1. Introduction
Dairy production, especially in developing countries, provides regular income for smallholders, thus contributing to rural development and decreasing rural exodus [
1]. In Brazil, owner labour accounts for 73% of the total rural labour and 67% of the work of family farming. Furthermore, dairy farms are present in 98% of Brazilian cities, and these are predominantly small and medium proprieties that contribute locally to social and economic development [
2]. Small and medium Brazilian dairy farmers are responsible for 58% of overall milk production (including cow and goat milk), which demonstrates its importance to the economy and Brazil’s milk market [
3]. Brazilian dairy farms are mainly pasture-based systems, where 80% fit into the category of family farming [
2,
3,
4]. Despite higher temporary unemployment that causes some instability and social risk to farm workers [
5], small and medium Brazilian dairy production is still an important source for employment in rural zones.
Given the social and economic importance of milk production, keeping smallholders involved in dairy production is a challenge. According to the United States Department of Agriculture [
6], thousands of farmers leave the business every year. In Brazil, rising rural production costs combined with milk price devaluation are listed as key reasons for dairy farmers to withdraw from the activity [
7]. For this reason, the intensification of dairy systems is one of the most common strategies used to increase the income of dairies [
7]. According to Powell et al. [
8], the intensification of production systems and the closer integration of crop and livestock is a form of adaptation that can enable smallholders to make socioeconomic changes and adapt to market and trade circumstances. In this perspective, the integrated dairy–crop system is an evolutionary process [
9] that aims to increase agricultural production per unit of land [
10]. In addition, the integrated dairy–crop system is a strategy of intensification that increases milk output relative to inputs of feed, labour, land, or herd size, thus raising efficiency and revenue and aiding productivity gains [
7].
Dairy system intensification commonly results in higher productivity and profitability [
11]. In turn, dairy intensification is strictly linked to livestock feeding programmes. Thus, on the one hand, intensification is associated with higher levels of productivity and profitability, and on the other hand, this intensification drives the farm to be more dependent on external inputs [
12,
13]. This can cause higher pressure on the environment [
14,
15]. From an economic perspective, livestock feeding programmes can represent more than half of the economic [
16,
17] and environmental costs [
14,
15]. In Brazil, Agostinho et al. [
12] analysed 92 Brazilian dairy farms and concluded that Brazilian dairy systems could be divided into five groups: group G1 and G2 (semi-intensive), G3 (extensive), and G4 and G5 (intensive). Among them, groups G2 and G3 were the most prevalent, representing 40 and 50%, respectively. Group G2 describes a small semi-intensive system of raising crossbred cows with low milk production (12 litres
milk/day) and is comprised of managed-pasture and the use of a concentrate diet. Group G3 represents an extensive system of raising crossbred cows with low milk production (5 litres
milk/day) without supplementary feeding during winter, and without pasture management in the livestock feed programme. The authors concluded that group G2 should be promoted, considering its efficiency and environmental results.
The same study considered corn and soybean as the only sources in the feed programme, which resulted in a total environmental cost of the feed that increased from 11 to 22% and from 34 to 42% for semi-intensive and intensive dairy production systems, respectively. Therefore, to our knowledge and despite the importance of feed in livestock systems and according to Odum’s macroscope concepts, the feed contribution as a pathway to the development of a more sustainable dairy farm has not been adequately exploited by these authors. The use of feed produced inside the farm’s boundaries can improve sustainability due to its higher use of local renewable resources and, consequently, its higher renewability fractions. Thus, using local feed ingredient sources in a dairy feed programme (i.e., silage and pasture-based feed programmes) can increase the sustainability level of dairy farms. In addition, the integrated dairy–crop production system can enhance dairy farm sustainability, which can help keep activity viable for smallholders.
