Identifying key factors to improve productivity and reduce environmental impact of potato farms in West Java, Indonesia

. Potato ( Solanum tuberosum L.) production in Indonesia has increased significantly in recent years, with a total annual production of 1.3 million tons. Although the acreage of potatoes is increasing relatively, the productivity of potatoes remains stagnant and is relatively low, around 16-17 tons per ha. A series of studies with the objective to identify key factors associated with potato cultivation and to evaluate of existing farmers’ practices were carried out in the main potato-producing areas i.e. Pangalengan and Garut, West Java during five consecutive production seasons starting from the dry season of 2013 to the dry season of 2015. In each sub-district, ten potato farmers growing cv. Granola and/or Atlantic were selected resulting in 40 participating farmers. A substantial variation of the observed yields was found for both varieties and regions, mainly due to seed quality. The supply of nitrogen (N) and phosphate (P 2 O 5 ) from organic and chemical fertilizers of 400 kg N and 581 kg P 2 O 5 /ha outreached results of local experiments in obtaining optimal yields. Farmers applied much more active ingredient (AI) per spray, up to 6 times and used more than 500 l of water per spray to control late blight, while experiments showed the spray volume can be reduced by 10-20% without loss of efficacy. The key factors identified in this study to stimulate potato development and productivity and reduce environmental impact in the areas were seed potato quality, fertilizer use efficiency and late blight control.


Introduction
The potato (Solanum tuberosum L.), the fourth-most important food commodity in the world after rice, wheat, and maize, is increasingly important in addressing problems of global food production and climate change [1,2].Because of its widely diverse distribution pattern and present cultivation and demand, the potato crop is one of few food sources that can assist meet all of these limits globally, particularly in emerging nations with high levels of poverty, hunger, and malnutrition [3].The interest in growing potatoes increases in tropical areas such as Indonesia due to its economic and high nutritive values [4].Over the last few decades, Indonesia's potato production has exploded, and the country has risen to become Southeast Asia's leading producer with an annual area of potatoes of 60,000 to 70,000 ha with a total production of more than 1.3 million tons per year [5].
Potatoes in Indonesia are mainly grown in the highland areas at altitudes of more than 1,000 m above sea level where the temperature is low.The main potato variety in Indonesia since the 1980s is Granola, a table potato that covers 80 to 85% of the area.However, there is an increasing demand for processed potato products like chips and also French fries, especially in urban areas of Indonesia [6] as well as Asia and the Pacific [7].Therefore, a processing potato variety, Atlantic becomes more popular to grow.
Even though the area of growing potatoes in Indonesia is still increasing, the productivity of potatoes remains stagnant, with declining yields as a result.Many factors were considered as the cause of this decline, such as: 1) a very short crop rotation combined with poor management in the highlands leading to a build-up of diseases, depletion of nutrients and soil exhaustion, and subsequently reduced profitability and increased impacts on the environment; 2) an excessive use of agricultural inputs with a negative impact on the environment contributing to acidification, eutrophication and global warming [8]; and 3) the use of poor-quality seed potatoes.
To mitigate land degradation, reduce negative impacts on the environment and to guarantee the sustainability of farms in the main potato areas in the highland tropics of Indonesia, the VegImpact Project (Wageningen University and Research, the Netherlands) initiated a project with a series of studies in two sub-districts (regions) i.e., Pangalengan and Garut, West Java.The objective of the project was to understand farmer production practices, identify key factors for yield improvement, for increased input use efficiencies and for reduced environmental impact.The results of the project were to develop a training program for potato farmers in the area to meet the increasing consumers demand.

