The Economic Value of Sustainability of the Integrated Crop-Livestock System in Tropical Regions

The objective of this study was to evaluate the potential of improve economic value of integrated crop-livestock systems in comparison to conventional systems specialized in monoculture. Empirical studies have demonstrated the environmental benefits of integrated crop-livestock systems, however the potential for 35 creating economic value these systems are controversial, especially in emerging countries, where the necessity 36 to expand the food supply needs be associated with better land use. This research evaluated six models of 37 integrated systems and two conventional systems (corn grain production and pasture beef cattle production) in the south-eastern region of Brazil for two years. The models were conducted in an experiment to replicate the 39 main management possibilities in the integrated systems. We show for the first time the economic impact 40 analysis combined the risk optimization and discounted cash flow techniques based on Monte Carlo simulation, 41 considering the price and productivity uncertainties of each system. Results indicated that, for the indicators 42 of added value and return on investment, integrated crop-livestock systems had an economic advantage when 43 compared to conventional systems. It was also found that integrated crop-livestock systems needed a smaller operational area for the economic break-even point to be reached.


55
Systems that integrated animal and plant cultures were the basis of food production for ancient

126
For the ICLS, four types of Marandu grass and maize intercropping were studied. The same cultivar, 127 row spacing, seeding density, and fertilizers as described for the crop system and the same seeding density, 128 fertilizers as described for the livestock system were used for all integrated systems. For ICLS-1, Marandu 129 grass was sown simultaneously with maize in the sowing row. For ICLS-2, simultaneous sowing was also 130 performed, but 20 days after maize germination 200 ml ha -1 of the herbicide nicosulfuron (8 g ha -1 of active 131 ingredient) was applied. For ICLS-3, Marandu grass was sown 20 days after maize had been sown (lagged 132 sowing), for this purpose, the grass seed was mixed in the fertilizer for the second fertilization and between-133 row sowing was performed using a cultivator. For ICLS-4, Marandu grass and maize was sown simultaneously, 134 but with the grass seed sown within and between the maize rows, resulting in a spacing of 37.5 cm. Exclusively 135 for this system, the sowing fertilizer and the amount of grass seeds were divided between and within the maize 136 rows to guarantee an equal mixture of grass seed and fertilizer. In addition, 200 ml ha -1 of the herbicide 137 nicosulfuron (8 g ha -1 of active ingredient) was applied 20 days after maize germination. In all integrated 138 systems, the maize was harvested in May 2016. Ninety days after harvest, the pastures were ready for grazing.

141
For economic analyses, two years of the project were used, in which the results of the first corn harvest 1 to ICLS-4, animals in the growing phase until fattening (finishing) were considered, using 50% of the carcass

181
of the experiment to model cash flow (Table 1). Using this, it was found that the modal size for a rural property 184 Table 1. First and second year production results for the empirical study system field-trials.

185
Year  give an average value of R $ 38.14 for the 60 kg sac of maize and R$ 152.31 for the @ beef cattle (@ = 15 kg),

193
with 17.32% and 9.33% being the respective variation coefficients. Price variability is shown Table 1, and this depreciation, were not included, since the aim was to determine the economic break-even point of each activity.

198
The variable costs of maize production include spending on soil preparation, planting, crop management and 199 harvesting activities. Variable costs for meat production included expenses for the purchase of animals,

213
It should be noted that the volatilities used were considered as independent, since the correlation 214 between price variations of maize and beef was 0.27 and without statistical significance. To generate 215 simulations, the possibilities of maize and beef prices and beef productivity were generated using a normal 216 distribution pattern, following identification of normality via a Jarque-Bera test. For corn productivity, a 217 discrete distribution pattern was used using the average, minimum and maximum values from the Crop System,

227
To quantify each production system as a rural property production strategy, the calculation of 228 valuation in perpetuity was used and, to provide a conservative approach, a real growth rate was not assumed

264
For the risk-free rate, the Selic rate that backs Brazil's national treasury bills for January 2019 was 265 used (when net remuneration was estimated at 6.4% per annum). The historical difference used by market 266 analysts for the Brazilian market premium (RM -Rf) was taken (8.9% per annum).

267
To determine the exact systematic risk for each production system, it would be necessary to analyse 268 the covariance of past results for each system using the returns on the Brazilian market portfolio (Ibovespa).

269
However, as a lack of information makes this impossible, risk of each production system were estimated Where: w = weight of each asset (maize or beef) within the total system revenue; σ = risk, measured by the 277 standard deviation in price changes for each asset.
Equation 9 allowed risk determination for each production system and for those with more than one 280 product, allowing evaluation of the effect of diversification on the risks involved, when considering the second 281 part of Equation 9, which aggregates the effects of covariance between maize and beef individual risks.

