Introduction

Extensive livestock production has a high environmental impact. This occurs mainly through deforestation and forest degradation, including landscape fragmentation, greenhouse gas (GHG) emissions, overuse and contamination of water and soil, introduction of non-native species, and loss of biodiversity and other ecosystem services (Herrero et al. 2009). Despite the widely recognized negative environmental impacts, the area of land devoted to cattle production is increasing worldwide, particularly in Latin America (LA), due to the global demand for beef (USDA 2019; Figueroa et al. 2022).

Mexico ranks seventh in the world in beef production and third in LA (behind Brazil and Argentina), and livestock production is an important contributor to the national economy (FIRCO 2018). Extensive livestock production under conventional management is the predominant production system, occupying 56% of the Mexican territory (109.84 million hectares) (Deschamps-Solórzano et al. 2020). The majority (64%) of beef is produced in ten states: Veracruz, Jalisco, San Luis Potosí, Baja California, Durango, Sinaloa, Chiapas, Michoacán, Chihuahua, and Sonora (SIAP 2020). The state of Campeche in the Yucatan Peninsula is ranked 25th and contributes with only 1% to the national beef production (SIAP 2020). Although technified commercial farms predominate in the most productive states, about 30% of bovine production in Mexico comes from small family farms, whose production is mainly for subsistence, with only sporadic local sales of surplus (Fuentealba and González-Esquivel 2016).

A livestock landscape is a mosaic composed of the key structural elements for livestock activity, including cattle, natural or man-made landforms, predominantly pasture, but also forest fragments, trees and other secondary vegetation, agricultural areas for forage production, barriers, and water reservoirs (Warren 2018; White et al. 2001). Conventional livestock landscapes are dominated by pastures and have only a few or no trees at all (Cárdenas et al. 2021). Silvopastoral (SP) livestock landscapes besides pasture also include herbaceous or perennial plants elements and can be classified as traditional and improved. The traditional SP landscape elements include live fences, windbreaks, and scattered trees among others, that are developed by local farmers using locally available resources, whereas the improved SP landscape elements, mainly protein banks and alley grazing, are developed by experts and implemented locally (Pezo and Ibrahim 1998; Fuentealba and González-Esquivel 2016). The SP landscape elements serve as a safeguard for native forest species and increase landscape connectivity (Cárdenas et al. 2021). The size, structure, and spatial distribution of SP elements in a livestock landscape is determined by the ranchers, whose decisions depend on the availability of land and water, type of forage, number of cattle, local history, government support, social dynamics, and socioeconomic characteristics (Manjarrez Muñoz et al. 2007).

Given that each type of SP landscape element is associated with different productive and environmental benefits, their heterogeneity is important. Only when several livestock landscape elements are present in the same local landscape and these are integrally managed, we can talk about SP systems (Murgueitio et al. 2014). Therefore, SP systems can reduce emissions and absorb GHG, increase production in less extensive areas, generate complementary income from the marketing of timber and non-timber products, and thus reduce the vulnerability of the productive system (Murgueitio et al. 2014).

Under the dominant scheme of extensive conventional cattle ranching, SP landscape elements are little valued in Mexico (Fuentealba and González-Esquivel 2016) and there is a tendency to decrease forest components in favor of crops and other non-forest land covers. To revert this trend, it is important to identify local practices that result in SP landscape elements and explore the reasons and limitations for their adoption. The existing studies in Mexico describe the SP practices adopted by local population either through external initiatives (Cancino et al. 2021 in Chiapas) or resulting from the adaptive management to local biophysical and socioeconomic conditions (Sánchez-Romero et al. 2021 in Jalisco).