According to Robinson and Tinker [
18], the sustainability concept requires the (re)conciliations of the: (i) ecological imperative, regarding the biophysical carrying capacity of the planet; (ii) the economic imperative, to provide an adequate material standard of living for all; and (iii) the social imperative, to provide systems of governance that propagate the values that people want to live by. Although the socioeconomic aspects are inherently considered, the sustainability goals are not taken into great consideration in terms of the appropriation of natural resources and the necessary carrying capacity of a humane society [
19,
20]. In this context, considering emergy-based tools as an approach to assess carrying capacity can be an option if we are dealing with a problem of resource availability and its level of exploitation [
21]. Thus, an emergy- and economic-based decision approach can help to achieve the goal of sustainability by highlighting production systems with lower environmental costs and a higher standard of living for all (i.e., SDG 12 [
22]).
A few studies highlight the environmental and economic costs through assessing a holistic and integrated system view that aims to improve dairy farm sustainability [
23,
24]. Linking environmental and socioeconomic costs in a single perspective allows for the assessment of system sustainability from a holistic point of view, helping decision-makers to choose the most sustainable pathways in livestock systems; however, no studies have focused on the contribution of feed sources to the enhancement of dairy farm sustainability [
15,
25]. Given this, our study aimed to assess dairy farm sustainability from a holistic point-of-view by applying emergy and economic theory perspectives.
Furthermore, this study proposes an evolution from the study of Agostinho et al. [
12] which considers the livestock feeding programme in a farm with low-intensity milk production (similar to group G2 from Agostinho et al. [
12]). The novelties of this study were: (i) the study of the feed ingredients of the livestock feed programme by season (dry and rainy) using emergy synthesis; (ii) scenario simulations for intensive (INT), semi-intensive (SIS), and extensive (EXT) dairy systems from the livestock feed programme perspective; and (iii) a sustainability comparison of the scenarios using economic profits, efficiency, and the emergy sustainability index as an integrated graphical tool.
2. Materials and Methods
This study was approved by the Committee of Ethics in Use of Animals, School of Veterinary and Animal Science, University of São Paulo (FMVZ-USP), under protocol number 8143010221. Data were collected from August 2020 to January 2021 and then estimated for one year. A descriptive methodology was used for data collection. A case study was conducted using a semi-structured questionnaire and was applied in order to capture smallholder farmer’s perceptions. In addition, a complementary description of the observation, analysis, and interpretation of the data was used.
2.1. System Description and Primary Data Collection
The small Brazilian dairy production system (PROP) studied is located in Analândia, São Paulo, Brazil (22°06′20.07″ S, 47°39′58.75″ W). The predominant biome is the transition between the Brazilian Savannah and the Atlantic Rainforest. The climate is hot and tropical, with an annual precipitation of 1648 mm. The driest months are from April to September (32–81 mm). The property has a Cadastro Ambiental Rural (CAR; Rural Environmental Registry) designation and has legalised forest preservation areas. Additionally, the PROP was included in the Programa Nacional para Fortalecimento da Agricultura Familiar (PRONAF; National Program for Strengthening Family Agriculture) according to the criteria of Decree No. 3.991 [
26]. The farm presents a diversified production system, with dairy production as a main source of income. The agricultural land was divided into pasture area (20 ha), cassava area (2.8 ha), corn crop area (3 ha), and sugar cane crop area (2.7 ha) (
Figure 1). Other areas were not considered in this study.
The dairy zootechnical performance indexes for the PROP were: (i) lactating cows with an average milk production of 15 litres
milk/cow.day (183,600 litres
milk/yr); (ii) familial labour as the main labour source; (iii) low financial investment in production, characterised by semi-automatic dairy milking; and (iv) buildings and equipment with a high service time (
Table 1).