Materials and methods
The study was carried out in Pangalengan sub-district, Bandung district and Cikajang Sub-district, Garut district, being the main potato production areas of West Java.Two main potato varieties i.e., Granola and Atlantic were included.Granola was released in Germany in 1975, introduced into Southeast Asia in the early 1980s, and the processing variety Atlantic was bred and released in the United States of America in 1976.This variety has good characteristics for crisps and is currently the main processing potato.Ten potato farmers in each sub-district (region) growing Granola or Atlantic were selected yielding 40 participating farmers.
A format for farm records of all farm activities associated with the potato production was provided to each farmer to be filled in on a daily basis during each of the five production seasons.The area planted as estimated by farmers and by GPS, quality and quantity of seed potatoes used, nutrients used (organic and inorganic), pesticides used and costs of all inputs and labour were the most data analysed.At harvest, marketable yields were measured by weighing each harvested bag of potatoes.Farmers were intensively supported with the registration as they had little to no experience with it.
Potentially, this intensive monitoring yielded 200 farm records, 20 farmers in two sub-districts, two potato varieties during five production seasons, e.g., planting in the dry season 2013, rainy season 2013/2014, dry season 2014, rainy season 2014/2015, and dry season 2015.However, as seeds of Atlantic were not always available, some farmers did not plant potatoes where some farmers planted the Granola instead, 180 farm records were collected.

