Assessing True Shallot Seed (TSS) performance and farmers’ perceptions of its innovation attributes

True Shallot Seed (TSS) cultivation is a new shallot production technique that does not use vegetative tubers as planting material, but rather seeds. This study aimed to examine the field performance of TSS and assess farmers’ perceptions of TSS innovation attributes. On-farm participatory trials involving eight farmers were conducted in Cirebon and Brebes during June-August 2020. Compared to Trisula and Maserati, Sanren’s performance was the best in both the nursery and transplant stages. Maserati showed the second-best performance in seed emergence and yields. Meanwhile, although farmers had rated Trisula high scores for color, pungency, and bulb shape, poorer yields made Trisula the least financially viable of the others. Relative advantage, compatibility, trialability, and observability of TSS are rated positive by farmers. In terms of complexity, most farmers perceive that learning and understanding TSS innovation is quite a challenge for them. Bivariate correlation analysis indicates that the more positive farmer’s perception of relative advantage, compatibility, trialability, and observability of TSS, the more positive their intention to adopt TSS will be. However, complexity has a negative significant correlation with the intention which means that the more complex TSS innovation is perceived by farmers, the weaker their intention to adopt TSS will be.


Introduction
As a flavoring ingredient in various daily culinary options and an herbal spice ingredient that has health benefits, the shallot is one of the priority vegetable crops for Indonesians.Comparing the last-two decades' data (2001-2010 and 2011-2020), the shallot production annual growth rate increases from 3.05% (1 st last decade) to 5.50% (2 nd last decade).There is also an increase in harvested area growth rate from 2.65% (1 st last decade) to 5.34% (2 nd last decade).However, there is a decrease in yield growth rate from 0.40%/year to 0.16%/year.In addition to that, during those two last decades, the national shallot yield has not moved from the range of 8.5-10.3t/ha, which may indicate a tendency of yield stagnation.The adoption of high-yielding variety will ameliorate most of the challenges currently bedeviling shallot productivity in Indonesia.
The challenges in Indonesian shallot have led to the development and innovation of True Shallot Seeds (TSS) since the late 90' which is projected as a way of enhancing shallot productivity [1,2,3].Some advantages of TSS over those of tuber seeds are (a) seeds are not bulky and easy to transport, (b) no storage is needed, (c) costs of planting materials are potentially lower, (d) long-term shelf life, and (e) less transmitted by viruses and other seed-borne diseases, therefore it potentially results in a higher yield [4,5].However, after the introduction of TSS more than two decades ago, some studies [6,7] suggest that the adoption rate of TSS at the national level is still quite low.Despite the stated impact on productivity and other advantages, low adoption suggests there are emerging factors limiting the adoption of agricultural technology among farmers [8,9,10,11].
Findings from various studies have shown that the adoption of agricultural technology depends on five main broadly classified categories: (i) farmers' characteristics; (ii) farm characteristics; (iii) technology characteristics; (iv) institutional factors; and (v) finance [12,13,14,15,16,17].Referring to these findings, the low adoption of technology occurs due to limiting factors that prevent the five categories from working as they should.At the micro level, several barriers that have contributed to low adoption technology are: (1) limited conformity of innovation attributes with farmer's preferences and lack of agro-ecological targeting, (2) inadequate interaction and rather weak coordination among institutions responsible for technology generation and dissemination, resulting in lack of awareness and low technology adoption at farmer's level, (3) technology transfer and delivery systems operate suboptimally with lack of work ethics and commitments, (4) programs put too much emphasis on yieldincreasing strategies with inadequate attention to the development of value chain and technology commercialization, and (5) enabling policy and institutional environment are weak-poor priority setting, inadequate funding of research/technology, and inefficient regulatory frameworks [9,18].While these barriers actually have not been exhausted, this paper is limited to focusing on applying a participatory on-farm trial and assessing farmers' preferences on innovation attributes.
Experience has shown that innovations/technologies that are developed in research stations without the participation of farmers are often not adopted by farmers.Ignoring farmers during the technology generation process without considering their experience and knowledge results in innovations that are less likely to match their values and needs.The participatory research approach is developed to respond to the shortcomings of the top-down research approach that frequently failed to deliver significant improvements at the farmer's level.This method is intended to bridge the divide between research and practice by engaging farmers as partners in the research process and by sampling the wide variation in environmental contexts and farm practices.The involvement of farmers in all the stages of technology development improves the probability of adoption since it creates a sense of ownership and credibility among farmers.Incorporating their need and perception during the technology generation and improvement process is an important strategy to enhance agricultural technology adoption [19,20,21,22].
The innovation-diffusion process is described as an uncertainty reduction process, so the attributes of innovations are proposed to help to decrease uncertainty about the innovation [23].Previous studies suggest individuals' perceptions of an innovation's attributes are important predictors of attitudes toward possible adoption and ultimately adoption decisions [24,25].Attributes of innovations include five characteristics of innovations: (1) relative advantage, (2) compatibility, (3) complexity, (4) trialability, and (5) observability.Relative advantage or perceived total benefits of innovation has been indicated as a decisive factor determining adoption [26,27].As an innovation is compatible with the existing norms, values, experiences, needs, and available resources, the higher its chance to be adopted [28].The complexity of innovation has a negative relationship with adoption.Innovations that are simpler to understand by an individual will be adopted more rapidly than innovations that require the potential adopter to have new skills and understanding [29].The degree to which a potential adopter can try something out on a small scale before adopting is a major determinant of technology adoption [30].In terms of observability, when impacts are easily noticeable, the person will adopt the innovation to show others the resulting successes [31].
A dissemination activity of True Shallot Seed (TSS) that involves more than one farmer in a participatory on-farm research set-up has never been done before.Based on this rare opportunity, the study aims to assess the performance of True Shallot Seed (TSS) and farmers' perceptions of its innovation attributes.

