Refuge plants and jajar legowo systems increasing the useful arthropods population and the rice yield

Ecological engineering has become a promising strategy for restoring paddy field biodiversity and enhancing yield. This study aims to evaluate the effectiveness of refuge plants and cropping systems on the population of useful arthropods and rice production. The study was conducted on a paddy field in Central Sulawesi, Indonesia, using a Split Plot Design. The main plots were refuge plants with four treatments, namely rice without refuge crops, rice crops with wild Cosmos (C. caudatus), marigold (Tagetes erecta), and both flower plants as the refuges, whereas two cropping systems, i.e., the jajar legowo system (JLS) of 2:1 and the symmetries system (SS) was set up as a sub-plot. Direct observation and collection of the arthropods were carried out using yellow-pan traps and a sweep net. The parasitoid eggs were reared at the laboratory. In addition, rice growth and yields were measured. The refuge plants and cropping system had no significant effects on the overall arthropods population but changed their diversity and community structure and increased the proportion of useful arthropods as well. Meanwhile, rice production in the JLS (5.16 tons GKG ha−1) was significantly higher than those of the SS (4.10 tons GKG ha−1), and this was in line with the differences in the panicles number in both cropping systems. The study highlighted the importance of the implementation of ecological engineering at the paddy fields by using the flower plants as the refuge and food sources for the beneficial organisms combined by JLS for enhancing the rice yields.


Introduction
Rice production in Central Sulawesi has declined in the last five years, and in 2019 only reached 844,944.30tons [1].Rice productivity is influenced by various factors, especially cultivation techniques and climate change [2,3].Generally, farmers use a paddy monoculture system with a high-intensity of use of chemical fertilizers and pesticides.More than 91% of the rice farmers in Central Sulawesi use inorganic fertilizers and chemical pesticides in rice cultivation.However, the majority of their rice plants (96.2%) were attacked by pests and diseases [4].
The intensive application of chemical pesticides as the consequence of the rice monoculture system has detrimental effects on useful arthropods such as predators, parasitoids, and pollinators [5].1306 (2024) 012004 IOP Publishing doi:10.1088/1755-1315/1306/1/012004 2 Meanwhile, chemical fertilizers might increase the performance of rice pests [6] and emissions of the greenhouse gases such as ammonia [2].
Ecological engineering or modification of agroecosystem habitat is considered to be a good solution to enhance the diversity of natural enemies and diminish the harmful effect of agrochemicals [7,8,9,10].This is in accordance with the Regulation of the Indonesian Ministry of Agriculture, in particular Law no.22 of 2019 concerning Sustainable Farming Systems, which recommends the improvement of crop production but minimizing it is harmful impact on the environment and beneficial organisms.
Ecological engineering can be done by planting flowering plants that attract predators or parasitoids to the cultivated crops.Flowering plants provide a food source (nectar, pollen) for the imago of natural enemies and also serve as a shelter for them [11,12].This habitat manipulation strategy may increase the fitness of natural enemies and suppress the rice pests [13,14,15,16].Meanwhile, modifying the rice cropping system, for example, by using the jajar legowo system, has various advantages compared to other cropping systems [17,18].Therefore, it is urgent to study the effectiveness of refuge plants and to crop systems on the diversity of useful arthropods and the rice yield that are cultivated without using synthetic fertilizers and pesticides.

Study area
The research was carried out in Pombewe Village, one of the centres for rice cultivation in Sigi Biromaru District, Sigi Regency, Central Sulawesi Province, Indonesia.The study sites were located at the 121 m asl, with latitude -0.9777741 o and longitude 119.934782 o .The research was conducted from June to November 2020.

Experimental design
The study used a Split Plot Design.The main plots were refuge plants with four treatments, namely paddy without refuge (N), rice plants with C.caudatus (C), marigold/T.erecta(M), and both flower plants (CM) as the refuge, whereas two cropping systems were set up as a sub-plot, i.e., the jajar legowo system (JLS) of 2:1 and symmetries system (SS).The planting distance of both systems was 20 cm x 10 cm x 40 cm and 20 cm x 20 cm, respectively.Each main plot was split into two subplots (20 x 10 m 2 ) and spacing with other main plots as far as 10 m (Figure 1).

Preparation and planting of refuge plants and paddy
The seeds of the refuge plant are first sown in polybags containing soil and sand, which are then watered every day.Seedlings are carried out for about 20 days and planted at the paddy fields using the sequential method two weeks before paddy planting.The refugees were planted as the border plant about 50 cm from paddy fields [15].The paddy variety used is Mekongga.Organic fertilizers of cow manures (organic-C: 15.23%, C/N ratio: 16.38%, pH:7) enriched with local microorganisms were applied at a rate of 3 t/ha a week before rice seedling was planted, whereas the organic fertilizer in liquid form (0.1/l, v/v) was sprayed at 35 days after planting (DAP).A botanical insecticide of Azadirachta indica (0.1/l, v/v) was applied once a week from 38 to 73 DAP for pest control.