2.2. Emergy Synthesis Model Development
Emergy was proposed by Odum [
27] as being the whole energy needed (economic, anthropic, and environmental resources) to produce goods and services in emergy flows with a universal unit of measurement (solar emjoules; sej). This methodology considers the importance of nature in production processes. The methodology depicts the systems, including their driving forces and interactions [
28]. Emergy synthesis consisted of three steps: (i) design of the system diagram defining the temporal and spatial system boundaries and the input and output energy flows used in the process (
Figure 2); (ii) organisation and construction of tables for calculating emergy flows through transformities and renewability; and (iii) calculation of the emergy indices, followed by a discussion of the results for the system (
Table 2). The use of local renewable (R) and non-renewable (N) resources were considered, as well as inputs from the economy (F), considering the renewable (F
R) and non-renewable (F
N) fraction of each source (
Figure 2).
The transformities (Tr) of the Items listed in the calculation tables were obtained from the literature, except corn, cassava, sugar cane, and pasture grazing. The emergy synthesis of cassava, corn, and sugarcane was developed by considering the crop-to-silage process to obtain local transformity and renewability (%R). Detailed information on silage emergy synthesis is provided in
Supplementary Material S2.
The Tr and Unit Emergy Values (UEVs) were standardised using the Geobiosphere emergy baseline (GEB) proposed by Brown et al. [
32] in which the latest value for GEB is 12.0 × 10
24 seJ yr
−1. All detailed memory calculations for emergy synthesis are shown in
Supplementary Material S3.
2.3. Economic Cost Development Model
The economic cost model was developed according to economic theory. The costs were allocated as variable (Vc) or fixed operational costs (oFc). The total cost (Tc) was considered to be the sum of variable and operational fixed costs. The variable costs were feed (corn and soybean meal, urea, vitamin-mineral supplement, corn silage, cassava silage, and sugar cane silage) and veterinary expenses (veterinary products and vaccines). The costs of pasture management were included as feed costs. The fixed operational costs were labour, depreciation (buildings and equipment), maintenance (buildings and equipment), and other fixed costs (energy and fuel).
The average prices of ingredients used in diet formulation were obtained from market prices reported in São Paulo state from August 2020 to January 2021 and converted at the prevailing exchange rate on the date of recording (USD:BRL = 1:4.73). The average prices per kg of ingredients were as follows: corn meal, USD 0.24; soybean meal, USD 0.52; urea, USD 0.60; vitamin-mineral supplement, USD 0.85; salt, USD 0.17; and corn silage, USD 0.05. The average price per litre of milk was USD 0.47 [
33].
Thus, profitability follows according to Equation (1):
where P is the profitability (USD/litre
milk); Rv is the revenue (USD/litre
milk); Tc is the total cost for dairy production systems (USD/litre
milk); Vc is the variable costs that include all components involved in the activity that only occur if there is production, and that are directly related to milk production (USD/litre
milk) [
34]; and OFc is the operational fixed costs that represent the elements of expenses that are borne by the farmer, regardless of the production volume [
34].
2.4. Comparison between the Most Common Brazilian Feed Programmes: Proposed Scenarios
The proposed scenarios were developed to investigate the economic and environmental costs of different dairy feed programmes, considering the most common feed programmes for dairy production systems (
Figure 3).
INT (intensive dairy production system simulation) involves: feed concentrates (corn, soybean meal, minerals, urea, and salt) and corn silage are offered in the trough all year round. Intensive dairy production systems contribute 3% to Brazilian production (BNDES, 2018). This system can be pasture-intensive (2000 to 4500 litres
milk/cow.yr) or confinement-intensive (above 4500 litres
milk/cow.yr) [
35]. Production can reach 14,000 litres
milk/cow.yr [
36]. The increase in productivity is related to the use of technical knowledge and skills, a specialised herd, a concentrate to feed the animals, and the strict accounting control of production [
12]. This system requires significant investment.
SIS (semi-intensive dairy production system simulation) involves: feed concentrates (corn, soybean meal minerals, urea, and salt), access to pasture that occurs during the rainy season (October to April), and corn silage in the trough during the dry season (April to October). Semi-intensive systems are often small and medium farms and family farms, and Brazilian dairy production is composed of over 60% of family farmers [
37]. The herd’s feed depends on pasture areas, and forage supplementation is offered in periods of lower growth of tropical grass [
35].