Results and discussion
Atlantic was not produced as intended as seeds were unavailable and farmers shifted to Granola.Farmers in general obtained Atlantic seeds through private companies who imported them from Canada, New Zealand and/or Australia.Most often they did not arrive in time for planting.Additionally, some farmers who received seeds in time complained about the poor quality not worth to plant, probably due to improper conditions during overseas transport and or (local) storage.Granola seeds, on the other hand, were in general obtained from the previous planting season or bought from specialized seed potato producers providing locally produced certified seeds.
The average area planted with potatoes was 2,844 m 2 per farmer.In Garut, the area planted with Atlantic was larger (5,113 m 2 ) than that of Granola (1,839 m 2 ), whereas in Pangalengan the average area with planted Atlantic was smaller (2,306 m 2 ) than that of Granola (2,551 m 2 ).The smaller areas planted with Atlantic in Pangalengan compared with those in Garut was mainly due to the poor quality of Atlantic seeds received by farmers in Pangalengan.Farmers in Garut had easier access to Atlantic seeds than with larger planted areas with Atlantic as a result.To a lesser extent, the same results were found for areas planted with Granola in Pangalengan compared to Garut i.e. 2,551 m 2 and 1,839 m 2 , respectively.The better seed quality of Granola in Pangalengan compared to Garut most likely contributed to the difference.
In Garut, the average yields of Atlantic was higher than that of Granola in all seasons recorded except for dry season 2015, where yields of Granola were high compared to all other yields achieved (Figure 1).In Pangalengan, the average yields of Granola were in general higher than those of Atlantic, except for the dry season in 2014 where yields of Atlantic were slightly higher than those of Granola.The small number of farmers growing Atlantic and the large number of farmers growing Granola, only 2 and 19 respectively, may have contribute to this result.On average, the yields of Atlantic in Garut were 21, 18, 22 and 19 tons/ha, whereas the average yields of Granola were 18, 17, 19 and almost 24 tons/ha in the subsequent growing seasons, respectively.In Pangalengan, the average yields of Atlantic were 13, 12, 18, 14 and 18 tons/ha whereas the average yields of Granola were 18, 23, 18, 23 and 20 tons/ha in the subsequent growing seasons respectively.The higher yields of Atlantic grown in Garut compared to those in Pangalengan was associated with the better quality of Atlantic seeds.Meanwhile, the lower yields of Granola grown in Garut compared to those in Pangalengan was also associated with the lower quality of Granola seeds as the source of certified and or high quality Granola seeds was very limited in the area.Averaged between seasons and regions, the yields of Atlantic were lower than those of Granola (Figure 1).Variation in yields between farmers either in Garut or Pangalengan was probably associated with the lower yields of Atlantic than those Granola.Yield variation may also be caused by the generation of potato seeds used by the farmers.In general, the older the generation, the lower the yields obtained.
Farmers either in Garut or in Pangalengan in general used chicken manure which is easily obtained in both sub-districts.The nutrients were applied by farmers based on the nutrient content of nitrogen (N), phosphate (P2O5) and potassium (K2O) in the chicken manure which was 1, 1.5 and 1 kg per ton, respectively.The manure applications did not differ between seasons, regions or varieties (data not shown).All plantings were fertilized with 22 tonnes organic manure per ha thus providing 220 kg N/ha, 330 kg P2O5/ha and 220 kg K2O/ha.About half of the N becomes available for the potato crop during the growing season [9].
In total, about 32 different chemical fertilizer products were applied of which the majority were compound NPK products.In contrast to the manure applications, differences were found for chemical fertilizer applications between regions (Figure 2 left).The N applied from chemical fertilizers in Garut was higher than that in Pangalengan.The mean N-application from chemical fertilizers was 180 kg N/ha, therefore the total N applied from manure and chemical fertilizers was as high as 400 kg N/ha where N fertilizer experiments in Garut and Pangalengan indicated that an optimal N-application from chemical fertilizers between 107 to 170 kg N/ha plus 15 tons/ha of chicken manure would maximize Granola yields [9].The N application in Atlantic in Pangalengan was lower than that in Granola, whereas the N application in Granola in Garut was higher than that in Pangalengan.As there are no fertilizer recommendations for potato in both regions, farmers applied organic and chemical fertilizers according to their previous experiences or their parents who cultivated potatoes in the past.
In terms of P2O5 applications, an average of almost 400 kg/ha was applied in Atlantic in Garut (Figure 2 right).The mean application of P2O5 from chemical fertilizer was 249 kg/ha.Differences in P2O5 application were found between regions, the P2O5 application in Pangalengan in Atlantic was lower than in Granola, while the P2O5application in Granola in Garut was higher than that in Pangalengan.The fertilizer experiments carried out in 2014 and 2015 in Pangalengan and Garut indicated that the optimal P2O5 application from chemical fertilizers for maximum yield of Granola in Pangalengan and Garut was between 114 to 165 kg P2O5 and 104 to 126 kg P2O5/ha plus 10 ton/ ha chicken manure, respectively.Although the majority of the P2O5 in potato production is applied with organic manure, application of small amounts of P2O5 through chemical fertilizers is essential for high potato yields in Pangalengan and Garut [9].
Potatoes grown in the main potato production areas such as in Pangalengan and Garut suffer greatly from late blight (Phytophthora infestans), which is the most important disease in potatoes and other Solanaceae such as tomato in Indonesia [10].The disease is especially severe in the highlands where the wet conditions and high humidity are favourable for the development of the disease.Furthermore, host availability is high in these regions as potato and tomato are cultivated the whole year round, and the two main potato varieties i.e.Granola and Atlantic are susceptible to this fungal disease.The disease can cause significant yield losses ranging from 10-100%.Many case studies in developing countries found that the most pesticides in potato are used to control late blight [11].Thus, potatoes are commonly heavily sprayed both in number of applications as well as doses active ingredients (AI) applied, as many as 20-30 times during a single potato cultivation season [12].The applications of fungicides to control late blight in both regions i.e.Pangalengan and Garut were in agreement with these findings (Figure 3).In general, the number of fungicide applications either with curative or preventive or both fungicides per farmer in Atlantic were higher than in Granola either in Garut or in Pangalengan.The number of fungicide applications in Garut with either curative or preventive fungicides in Atlantic and Granola were on average 20.5 and 16.9 applications per farmer, respectively, whereas in Pangalengan the number of applications of fungicides with either curative or preventive fungicides in Atlantic and Granola were on average 15.3 and 14.4 applications per farmer, respectively.Both Atlantic and Granola are susceptible to late blight, but Atlantic is somewhat more susceptible than Granola and therefore requires a more robust control strategy under high disease pressure [13,14].However, the most stricking from Figure 3 is that late blight conditions in the dry season are less favourable than in the rainy season and such a difference is not found in the dry season 2013 compared to the rainy season 2013, but afterwards a decrease of applications in general and particularly in dry seasons is found.Finally, in the rainy season 2014 the average number of applications stays just under the 20 to 30 sprays which are considered as standard.
In Garut, the use of preventive and curative fungicides in Atlantic and Granola was on average 64.4 and 1.3 kg AI/ha, respectively, whereas in Pangalengan, the use of preventive and curative fungicides in Atlantic and Granola was on average 33.3 and 5.4 kg AI/ha, respectively (Table 1).Also, more AI was used on Atlantic than on Granola in both regions.The knowledge of using the preventive and/or curative products to control late blight in potatoes should be mastered by farmers in the areas in order to have an effective and efficient control of late blight.An additional way to evaluate the use of fungicides is to estimate the Treatment Frequency Index (TFI).This evaluates if farmers used the required dose according to Government Guidelines of the product used.It is calculated as the relative dose compared to the recommended dose.For each fungicide spray the TFI was calculated and averaged per farmer.Farmers use on average more kg AI/ha as recommended (TFI > 1), up to more than 6 times (Table 2).There are two factors that contribute to this.First, farmers use tank mixes of different products that have the same control strategies, most often preventive agrochemicals.Product names differ but the AI do not which results in a substantial overuse of AI.The second factor is the spray volume, that is included in the recommendations.Farmers start spraying approximately 20 days after planting and use small volumes.However, the volume increases when biomass increases, and exceeds recommendations.Farmers commonly use a knapsack or a motor sprayer.The latter has an electric pump which pumps out the pesticides onto the crop at high pressure.Both methods use high volumes of water [15].In general, farmers used more than the recommended 400 to 600 l of water per ha.The potato crops were sprayed at a high volume until droplets start to form and run-off.The solution that runs off the leaves was not effective in protecting the plants against the disease.Experiments in Garut and Pangalengan indicated that a 20% reduction in spray volume in both Atlantic and Granola tended to result in a decreased efficacy of late blight control [13].However, the small difference in efficacy did not negatively influence potato yields.This finding was also in agreement with the experiment of the influence of spray volume and adjuvants on the deposition of mancozeb carried out by Schepers et al. [16].Without rain simulation there was a clear trend that spraying with 750 l per ha compared to 250 l per ha resulted in a lower mancozeb residue on the leaves, most probably caused by run-off.