The few studies investigating the livestock landscapes and SP systems in the state of Campeche identify livestock as the cause of deforestation through pasture establishment, despite the potential use of local tree and shrubs as fodder (Warren 2018; Alayón-Gamboa et al. 2022). Hopelchén is the fifth municipality as per total livestock production area in Campeche (BioPaSOS n.d.) and forms part of the buffer zone of the Calakmul Biosphere Reserve (CBR). As in the rest of Campeche, cattle ranching in Hopelchén was promoted by the several federal government programs since the 1970s (e.g., PROCAMPO and “Alianza para el campo”)Footnote 1. Since then, the number of livestock breeders has been steadily increasing (Porter-Bolland et al. 2008), supported by the federal government’s Livestock Productivity Incentive Program (PROGAN) (2003–2019) (Špirić and Ramírez 2021). This activity plays an important role in the local livelihoods’ security, as it is a form of wealth savings and a part of local culture (Porter-Bolland et al. 2008). Although mechanized agriculture (maize, soya, and sorghum) is the main land use change in the municipality, deforestation caused by pasture expansion is also one of the central socio-environmental problems (Ellis et al. 2018).

The objective of this article is to identify the livestock management practices and analyze the resulting landscape elements, both conventional and SP (traditional and improved), in the municipality of Hopelchén and describe their impacts on forest cover. The results are an initial description of the local livestock production cycle composed of four phases: (1) Pasture and infrastructure establishment, (2) Production, (3) Pasture and infrastructure maintenance, and (4) Fallow and abandonment.

Methods

Study area

The municipality of Hopelchén is located between 89° 06′ and 90° 09′ longitude and 17° 48′ and 20° 11′ latitude and it is the third biggest municipality in Campeche as per land area (747,900 ha) (INEGI 2009). It is inhabited by 37,777 people (INEGI 2010). The dominant type of land tenure is the Maya indigenous people ejidosFootnote 2 (40), which occupy 64% of the municipality’s territory. Hopelchén has a karst topography with no superficial water flows. Its climate is subhumid warm tropical, with a mean annual temperature of 26 °C and a marked rainy season (May–October), with the occurrence of hurricanes (Porter-Bolland et al. 2008). The dominant land cover is the mature tropical dry forest (monte alto) and secondary forest in various stages of succession (acahual) (84.6%), followed by agriculture and pasture (15%), and urban areas and water reservoirs (0.4%) (Špirić and Ramírez 2022).

The southern part includes the region called La Montaña (473,336 ha), located between 89° 10′ 10″ and 89° 25′ 07″ longitude and 19° 14′ 45″ and 18° 59′ 48″ latitude (Porter-Bolland et al. 2007). La Montaña is covered by a mature forest rich in timber and melliferous trees, which partly lies within the buffer zone of the CBR. La Montaña has the largest number of cattle (5908 head) and pasture area (12,159 ha) (Ellis et al. 2018). Based on local typology and use, there are five main soil types in La Montaña: (1) akalche’—periodically flooded lowlands with organic matter, supporting savannas with natural grasses and induced pasture; (2) kankab—plain terrain with deep red soils, preferred for mechanized agriculture; (3) ya’x hom—plain terrain with dark soils, rich in organic matter, used for milpa (slash-and-burn cultivation of maize, squash, and beans) and pasture; (4) tsekel—dark, rocky, shallow soils, poor in organic matter, used for pasture; and (5) kakab or ramonales—very fertile soils with a deep organic layer over a limestone, valued for milpa and as a source of fodder trees (Porter-Bolland et al. 2009).

Our study area includes La Montaña and the central region, which together comprise 21 ejidos, while the number of cattle reaches 11,240 heads that are raised on 21,764 ha of pasture (Ellis et al. 2018). Our study does not include the northern region, where mechanized agriculture is the dominant land use, originally promoted in the 1980s by the Mennonites, a group of ethnic-religious migrants (Ellis et al. 2017).

Data collection and analysis

Our sample design combined convenience and purposive strategies (Newing 2011), aiming to include at least one livestock breeder from each Maya ejido in our study area. We approached the livestock breeders that were recommended by local key informants and authorities. We conducted 28 semi-structured interviews (February–October 2022) with producers from 14 ejidos (out of 21)Footnote 3 and visited seven cattle breeding ranches that were easily accessible.

Most of the interviewees were ejidatarios (22), followed by persons without agrarian rights (locally known as comuneros) (4), and private landowners (2). The average age of the interviewees was 55. Only two were women. Livestock was the main productive and income activity, accounting for half of the cases, followed by beekeeping (36%) and mechanized agriculture (14%). Most respondents (25) combine all three productive activities.