Materials and Methods
A participatory on-farm research approach was selected to speed up the dissemination of TSS innovation by engaging farmers as active participants in the research process on their farms.Brebes, Central Java, and Cirebon, West Java were the targeted locations representing the largest shallot-producing area in Indonesia.Five farmers from Brebes and three farmers from Cirebon were purposively selected based on their voluntary willingness to participate.These farmers were asked to set aside a part of the land they would be using to plant traditional shallot (bulb seeds) and spare it for TSS.Tuber planting (traditional) and seedling transplanting (TSS) were carried out at the same time so farmers could directly compare the field performance of both.TSS was cultivated using the transplanting method hence farmers were involved since the nursery stage.Three varieties of TSS, namely Maserati, Sanren, and Trisula were used.Data on germination percentage, number of usable seedlings (nursery), fresh yield, and dry yield (field) were collected.
After harvesting farmers were asked to evaluate product attributes of TSS bulbs.Those product attributes were bulb shape, bulb size, number of bulbs per plant, bulb skin color, aroma/ pungency, and yield.The scoring used was 1=very poor, 2=poor, 3=acceptable, 4=good, and 5=very good.
One week after harvesting, farmers were interviewed using a structured questionnaire.The questionnaire consisted of 41 statement items covering: relative advantage (7 items), compatibility (6 items), complexity (3 items), trialability (3 items), observability (2 items), attitude (6 items), acceptance of consequences (7 items), and intention (7 items).A five-point Likert scale was used to measure each item in the questionnaire (1=strongly disagree; 2=disagree; 3=undecided; 4=agree; and 5=strongly agree).Cronbach's alpha tests were carried out for each construct to see if multiple-question Likert scale surveys were reliable.Descriptive statistics and Pearson bivariate correlation were used for data analysis.

Characteristics of farmer participants
On average, the farmers involve in the on-farm trial are 44.5 years of age with the youngest is of 28 and the oldest is of 61 years old.Most of them only attend and finish elementary school.The average total land-holding is 3,650 m 2 , which is generally the total area of several farm parcels under their control.The number of household members is between 3 and 5, averaging 4 persons, and at least one of them is helping the farm-works.All farmers have participated in general shallot training/extension and on average they have 13.5 years of experience in shallot cultivation.Six farmers have participated in TSS cultivation training/extension, but only one out of eight farmers has experience to cultivate TSS.