EXT (extensive dairy production system simulation) involves: feed concentrates (corn, soybean meal, minerals, urea, and salt) and access to pasture that occurs all year round. According to data from IBGE [
38], 75% of Brazilian farmers use so-called extensive systems. Extensive systems are mostly small farms, especially in subsistence and family production. Cattle feed depends exclusively on pasture areas and supplementation only with white salt [
35]. In these systems, it is common to have low production from cows (less than 1200 litres
milk/cow.yr) and use more rustic breeds (high-blooded Zebu) [
35].
The diets of the proposed scenarios were formulated using the nutritional requirements and nutritive value of the food estimated by the NRC [
39] for lactating cows, according to the body weight and expected production of 15 litres
milk/cow.day to meet energy and protein requirements, as suggested by Santos et al. [
40]. From the formulation of the diet, the milk production of the scenarios was corrected by the NRC [
39] from the stable energy balance (demand = offered), in which PROP was 15 litres
milk/cow.day, INT was 20 litres
milk/cow.day, SIS was 20 litres
milk/cow.day, and EXT was 15 litres
milk/cow.day (
Table 3).
INT is the intensive dairy system scenario where the feed program is based on corn silage and protein concentrated feed ingredients in the whole year (365 days). SIS is the semi-intensive dairy system scenario where the feed program is provided according to the climate season features. In the dry season (from April to October), the feed program is based on concentrated feed ingredient intake and corn silage (180 days); in the rainy season (from November to May), the feed program is based on concentrated feed ingredient intake and pasture grazing (Urochloa brizantha; 180 days). EXT is the extensive dairy production system scenario where the feed program is based on pasture grazing (Urochloa brizantha) and protein concentrated feed ingredients during the whole year (365 days).
2.5. Interlink Decision Making Index: The Sustainability Evaluation under a Critical Criteria Perspective
The Interlink Decision Making Index (IDMI) is a multi-criteria decision-making tool that aims to simplify through less human interference in determining the best choice for a more sustainable pathway for any process. IDMI aims to integrate key criteria of different dimensions (environmental, economic, social, etc.), while allowing a few of these criteria to be chosen as critical criteria (CC) that are weighted to be more influential than the others in the decision-making process. IDMI proposes a comparison of the options of a particular case (i.e., new product, manufacturing process, or a construction process) under similar selection criteria [
41].
Despite having no unit or physical meaning, IDMI can be understood as something similar to the area or volume (or other geometrical properties) that is constructed by all criteria in which the critical criteria (CC) are more influential and decisive to the IDMI value [
41]. Mathematically, if there are n selection criteria in a decision-making case, the IDMI value area could be calculated as follows:
where Ci are the variables, C
n is the (value of the) first critical criteria (CC), and n is the number of selection criteria.
The logic behind Equations (2) and (3) is such that in Equation (2), all criteria contribute to IDMI in a similar way. In Equation (3), all criteria contribute to IDMI, but CC is critical and contributes more. Thus, for this study, the higher the value of IDMI for a particular system, the better the option may be [
41].
The criteria selected as CC must be directed towards more sensitive and contributing criteria in the IDMI value. The selection of criteria and CC can be done by a survey of managers or decision-makers in the industry or government or from the experience of the particular decision-making group [
41]. In this study, the selected criteria were ESI, GP, and P. The rules chosen for the CC were: no CC, ESI as CC, and P as CC. The objective of CC selection was to identify changes in the sustainability ranking among the proposed scenarios, as far as the variable elected as CC was modified.
Since the values obtained for each criterion have different unities and magnitudes, it is difficult to compare them directly. Thus, it is essential to make the value dimensionless before IDMI calculation [
41]. Index standardisation was defined by dividing the selected criteria by the reference values for each indicator. The reference values were selected according to the group G2 proposed by Agostinho et al. [
12]. The profitability from G2 was assumed to be the same value obtained for the SIS scenario.