The interview guide consisted of two parts: (1) Structured questions about productive activities focusing on livestock (e.g., number of heads, hectares of pasture, occupation time), and (2) Open-ended questions on livestock production practices, focusing on the configuration and use of forest fragments and tree elements. The interviews were semi-structured so that emerging themes could be explored during the conversation (Newing 2011). They were conducted in Spanish, with occasional translation from Mayan by a field assistant. All interviews were recorded and transcribed with prior consent.

The quantitative information was organized in Excel (Microsoft) and analyzed using descriptive statistics. We also calculated the stocking rate (number of heads per hectare of pasture), and the recovery period (number of paddocks at rest multiplied by the occupation period in days). The scientific names of tree, shrub, and pasture species were identified using the local names provided by the interviewees and by consulting local flora catalogs (Porter-Bolland et al. 2009; Duno de Stefano et al. 2010).

The qualitative content, including the description of the activities and processes performed, and the reasons for them, was coded in MAXQDA 2022 (VERBI Software 2021). The codes were organized into four categories corresponding to the phases of livestock production. Three levels of subcodes were derived from each code (see Online Appendix 1. System of codes). A qualitative interpretative analysis was performed, which combines a classical qualitative content analysis that answers the question ‘what’ with an interpretative step that answers ‘how’ and ‘why’ (Wiesner 2022).

We used the “Retrieved Segments” MAXQDA window to group the same code or subcode segments in individual documents. The qualitative information in the documents was analyzed as a whole and served to describe each production phase, including the activities and processes (‘what’) performed in each phase of the production cycle (codes). In the second part, corresponding to the interpretative step, we identified the material and measures (‘how’) used in the activities and processes, and reasons and initiatives (‘why’) for this (subcodes). Finally, using the “Quantitizing” MAXQDA function, the subcodes of the second and third levels were transformed into document variables of Boolean type, and their absolute frequencies in the whole sample were calculated.

Results

Phase 1: pasture and infrastructure establishment

Area selection for pasture establishment

Establishing a ranch begins with choosing a location. The producers prefer a heterogeneous terrain with different soil types to ensure good pasture development on the plains (ya’x hom), water retention on the lowlands (akalche’), and shelter for the cattle on the hills (tsekel and kakab) in case of floods. Another criterion is proximity to the village (17 ranches), which reduces the cost of transporting water and fodder during the dry season (see Online Appendix 2. Calendar). However, available land areas close to villages is scarce and producers are forced to use remote areas (5 ranches), with those accessible by vehicle throughout the year being the next to be occupied. Additionally, four ranches are near to villages but not easily accessible, and two producers have paddocks both near and far from the village.

Forest clearance and pasture establishment

In most of cases, acahuales formerly used for agriculture have been cleared for pasture, while mature forest (monte alto) has been preserved. This is because the vast majority (89%) of our respondents are also beekeepers, and the mature forest is an important source of nectar. In addition, because this forest can be used for timber harvesting or to participate in environmental programs, there are restrictions on logging in monte alto by local and federal authorities. Only those producers who established their pasture more than 30 years ago have cleared mature forest.

The clearing of the forest is done manually, and the land is prepared by the traditional practice of slash-and-burn (roza, tumba y quema). When small areas (1–4 ha) of forest are cleared, they are generally used for milpa for 2 years before pasture is established, unlike larger areas (6–8 ha) where pasture is established from the first year. In this study it was found that forest clearance of 10 ha has been the upper limit that a single producer could open each year. The reason for this is  the high costs of clearing and because larger areas can be detected from the air by environmental authorities, especially in the ejidos located in the CBR buffer zone. In the past forest clearance was done manually with axes and machetes, more recently with chainsaws, but it is also becoming common to hire caterpillars or backhoes owned by Mennonites. However, this is only affordable for easily accessible large areas (at least 20 ha).