Performance of True Shallot Seed (TSS)
Two nurseries are constructed close to the farmer's field each in Brebes and Cirebon.In each location, by taking into account the seedling needs, 15 m 2 of nursery is allocated for each variety.There are 10 planting lines per 1 m 2 in which 4 g of seeds are sown.Since the seeds from the three varieties weigh differently, the number of seeds sown in each line is also not the same.Weight per 100 gr of Trisula, Maserati, and Sanren is 0.25, 0.32, and 0.36 gr; hence the number of seeds per line is approximately 160, 125, and 111 seeds respectively.Table 2 shows that Sanren has the best germination percentage in Brebes nursery.In both locations, at 21 days after sowing, Trisula shows the lowest germination with 24% (Brebes) and 47% (Cirebon).At 45 days after sowing (transplanting stage), Sanren variety is averaging the highest percentage of useable seedlings at 65%, while Trisula shows the lowest percentage at 30%.In TSS cultivation, the cost of planting material (seedlings) is not only determined by the seed cost but also the cost of growing seedlings in the nursery.Therefore, the germination rate, final useable number of seedlings, and nursery production costs will very much affect the costs per seedling.For professional seedling growers, it is important to determine the seedling costs in order to establish a selling price.
Table 3 shows the seedling production costs for a 30 m 2 field bed nursery surface with a depreciation period of 12 months for reusable nursery materials and 2 month raising period.Fixed costs are 2.07%, while variable costs are 97.93% of the total production cost.In variable costs, the highest cost component is seeds (39.37%), followed by labor (29.31%).The summation of all costs (total production cost) supported by the information about the seed weights, seed number/line/m 2 /30 m 2 , and usable seedlings ultimately generate the cost per seedling for each variety.Seedlings are harvested from the nursery 45-50 days after sowing, and transplanted to the field that is ready for planting.The planting distance used is 10 x 10 cm.Agronomic treatments and pest-disease controls are no different from those applied to common shallot (bulb seed) cultivation.Shallot bulbs are harvested 85 days after transplanting.Table 4 shows the average fresh yield (weighed immediately after uprooting the plants), and the average dry yield (weighed after 6-12 days of sun-drying) of the eight farmers involve in the on-farm trial.The observed average fresh yield of Sanren (37,670 kg/ha) is the highest, followed by Maserati (21,389 kg/ha) and Trisula (13,443 kg/ha).Trisula shows the lowest yield although the plant density used is exactly the same as the other two varieties.The dry yield of shallots is about 42.0-48.5% less than the fresh weight.Seedling costs are calculated based on a plant density of 100 plants per square meter plus extra seedlings (0.5%) to provide easiness for selecting the best seedlings to be transplanted and to supply sufficient seedlings for re-transplanting when some plants died during the first two weeks of planting.Different seedling prices for each variety are used based on nursery results.Maserati, Sanren and Trisula are priced at IDR 42.6; 36.1, and 50.6 per seedling respectively.Other costs are obtained from farmers' farm records.Seedling and labor costs are the two biggest components of production costs.The average production cost breakdown is seedlings (33.9%), fertilizers (4.5%), pesticides (6.0%), and laborers (55.6%).One thing that stands out a bit is that the average labor cost in Brebes is higher than in Cirebon.At a shallot market price of IDR 15,000/kg, 3 farmer-participants growing Trisula, 6 farmer-participants growing Maserati and 8 farmer-participants growing Sanren get profits from their TSS farms.Both in Brebes and Cirebon, on average, Sanren is the most profitable variety cultivated by participant farmers, followed by Maserati.Meanwhile, due to low yields, Trisula in both locations suffers losses on average.At harvest, farmer participants are asked to evaluate some product attributes of the three varieties tested.Table 6 shows that yield and number of bulbs per plant are perceived as the highest for Sanren, followed by Maserati and Trisula.In terms of bulb shape and bulb size no significant differences in farmers' perceptions are present.Although Sanren is evaluated as the least-performing variety in terms of bulb shape, it is still rated reasonably high with a score of 3.3.Maserati and Trisula are equally rated very high for bulb skin color, while Sanren is the lowest.
Trisula is rated as the most pungent among the three varieties, followed by Maserati and Sanren.