Conclusions and recommendations
There was a substantial variation in yields between varieties in both regions, where the quality of the seed potatoes was one of the most dominant factors to affected yields.The supply of nitrogen (N) and phosphate (P2O5) from organic and chemical fertilizers of 400 kg N and 581 kg P2O5/ha outreached results of local experiments in obtaining optimal yields.Farmers applied much more active ingredient (AI) per spray, up to 6 times and used more than 500 l of water per spray to control late blight, while experiments showed the spray volume can be reduced by 10-20% without loss of efficacy.The key factors identified in this study to stimulate potato development and productivity and reduce environmental impact in the areas were seed potato quality, fertilizer use efficiency and late blight control.

Fig. 1 .
Fig. 1.The average yields of Atlantic and Granola grown by farmers in Garut and Pangalengan from the dry season of 2013 until the dry season of 2015.

Fig. 2 .
Fig. 2. The nutrient applications for chemical N (left) and chemical P2O5 (right) in Garut and Pangalengan for Atlantic and Granola in the different season, the means of varieties and the overall mean.

Fig. 3 .
Fig. 3.The average number of preventive, curative and total fungicide applications per farmer in different regions, seasons and potato variety for different seasons.

Table 1 .
Overview of preventive and curative fungicides used in the different regions, seasons and potato variety expressed as kg Active Ingredient (AI) per ha per cultivation.

Table 2 .
The Treatment Frequency Index of fungicides for late blight control used in the different regions, seasons and potato variety for the different seasons.