Forest components

During forest clearance, ranchers decide on which elements of the forest landscape to maintain and their spatial configuration. Most interviewees (76%) maintain mature trees scattered in the pasture, but as observed during the field visits, these are usually present at a very low density (average of 6 trees/ha). Most of these trees were cleared of shrubs and herbaceous plants around them (roza) when they were left standing during manual clearing (tumba), which should have helped them to survive the initial burning (quema). However, selective tree preservation is difficult when unwieldy agricultural machinery is used for the initial forest clearance. The regrowth of scattered trees was less common (35% of the ranches). The main function of these trees is to provide shade, but they are also used as fodder for livestock or honeybees, and as construction material (Fig. 1).

Fig. 1
figure 1

Uses of trees scattered in the pasture: A shade (ranch 2; Mar 2022), B poles (ranch 2; Mar 2022), C fodder (ranch 4; Mar 2022), and D selectively preserved regenerated trees (ranch 1; Feb 2022)

Full size image

The preferred tree species for shade are those with a broad crown, such as: pich (Enterolobium cyclocarpum), beek or roble (Ehretia tinifolia L.), puk’te’ (Bucida buceras L.), tsalam (Lysiloma latisiliquum (L.) Benth.), ja’bin (Piscidia piscipula (L.) Sarg.), ya’ or zapote (Manilkara zapota (L.) P. Royen), and pixoy (Guazuma ulmifolia Lam.).

Forest strips on exterior boundaries are very common (89% ranches). They are mainly used to separate two ranches, to prevent fire from spreading to neighboring pastures or damaging of the barbwire fences placed in the middle of them, but they are also a source of material, shade, and fodder. Also, a high number of ranchers (70%) reported having forest strips with fences on the interior boundaries. The main functions of interior forest strips are to divide individual paddocks within the same ranch that are used in rotation by the same herd, to prevent the spread of fire or damage to barbwire, and to provide shaded corridors for cattle movement between paddocks (Fig. 2).

Fig. 2
figure 2

Forest strips: A with fence on exterior border (ranch 1; Feb 2022), and B shaded cattle corridor in the interior division (ranch 4; Mar 2022)

Full size image

The width of the forest strips ranges from 20 to 50 m. However, the strips located on the prevailing wind side are gradually shrinking because they are repeatedly damaged by fire, even when firebreaks are established (see “Phase 3: pasture and infrastructure maintenance” section. Pasture maintenance). Additionally, the interviewees are aware of negative practices that occur in the study area, such as the clearing of the exterior strips to facilitate access for agricultural machinery and to reduce the shading effect on crops; and the clearing of interior forest strips to establish some additional pasture.

Infrastructure

Most ranchers use fencing that combines dead wooden posts with live trees present in forest strips, both on the exterior (74% of ranches) and interior (63%) boundaries. Typically, producers prefer to use only live trees in a straight line 2 m apart to ensure that 3–4 barbwire strips are tensed and resistant. Dead posts are considered more stable against strong winds than live trees. They also tend to absorb barbwire as they grow, although this can be avoided by using plastic bottle patches as insulation. Only two producers prefer to use live trees because they do not rot. Although none of the interviewees have exclusively live fences, five producers are growing or have planted the following fast-growing trees: chakah or palo mulato (Bursera simaruba (L.) Sarg.), balche’ or cocoite (Gliricidia sepium (Jacq.) Kunth ex Walp.), chak mo’ol che’ (Erythrina standleyana Krukoff), pomol che’ (Jatropha gaumeri Greenm.), and teka (Tectona grandis) (Fig. 3).

Fig. 3
figure 3

Interior pasture divisions: A with dead fence (ranch 6; Oct 2022), and B live fence with growing trees (ranch 3; Mar 2022)

Full size image

In addition to fencing poles, impacts on forests during the establishment phase occur through the extraction of hardwoods for stockyards, which can be made of wooden boards (21 producers), wooden logs (4), or only with barbwire (3) (Fig. 4).

Fig. 4
figure 4

Stockyards: A made of wooden boards (ranch 6; Oct 2022), and B combined with vertical logs (tulum) (ranch 5; Mar 2022)

Full size image

The preferred tree species used for boards, which should be resistant to pests and moisture, are different from those used for pole, which should not rot when buried in the ground (Table 1). In both cases, according to our respondents, the timber should be cut during the full moon when there is less moisture and more resin in the wood. Other good practices mentioned by ranchers are to cut straight trees that can provide maximum number of poles and boards of preferred height (2.60 and 2.20 m, respectively), and to harvest from several areas to avoid overexploitation.