Innovation attributes and intention to adopt TSS
Relative advantage is the degree to which an innovation is better than the technology that it is superseding.Farmers perceive that TSS cultivation is more profitable, less seed costly, socially more prestigious, and also potentially provides better farm cash flow.However, farmers perceive that the total production cost and production risks of TSS are not lower than bulb-seed cultivation.The mean score of total items (3.250) may suggest that farmers are just slightly in favor of the TSS advantages.Compatibility refers to the degree to which an innovation is consistent with the existing values, past experiences, and needs of potential adopters in terms of compatibility; farmers even agree that TSS to some degree may help them reduce their use of pesticides and fertilizers.Most compatibility items are rated greater than four (total items' mean score of 4.00) suggesting that TSS innovation is most likely relevant to farmers' needs, experience, and values.Complexity is the degree to which an innovation is perceived as difficult to understand and adopt.Greater complexity implies an increased degree of difficulty in understanding the use of a given innovation.Therefore complexity is assumed to be negatively associated to use intentions.Most farmers are in disagreement with all items that constructed complexity attributes and perceive that learning and understanding TSS innovation is not easy for them.The overall mean score of 1.83 implies that it is quite a challenge for farmers to deal with the complexities of TSS innovation.Trialability (also referred to as divisibility) is the degree to which an innovation may be experimented with on a limited basis before deciding to adopt.Most farmers agree that obtaining information and experimenting with TSS is a necessity before adopting TSS.Furthermore, farmers also tend to disagree that a lack of knowledge will prevent them to conduct a small-scale TSS trial.In this study, the trialability of TSS is perceived as positive by farmers with a total item mean score of 3.75.Observability is the degree to which the results of an innovation are noticeable/visible/ observable, demonstrable, and communicable to others.Observability of TSS is rated positive by farmers (the overall mean score of 3.44) as reflected by their easiness of observing the benefits of TSS from demonstration plots.Furthermore, farmers also seem to agree about the workability and replicability of TSS innovation.Farmers establish their attitudes based on their perception of what may be true about a particular subject.This perception may or may not be based on information, knowledge, or even an emotional reaction to the subject, sometimes supported by beliefs and values.The total mean score of 4.10 indicates that based on their judgment about the consequences of TSS use, farmers have a favorable evaluation of the TSS adoption.Based on their experience, values, and needs farmers always weigh the desired and undesired consequences before adopting a new technology.Farmers' acceptance of the consequences of adoption will greatly influence their adoption decision.The average score of the total items (4.02) indicates farmers' readiness to accept all the consequences that may lead to the adoption of TSS.Intention is a measure of the likelihood of an individual is inclined to a particular technology or innovation.The intention is frequently considered an immediate determinant of actual adoption.The stronger the intention, the greater will be the probability of adoption.The total mean score of 3.68 suggests that farmers have a quite strong intention to accept TSS.Bivariate correlation analysis shows that relative advantage and trialability are significantly correlated with consequence and intention.Compatibility is positively and significantly related with attitude, acceptance of consequences, and intention.Complexity is negatively and significantly correlated with attitude, acceptance of consequences, and intention.Observability is positively and significantly related to the intention to adopt TSS.The attitude of farmers toward TSS is significantly correlated with the acceptance of consequences and the intention of adopting TSS.The level of acceptance of TSS adoption consequences is significantly related to farmers' intention to adopt TSS.