282
From this, the risk for each system was related to the Ibovespa-based risk, where β = 1, making it Where, βs = overall production system risk (s); σs = risk for each production system; σm = market portfolio risk 287 (Ibovespa).

288
It should be noted that this procedure was performed as a proxy to identify the risk in each production 289 system, which is expressed in the DCF model by the discount rate (i) and appears directly in the calculations 290 of equations 1, 2, 3, 4 and 7.

291
In agribusiness-related literature a risk-free rate is frequently used as a discount rate for investment

310
To calculate the current value of each variable, the discount rate (i) of each production system was

320
The economic risk of the ICLS expressed as a discounted rate showed a high level of diversification, 321 this was due to the weak price correlation between a sac of maize and Beef cattle (@) (0.27), which increased 322 the natural hedge of these production systems, whose response was shown in the associated interest rates.
Even though the ICLS financial results are higher than those from the CS, the impact of risk 326 diversification for each system, and the different fixed capital investment requirements must be comparatively 327 evaluated, that is, in the differences in requirements for machinery, equipment, implements, tools, installations 328 and utensils must be considered in such calculations.

329
Accordingly, Table 4  Note: The NPV averages were statistically different using the two-tailed t-test, with a 5% confidence level.

343
The differences in investments in fixed capital demonstrated that the ICLS required higher levels of 344 spending on long-term resources. This comes from the need to develop more than one agricultural activity in 345 the same area, which reinforces the need for an economic analysis of the viability of this investment. The

346
Livestock System was not economically viable. Even with the lowest risk involved, it was the system with the 347 lowest rate of return and, in effect, the lowest probability of having a positive NPV, across 10,000 simulations.
the lowest among all the production systems analyzed, the NPV of this system was positive.
investment level, so increasing the levels of profitability of the property (ROI), and resulting in positive NPV 354 delivery having a high occurrence probability (> 90%).

355
For the treatments, the ICLS-related enhanced cash flow generation capacity had an impact on the 356 area necessary to make each system viable, as can be seen in the breakeven points calculated in formula 7. All 357 systems showed a positive contribution margin but, due to the value that each system generated across the 358 different investment profiles, the ICLS had a lower BEP.

359
The per hectare production system valuation, that is, the perpetuity calculation for the capacity of each

364
It was clear that, in the long term, between production system differences existed in the potential for market and production uncertainties, allowed for a more robust economic analysis to be conducted.

380
The literature refers to two main economic benefits of ICLS. The first is scope economics, which 381 occurs when the cost of producing two products in the same production system is lower than if the same 382 products were produced separately (Panzar and Willig, 1981). In other words, it is the saving obtained due to 383 the scope of the production unit (Mendonça et al., 2020). This is one of the hypotheses that explains the increase 384 in the cash-generating capacity of the ICLS systems, as shown in the results in Table 2. The second benefit is

387
The Crop System was found to have a higher activity risk value (6.97%) than the Livestock System

388
(2.98%), while for ICLS risk was 50% less than the Crop System, but 6.04% greater than the Livestock System.

389
As a result, it was possible to assert the benefits of combining livestock with an agricultural system, so

393
Nevertheless, the results showed that, overall, livestock is the activity with the lowest risk.

394
Risk reduction in agricultural activities when ICLS is adopted has also been reported by Ryschawy et for ICLS compared to a Crop System, plus the participating farm had greater autonomy to reduce total cost. In 402 addition, sensitivity analysis showed that, unlike CS, ICLS were less likely to be affected by fluctuations in 403 the price of inputs and sales, as a result of diversification. One of the differences in this study compared to the 404 others, regarding risk diversification, is that, via CAPM, it was possible to target this risk reduction in terms of 405 the discount rate, a practice not generally used in agricultural systems feasibility studies (Farinelli et al., 2018).
include the effect of the correction between them (Formula 9). Accordingly, the ICLS betas can lie very close Although the Crop System has a positive NPV, its NPV was impacted by variation in the system´s 439 crop production indicator, since in the second experimental year the of corn sacs per hectare production was 440 lower than in year one. Grain production may have been affected by unfavorable climatic conditions during 441 the second harvest. This factor is linked to the higher risk of agricultural activity, as shown in Table 3.

442
The different ICLS, vary in relation the NPV in ways related to sowing techniques and how the corn 443 and pasture linkage was implemented. Higher corn productivity was obtained in the ICLS-2 and ICLS-4 444 treatments (Table 1)  viably, with this being 56% smaller in than that needed for the Crop System and 80% smaller than that required 468 for Livestock.

469
The use of CM is necessary to achieve a BEP and, as all

489
(2017) reported that these integrated agriculture formats require the development of assessment methods at 490 various local, regional and global levels, with analytical capacity in the areas of social and human sciences.