Table 1 Tree species used to build infrastructure and absolute frequency of their mentions (the highest three frequencies for each category of use are shown in bold)
Full size table

The availability of hardwoods is not the same for every producer and depends on the local vegetation type, the internal rules of the ejido, and the personal decision to preserve forest fragments (reservas). More than a half of our interviewees (16) have forest fragments that serve them mainly as a source of timber (10 ranchers), and less as a source of nectar for honeybees (3) and fodder for cattle (2), as a refuge for wildlife (2), and as shade (1). Since it is expensive to fence them, these forest elements are not used for free browsing. Forests are fenced only when they are used as shelters for cattle during the extreme floods,Footnote 4 but these areas are usually located in higher zones that are converted to pasture the next season (see Online Appendix 2. Calendar). As hardwood have become scarce, producers have started to remove cut timber before the initial burn to use or sell it, which was not common practice a few years ago. This also reduces the amount of biomass to be burned, protecting the soil from degradation and cows from stepping on unburned logs.

Phase 2: production

Producers’ characteristics

Most of our sample includes medium producersFootnote 5 (60%) owning between 12 and 40 heads, followed by small producers (20%), between 4 and 11, and large producers (20%), between 50 and 90. The average number of animals per producer was 30, while the average of pasture was 40 ha (ranging between 6 and 100 ha).

Most producers (17) are within the optimum stocking rate (1 head/ha).Footnote 6 However, on average per group, small producers have a suboptimal stocking rate (0.39 head/ha), while medium (1.13 head/ha) and large (1.41 head/ha) producers exceed the optimal number. Those with a low stocking rate (as low as 0.11 head/ha), including all small and nine medium ranches, are in the early stages of establishing or recovering their ranches and have pasture but no or few cattle. Some of the ranchers with low stocking rate rent their paddocks, while others avoid this practice because of the risk of degradation of the pasture by the renter. Only four producers have a high stocking rate (up to 5 head/ha), including those who lost their pasture to flooding but were able to save their cattle, which they now have in smaller areas.

Forage and fodder type

All interviewees have induced pasture as their main source of fodder and grow an average of three different improved seeds. Brizantha (Urochloa brizantha) (22 producers) and Tanzania (Panicum maximum cv. Tanzania) (19) are the most used in the hilly areas, Humidicola (Brachiaria humidicola) (5) in low flooded areas, while Alicia (Cynodon nlemfuensis) (6) can be found in both areas. Only two persons cut and carry pasture from other areas, but only one of them has cut grass that can be preserved as silage and used during the dry season. Two producers use natural grown pasture alambrillo (Sporobolus indicus (L.) R. Br.).

Most producers (22) use fodder trees, but only during the dry season. They do so by allowing free browsing (Fig. 5) and by pruning the branches to make them available for cattle, while only few producers cut and carry fodder from other areas (Table 2).

Fig. 5
figure 5

Fodder trees: A low forest vegetation (akalche’) surrounding the local road, freely browsed by cattle (ranch 5; Mar 2022), and B paddock with waaxim shrubs during the recovery period (ranch 3; Mar 2022)

Full size image
Table 2 Tree and shrub species used as fodder
Full size table

Only those producers (4) participating in the SP initiative implemented by the local NGO Pronatura PY have planted waaxim or guaje (Leucaena leucocephala) in alleys. Although the producers who do not use fodder trees (6) are proud of the quality and quantity of their pasture, they buy bales of straw or use the rest of the maize harvest to supplement their cattle’s diet during the dry season. Additionally, half of the producers graze their cows on crop residues on nearby agricultural land.