Discussion
The TSS planting method most frequently disseminated to farmers is the transplanting method using seedlings.In this method, the cost of planting material (seedlings) is not only determined by the seed cost but also the cost of growing seedlings in the nursery.Taking into account both seed and nursery costs, the price per seedling of Sanren is the cheapest because Sanren shows the highest germination percentage in the nursery and the highest percentage of seedling survival or useable seedlings for transplanting.Price per seedling has important implications for shallot bulb production since the cost of planting material is the second biggest component in total production cost [32,33,34].Using the same plant density, the yield of Sanren is the highest, followed by Maserati and Trisula.This is consistent with the farmers' evaluation of product attributes which states that the yield and number of bulbs per plant of Sanren is the most preferred compared to the other two varieties.Meanwhile, the production cost per hectare of Sanren is the lowest mainly because of its cheapest price per seedling.Consequently, on average, Sanren is the most profitable variety cultivated by farmers, followed by Maserati.Unfortunately, due to much lower yields, Trisula in both locations suffers losses on average.The agronomic and economic performance of Sanren shows the potential of Sanren as the best choice variety for farmers.
In terms of relative advantage attribute, farmers perceive that TSS cultivation is more profitable, less seed costly, socially more prestigious, and also potentially provides better farm cash flow.However, farmers also perceive that TSS production is more expensive and riskier.The overall score of relative advantage may indicate that farmers are only slightly in favor of the TSS benefits.TSS is an innovation that farmers perceive as consistent or compatible with their needs, experience, and values.Meanwhile, 1230 (2023) 012191 IOP Publishing doi:10.1088/1755-1315/1230/1/01219111 farmers perceive an uneasiness to learn and understand TSS suggesting that it is quite challenging for them to deal with the complexity of TSS innovation.However, farmers seem willing to overcome this challenge because the attributes of trialability and observability are rated positive by farmers.
Based on their judgment about the prospects of TSS use, farmers have a favorable evaluation of the TSS; hence farmers show positive attitudes toward TSS adoption.These positive attitudes direct farmers willing to accept the consequences of TSS use and strengthen their intention to adopt TSS.Relative advantage and trialability are significantly correlated with consequence and intention as suggested in previous studies [26,35,36].The higher the perceived benefits and divisibility of TSS, the higher will be the level of acceptance of consequences and the intention of adopting TSS.In line with findings from other studies [27,37,38], compatibility is significantly correlated with attitude, acceptance of consequences, and intention.The more positive the farmer's perception that TSS is in accordance with their values and needs, the more positive their attitude, risk acceptance, and intention of TSS adoption will be.Among the five attributes of innovation, complexity is significantly and negatively correlated with attitude, acceptance of consequences, and intention, similar to the results obtained from previous studies [30,39,40].The more complex TSS innovation is perceived by farmers, the weaker their attitude, acceptance of consequences, and intention will be.Findings similarly resulted from other studies [41,42,43] suggest that observability is positively related to the intention to adopt TSS.The more noticeable and demonstrable TSS perceived by farmers, the higher will be their intention to adopt TSS.
The more favorable the attitude of farmers toward TSS, the stronger will be their acceptance of consequences and intention in adopting TSS.Bivariate correlation also suggests that the higher level of acceptance of TSS adoption consequences, the stronger farmers' intention to adopt TSS will be.These findings are in line with other behavioral studies that explain attitude as a key variable and fundamental build block of behavioral change -an individual's positive or negative feelings about a perceived outcome [44,45,46].Attitude is influenced by the judgment of an individual about the effects and consequences of behavior [47].In the context of this study, an attitude refers to a farmer's global positive or negative evaluation of the behavior in adopting TSS, and the effect of consequences is the farmer's acceptance of consequences arising from the adoption of TSS.Generally, if farmers have higher awareness and more positive evaluation of TSS innovation, and higher acceptance of consequences, their willingness to adopt TSS will be higher.Several studies [48,49,50,51] have reported the role of attitude in predicting farmers' intention to adopt an innovation.Attitude can be regarded as a significant determinant of an individual's intention to adopt [52,53].Meanwhile, behavior intention reflects a farmer's motivation regarding their plan or conscious decision to apply effort in carrying out a particular behavior [54,55].Furthermore, behavior intention represents the immediate antecedent and best predictor of performing an actual behavior [56,57].In the context of this study, intention refers to the extent to which farmers are motivated to adopt TSS innovation.By implication, a more positive evaluation of TSS may lead to stronger behavior intention towards TSS and results in a stronger likelihood of the performance of actual TSS adoption.
As indicated in this study, the relative advantage is positively correlated with intention, while complexity shows a negative correlation with the intention to adopt TSS.A complex innovation or technology is one that is based on principles unfamiliar to the farmers.The complexity of innovation increases the greater the number of farm sub-systems affected when adopting it, and the more severely they are affected.The magnitude, riskiness, and stress associated with the prospect of adoption could be expected to influence the attractiveness (relative advantage) of TSS innovation.A detailed analysis of the adoption decision is only likely to commence once farmers accept TSS is relevant to their farm system.Therefore, more comprehensive studies that explore relative advantage and complexity from the system perspectives are needed to assess whether the perceived benefits of the innovation can steer farmers' intentions and outweigh its perceived complexity.