Measured Variables a. Diversity of arthropods
Arthropods were observed and collected using a sweep net and yellow pan trap modified by Brotodjojo et al. [15].At each subplot, arthropods were sampled at five sites of 1 m 2 square and three sites on the refuge plants.Sampling was conducted once a week for three months in the morning (07.00-09.00 AM) and afternoon (16.00 -18:00 PM).The arthropods were then grouped based on their role as predators, parasitoids, pollinators, and pests.In addition, the eggs of rice pests were also collected and reared in the laboratory to identify the parasitoids.

b. Number of panicles during the generative phase. c. Rice yields
The average weight of dry milled grain (GKG) is used to measure the rice yields (tons/ha).It was obtained from the conversion of rice yields per tile with the following formula: Rice yield = a number of clumps x number of tillers x number of grains per panicle x weight per 1000 grains [19].

Data analysis
The data obtained were analyzed using analysis of variance (ANOVA) and the HSD test using the Statistix 8. Arthropods diversity based on the Shanon-Wiener index (H') indices and Chao-1 estimated number of species, as well as the cluster analysis, were analyzed using PAST software [20].

Population and diversity of arthropods
A total of 7618 arthropods belonging to 71 species from 9 orders and 52 families were recorded.The results of ANOVA did not detect a significant effect of refuge plant species and cropping system on the arthropod population (refuge: p = 0.328; cropping system: p = 0.842; refuge*cropping system: p=0.266).However, the presence of refuge plants tended to change the diversity of arthropods.The diversity (H') and estimated number of species (Chao-1) of the arthropods in the morning, afternoon and combined both times in rice plants with refuge plants (C, M, CM) was higher than those at no refuge plants (N)(Table 1).

Community structure and functional groups of the arthropods
Differences in the arthropod's diversity due to the presence of refuge plants at the paddy fields were in line with cluster analysis obtained, showing the differences in arthropod community among the treatments.The arthropods community in the paddy field with no refuge plants was quite dissimilar from those with refuge plants.Nevertheless, the arthropod community at the three different refuge plants was highly similar.This similar pattern of arthropods community was found in the morning, afternoon, or a combination of both times (Figure 2).The presence of refuge plants also affected the guild composition or the functional groups of the arthropods.The percentage of pest groups in refuge treatments was lower than in no refuge treatment.On the other hand, the percentage of beneficial arthropods, particularly the pollinators and parasitoids group was quite higher in the paddy fields with refuge plant (Figure 3).The relatively high abundance of predators at all plots indicates that the free agrochemicals paddy cultivation still supports the population of those natural enemies.This is another advantage of the refuge plants; besides, it is potency to reduce the use of synthetic insecticides and increase the crop yield [14].
The results of this study are consistent with previous studies, which showed that habitat diversity using flowering plants surrounding the rice fields promotes the diversity of natural enemies, especially parasitoids and predators [11,12,21], and may reduce pest population [13,14,15,16,22].However,the use of flowering plants as refuge plants does not always succeed in increasing natural enemies and suppressing rice pests [23].For example, if pests also prefer to feed on and dwell in flowering plants, this could lead to more damage to cultivated crops and increase ecosystem disservices [11,24].Additionally, the presence of the refuge plant might not affect the spatial and temporal abundance of the overall natural enemies, but it was taxa dependent response [25].Therefore, several factors should be considered, such as the characteristic of the flowering plant and the target pests, to enhance the successful implementation of this ecological engineering.

Number of panicles and rice yields
The results of the analysis showed that there was a significant effect of cropping systems on the number of panicles (Table 2).The average number of panicles in the JLS was significantly higher than those of SS.In concordance with the number of panicles, the rice yield was significantly affected by cropping systems but not by refuge plants (Table 3).The yields in the JLS (5.16 tons ha -1 ) were significantly higher than those of the SS (4.10 tons ha -1 ).A previous study reported that the rice yield in the JLS was generally higher than in the conventional cropping system [26].The advantage of the JLS is that the legowo row provides many peripheral plants.Those plants will get more sunlight, better air circulation, and will get more nutrients than the symmetries/tile cropping method [17].In addition, the plant population in the legowo system is higher, so it provides an opportunity to get high yields [3,18].The higher rice yield at the JLS may also be related to its effect on the pest population.When the pest population is high, the Legowo cropping system with more space is reported to be able to suppress the population of pests such as L. acuta, Riptortus sp., and A. similis.In contrast, the tile system has dense clumps and no space.This condition is more favored by phytophagous insects [27].

Conclusion
Planting flowering plants such as C.caudatus, marigold, or a combination of both flowers as the refuge plants increases the diversity of arthropods and enhances the population of useful arthropods (parasitoids, pollinators, and predators) at the paddy field.Rice yield in the jajar legowo system (JLS) of 2:1 was higher than the symmetries system (SS) i.e. 5.16 vs 4.10 tons GKG ha -1 .The increase in paddy yields was in line with the higher number of panicles in the JLS compared to the SS (312.75 vs 248.45 m -2 ). 4

Figure 2 .
Figure 2. A dendrogram showing the similarity of arthropod communities at the paddy fields with different refuge plants in the morning (A), afternoon (B) and combined of both time (C).C = C. caudatus, M= Marigold, CM = C. caudatus + Marigold, N = No refuge plants.

Table 1 .
Diversity of arthropods at the paddy fields with different refuge plants in the morning, afternoon, and combined both times.

Table 2 .
Population of panicles (number m -2 ) with different refuge and cropping system.C. caudatus M= marigold, CM = C. caudatus + marigold, N = no refuge plants, JLS = jajar legowo system, SS = symmetries system.Average numbers followed by different letter are significantly different.

Table 3 .
Paddy yields (tons ha -1 ) with different refuge and cropping system