Pasture management

All the interviewees claim to practice rotational grazing, however three of them have only two plots and therefore practice alternate grazing. Producers have an average of four pasture lots. The average size of each lot is 10.4 ha, but ranges from only 0.7 ha in the case of two producers with intensive grazing with electric fences, to up to 31.7 ha in the case of one large private landowner. Only four producers have a recovery period within the optimal range of 28–45 days,Footnote 7 including two producers with intensive rotation and one with alternate grazing. Most producers are well above the upper limit, mainly because they plan to increase the number of cows.

Water sources

Most of the producers (85%) depend on communal potable water networks as their main (7 producers) or complementary water source, only during the dry season (16 producers). The second group combines it with built (jagüeyes) or natural (aguadas) open water reservoirs that are filled during the rainy season. Only those producers who could afford to dig a deep well (4) or several jagüeyes (1) have theme as their only source of water throughout the year. To prevent evaporation several interviewees reported keeping or planting trees around water reservoirs and placing drinking troughs in the shade (Fig. 6).

Fig. 6
figure 6

Vegetation around water reservoirs: A and B jagüeyes surrounded by low forest (ranch 2; Mar 2022 and ranch 6; Oct 2022, respectively)

Full size image

Phase 3: pasture and infrastructure maintenance

Manual weeding of paddocks for pasture maintenance is the most common practice used by ranchers in our sample (96%), followed by fire (81%) and pesticides (56%). Manual weeding does not damage soil quality or surrounding vegetation and allows for selective pruning of regenerating trees. However, only four interviewees use manual weeding exclusively, either because the use of fire is prohibited by the ejido or by the local SP initiative. However, the manual weed control with machetes (13 producers) or brush cutters (7) is labor-intensive, and most producers have to pay extra workers. The pesticides are used primarily to control the growth of invasive or thorny tree and shrub species, such as tajonal (Viguiera dentata (Cav.) Spreng.), x-peres (Croton guatemalensis Lotsy), chimay (Acacia pennatula (Schltdl. & Cham.)), chukum (Havardia albicans (Kunth) Britton & Rose), and kaatsim (Mimosa bahamensis Benth.).

Only two ranchers fertilize their pastures. Most of the producers consider the fire as essential for pasture regeneration and burn their paddocks annually or every 2 years. In their opinion, if local fire management practices are applied, fire does not pose a threat to the surrounding forest elements. These practices include opening of firebreaks, constant vigilance, and burning against the wind, before the dry season (see Online Appendix 2. Calendar) and during the night. This also ensures that no serious damage is done to the infrastructure and that annual repairs are kept to a minimum.

Phase 4: fallow and abandonment

Because of the high investments, paddocks are rarely abandoned intentionally. This occurs during a prolonged drought or when the herd is reduced due to lack of labor, illness or other family costs. However, it is common to maintain the ‘cattle-less’ pasture to resume production without having to cover the establishment costs. However, if weeds take over the pasture, it is left 5–6 years before it is cleared to allow for secondary vegetation regeneration (jubche’ in Mayan) to suppress the weeds. The degradation of pasture due to overgrazing is another reason to leave it fallow. However, according to ranchers, this rarely happens because they respect the maximum stocking rate and rotate. Pasture can also be lost to fire, in which case it is left to regenerate on its own, or to flooding, in which case it must be reseeded. Due to the high costs of repairing infrastructure and purchasing new livestock, the pastures in the ejido most severely affected by the 2020 floods have been converted to mechanized agriculture.

In summary, Fig. 7 shows the frequency of traditional and improved SP elements in the local livestock landscapes, while Table 3 shows the practices that lead to their formation and maintenance, and relates them to forest cover outcomes.

Fig. 7
figure 7

Frequency of traditional and improved SP elements in the local livestock landscapes in Hopelchén

Full size image
Table 3 Cattle managing local practices and resulting livestock landscape elements and forest cover outcomes
Full size table

Discussion and conclusions

We identified six traditional and one improved SP elements in local livestock landscapes in the municipality of Hopelchén (Fig. 7). The configuration and use of forest and tree elements resulting from practices employed by local ranchers in each phase of the livestock production cycle leads to either positive or negative forest cover outcomes.