Conclusion
Sanren has the best germination percentage and the highest number of usable seedlings in the nursery.
The calculation of seedling production cost in the nursery shows that the price or cost per Sanren seedling is the lowest (IDR 36.1),followed by Maserati (IDR 42.6), and Trisula (IDR 50.6).Immediately weighed after harvest, the average fresh yield of Sanren is the highest (37,670 kg/ha) compared to Maserati (21,389 kg/ha) and Trisula (13,443 kg/ha).The average yield loss due to drying for each variety is Maserati (42.0%,Sanren (48.5%), and Sanren (44.3%).Production costs between the three varieties do not differ much, averaging IDR 133,766,642 per hectare.In terms of product attributes, farmers score Sanren the highest for yield and number of bulbs per plant.No significant differences in bulb shape and bulb size.Maserati and Trisula are equally rated very high for bulb skin color, while Sanren is the lowest.Trisula is rated as the most pungent among the three varieties.
In terms of relative advantage, farmers perceive that TSS cultivation is more profitable, less seed costly, socially more prestigious, and also potentially provides better farm cash flow, but not for the total production cost and production risks of TSS.Most compatibility items are rated high suggesting that TSS innovation is perceived as relevant to farmers' needs, experience, and values.In terms of complexity, most farmers perceive that learning and understanding TSS innovation are not easy which means that dealing with the complexities of TSS is quite a challenge for them.A necessity for obtaining information and experimenting with TSS before adoption implies that the trialability of TSS is perceived positively by farmers.The observability of TSS is rated positive by farmers as reflected by their agreement on the workability and replicability of TSS innovation.Based on their judgment, farmers show a favorable evaluation of the TSS (positive attitude), a readiness to accept all the consequences of adopting TSS, and a quite strong intention to adopt TSS.Innovation characteristics (relative advantage, compatibility, trialability, and observability) have a positive significant correlation with intention.The more positive farmer's perception of the relative advantage, compatibility, trialability, and observability of TSS, the more positive their intention to adopt TSS will be.However, complexity has a negative significant correlation with the intention which means that the more complex TSS innovation is perceived by farmers, the weaker their intention to adopt TSS will be.
Data used for this study are basically collected from a participatory on-farm trial approach.The limited number of farmers involved (8) is mainly due to the limited resources available.The sample size is still justified for assessing the TSS field performance, but too small for evaluating farmers' perceptions of TSS innovation attributes.Too small a sample may prevent the findings from being extrapolated.Therefore, caution needs to be exercised when using and interpreting innovation attribute research results.It is recommended to carry out a further survey research study with a representative and justified sample size to obtain reliable data by using a set of scale or questionnaire items that have Cronbach's alpha tested.

Table 2 .
Percentage of emerged seedlings 7, 14, and 21 days after sowing and percentage of usable seedlings 45 days after sowing

Table 3 .
Nursery production cost and the cost-price per seedling

Table 7 .
Relative advantage -mean score and standard deviation

Table 8 .
Compatibility -mean score and standard deviation

Table 9 .
Complexity -mean score and standard deviation

Table 10 .
Trialability -mean score and standard deviation

Table 11 .
Observability -mean score and standard deviation

Table 12 .
Attitude towards TSS -mean score and standard deviation

Table 13 .
Adoption consequences -mean score and standard deviation

Table 14 .
Intention to adopt TSS -mean score and standard deviation

Table 15 .
Correlations between innovation attributes and attitude, intention, and consequence