this residual N lost through nitrate leaching during the winter–early spring periodReference Gasser, Laverdiere, Lagace and Caron39.
Figure 2. Soil (0–30 cm) residual NO3−-N at harvest as affected by organic amendments over the three site-years (CT, un-amended control; CP300 and CP600, hog manure–sawdust compost applied, respectively, at 300 and 600 kg total N ha−1; NW300 and NW600, pelletized poultry manure ‘Nutriwave’ applied, respectively, at 300 and 600 kg total N ha−1).
Conclusions
In an extended rotation characteristic of production practices on commercial organic potato farms in Atlantic Canada, relatively high tuber yields (~30 Mg ha−1) were achieved without supplemental N application, in 2 of 3 years when seasonal moisture levels were non-limiting. Crop N uptake in un-amended soils averaged 112 kg N ha−1. At the commercial farm site, significant reductions in marketable yield in the second year were due mainly to scab, attributable to high soil pH. In un-amended and compost-amended soils, residual nitrate N at harvest was low (<25 kg N ha−1) when compared with regional averages where inorganic fertilizer is applied. These results suggest that these systems are conservative with respect to N and minimize risks of overwinter soil nitrate losses. When soil moisture levels were adequate, application of a pelletized poultry manure product promoted gains in yields (+5.8 Mg ha−1 average) when applied to supply an estimated 112 kg PAN ha−1 (NW300), but residual soil nitrate levels rose to 61 kg N ha−1. Similar benefits achieved in 1 of 3 years for compost applied at up to 48 mg DM ha−1 (CP600) appeared due to ‘non-nitrogen’ benefits of compost. Yield effectiveness of NW decreased at higher rates of application (NW600). Most leaching losses of NO3−-N occur between seasons, and excessive levels of residual soil NO3−-N at harvest, as obtained for NW600 at 141 kg NO3−-N ha−1, must be avoided. Given current premiums for certified organic potatoes, enhancing yields and marketable yields through application of amendments which supply moderate rates of N or organic matter appears warranted.
Acknowledgements
The project was financially supported by the Prince Edward Island Department of Agriculture and Forestry, the New Brunswick Department of Agriculture, Fisheries and Aquaculture, the Organic Agriculture Centre of Canada at NSAC, and the Canada Research Chairs program. Thanks are extended to F. Dollar, owner of Kentdale Farms, Winslow, PEI, for his participation in this study, and to K. MacNeil, A. Runnels, H. Purves, B. van de Pol, S. Urbaniek and A. Hammermeister for their excellent technical assistance.