The forest strips combined with live or dead fences are the most found among traditional SP elements. These are principally present at exterior borders and serve for division and protection of built infrastructure against fire, differently from classical windbreaks described in the literature (Pezo and Ibrahim 1998). The stand-alone fences are found on interior division of big paddocks where interior strips are missing. The presence of interior forest strips between the paddocks of same size can be an indicator of a gradual and planned ranch development. Although live fences are economically cheaper than built fences, these are the least commonly found tree configuration and use because ranchers have negative perceptions of the difficulty of their establishment and the intensity of repair (Morantes-Toloza and Renjifo 2018).

While ranchers open firebreaks to protect other forest elements, no direct measures to protect scattered trees were reported. Although producers are knowledgeable of the importance of shade for animal wellbeing, they have negative perceptions on its impact on grass and keep trees in pasture at low density. Nevertheless, it is found that a medium canopy density has positive effect on pasture yields in tropics (Hernández-Salmerón and Holmgren 2022), while even at low density trees enhance soil infiltrability in flooded pastures (Cruz-Alonso et al. 2022).

Although the livestock breeders in Hopelchén use more tree species than those described in a similar study in neighboring area (Warren 2018), the principal use is as material, which was the main reason for conserving forest fragments. Despite the fodder species richness in the study area (Alayón-Gamboa et al. 2022; Santillán-Fernández et al. 2020), induced pastures dominate the cattle nutrition. No protein banks were identified, while naturally present trees and shrubs are used only in dry season. The deliberate pruning and cut and carry of fodder need a large amount of labor (Harvey et al. 2005). Accordingly, rather than uncontrolled free browsing and collection of cattle fodder (Gao et al. 2020), the extraction of timber for material and damage from pasture maintenance fires are more relevant sources of forest degradation in the study area.

Beyond the initial deforestation and fragmentation during the establishment phase and the conversion of forest to pasture, livestock production can have a reduced negative impact on forest cover, at least at the local landscape level, if decisions are made to maintain SP elements and practices are performed to improve their quantity and quality. Some of these practices can easily be implemented, such as use of plastic patches in live fences. Other practices are more economically and labor demanding, such as building water reservoirs as drought coping strategy to assure enough water and forage in principal work areas (Sánchez-Romero et al. 2021) and avoid deforestation of new areas.

However, for the positive impacts at the local livestock landscape level to be visible at the municipal level, where deforestation and degradation for pasture establishment and maintenance remains one of the main drivers, policies and initiatives should promote the SP systems that include multiple SP elements under integrated management. Mexico’s national REDD+ strategy (ENAREDD+) recognizes the potential of the SP systems to reduce deforestation and forest degradation (Murgueitio et al. 2014), and 17% of the REDD+ Special Program (2010–2014) investment in Campeche was invested in “innovative production systems”, including SP systems (Deschamps Ramírez et al., 2015). Currently the only SP initiative identified in the municipality of Hopelchén is promoted by local NGO Pronatura PY. The government continues to support the conventional livestock, such as the federal loan program “Crédito Ganadero a la Palabra”, which aims to improve the productivity of small ranchers by providing them with some additional animals (DOF 2019).

Although several forest elements are present in local livestock landscapes in Hopelchén and these perform certain connectivity function, their full potential is not reached, because their protective and construction use dominates. To prevent degradation and deforestation of additional areas their use as fodder and live fences should be enhanced. The full potential of forest and tree components could be reached, as presented in scientific literature from other study areas, when there is an optimal shade density by scattered trees, and forest and tree species are used beyond just protection or construction material, especially as a source of fodder (Nahed-Toral et al. 2013). However, the potential usefulness of local forest vegetation as a fodder depends on its structure and diversity, which influences which plants are likely to be selected and consumed by cattle (Albores-Moreno et al. 2021).

Further research is needed to determine the optimal conditions for reaching the full potential of the forest and tree components of SP systems in the study area. To inform future sustainable livestock policies and other initiatives, we would need to analyze a bigger sample to understand how producer and ejido characteristics impact on the choice of practices and the spatial composition and organization of the local livestock landscapes. It would equally be important to fill the research gap and evaluate other forest process, beyond deforestation, resulting from each phase of livestock production in the study area.