Secondary Succession in Fallow Agroforestry Systems Managed in Tropical Dry Forest in Western Mexico

Abstract

Tropical dry forests (TDFs) are ecosystems of high biocultural value, in which agroforestry systems (AFSs) have been essential in their management and conservation. We aimed to characterize agroforestry practices and analyze their capacity to conserve perennial plant diversity. In addition, we sought to evaluate how the management of TDFs as AFSs, together with their regeneration, influences species diversity and vegetation structure in a landscape with AFSs and TDFs in different conservation states. We compared the species diversity and basal area (BA) of plants in active and fallow AFSs at different regeneration stages in Zacualpan, Colima, Mexico. We found that AFSs harbored 71% of species richness (⁰D), forming a mosaic that contributed to the gamma diversity (124 species) of TDFs in the area. AFSs supported 23 endemic and 12 protected species. TDFs, active and advanced regeneration AFSs, had the highest number of useful species and diversity. Species richness (⁰D) in management categories increased as succession progressed, but not the BA, possibly due to frequent browsing and wood and firewood extraction. However, BA may be related to the management of useful trees maintained through agroforestry practices. We suggest increasing the matrix quality through a mosaic of active and fallow AFSs to promote ecological connectivity and biodiversity conservation.

1. Introduction

Throughout history, the management of natural vegetation to meet human needs has been a major driver of tropical forests change [1,2,3]. Among these, tropical dry forests (TDFs) are one of the most exploited ecosystems in the Neotropics for traditional food, medicine, and other subsistence resources as well as for agriculture [4]. However, the long-term human presence in Neotropical dry forests has generated biocultural wealth but also led to significant impacts from the expansion of intensive agriculture, cattle ranching, and urbanization [4,5]. Despite its ecological and cultural importance, between 1980 and 2000, Latin America had one of the highest deforestation rates of TDFs globally. For instance, areas of TDF in Mexico had some of the most densely populated human settlements worldwide as well as fire occurrence [6].

In such context, secondary forests have expanded; these are important reservoirs for tropical biodiversity and carbon stocks, and they provide essential ecosystem contributions; therefore, analyzing their recovery is a priority for research [7,8,9]. Over time, the species diversity and composition, as well as vegetation structure of these forests, change as ecological succession progresses [10]. Ecologists have accounted for several key indicators of forest recovery, including a synergistic increase in species diversity and basal area, leaf litter biomass accumulation, and the recovery of soil carbon pools [11,12]. However, ecological succession is driven and regulated by numerous factors at different scales (from patches to landscape) and organization levels (from species to socioecological systems) [9].

Studies on Mexican TDFs at the landscape scale have revealed that sites surrounded by higher forest cover exhibit faster recovery in species richness [13]. At both species and ecosystem levels, disturbances caused by land use change, such as vegetation removal through slashing and burning, have been identified to greatly influence the speed of vegetation recovery [14,15,16]. In addition, factors such as water and nutrient availability, the presence of plant propagules in the soil [17], temperature, relative humidity [18], and seed dispersal dynamics [19] are also crucial in successional processes. At the socioecological system level, management practices during the agricultural phase exert significant effects on successional pathways [9]. For instance, livestock browsing [20], the selective extraction of vegetation components [21], the gathering of sexual reproductive parts of plants (seeds, fruits, and flowers), the promotion and protection of useful species through local agreements, and the enrichment of socioecological systems may have a large influence on the way vegetation recovers [1,22,23].

Old-growth and secondary TDFs exhibit distinctive characteristics that have benefited humans for a long time. These include a rich species diversity that provides resources even during the dry season, which promotes people’s motivation for their procurement [24]. Furthermore, the presence of low trees that can resprout and withstand prolonged droughts [25] is also valuable as well, as that is easier to slash and burn after clearing [4]. Consequently, TDFs have been scenarios for the in situ management of perennial and annual plants, promoting the development of desired phenotypes through forestry and agricultural practices, thus facilitating the domestication of some species and landscapes [1,22,26]. In these management strategies, agroforestry systems (AFSs) play a central role integrating silvicultural and agricultural techniques that allow the conservation and shaping of essential components of Mexican agrobiodiversity [27,28,29,30].

AFSs are configured by agroforestry practices that integrate management units by influencing the spatial arrangement of vegetative elements such as live fences; vegetation fringes; remnants, patches, and islands of vegetation; and scattered trees [27,28,29,30]. Therefore, AFSs involve the management of both wild and domesticated plants and other organisms within and around cultivated fields [27,29], offering benefits to the people who manage them by providing a diversity of products for subsistence and local economies [27,28,29,30]. By maintaining native species, AFSs contribute to the conservation of regional biodiversity [27,28,29,30,31,32] and promote connectivity among populations, communities, and ecological processes [31,33,34]. Given these attributes, AFSs have considerable potential to contribute to ecosystem regeneration in areas that were previously destined for agriculture and other disturbed environments [33,34,35].

Humans and regrowing vegetation form a socioecological system, resulting in a dynamic mosaic of patches where agriculture and regrowth alternate in time and space. Thus, management practices used during the agricultural phase leave lasting legacies on vegetation succession [9]. AFSs in the regeneration phase often involve long-term agroforestry practices where fallow management includes planting, protection, fertilization, weeding, and the shade control of beneficial species. These practices vary in intensity as long as secondary succession progresses [36]. This phenomenon has been studied especially in tropical forests, such as Amazonian swidden fallows [3,36], Indonesian agroforests [37], Mayan-enriched fallows [38], and the Nahua-Totonaca Kuauhtakiloyan [39]. Therefore, a comprehensive understanding of how agroforestry practices and the management of TDF’s secondary vegetation influence ecological succession is critical for their conservation and may be applied in sustainable restoration efforts based on biocultural approaches like enriched fallows and productive restoration.

In this study, we aimed to characterize agroforestry practices and analyze their capacity to conserve perennial plant diversity. In addition, we sought to evaluate how the use and management of TDFs as AFSs, along with their subsequent natural regeneration, influence species diversity and vegetation structure in the context of a landscape with AFSs and TDFs in different conservation states. To accomplish this, we designed a study in the Nahua community of Zacualpan, Colima, Mexico, which included parcels currently used as AFSs with milpa cultivation (an agricultural system in which maize is grown in association with squash and/or beans and other crops) as well as fallow AFS parcels with different stages of regeneration after milpa cultivation has ceased.

2. Materials and Methods

2.1. Study Site

The Nahua community of Zacualpan is located in the municipality of Comala, Colima, Mexico. The area is part of the Nevado de Colima-Sierra de Manantlán Biological Corridor and the Central–Western Mexico Biocultural Corridor (Figure 1). According to the Köppen climate classification system modified by García [40], the climate in Comala is A(C)wo(w) with a mean annual temperature of 21.5 °C, a total annual precipitation of 761 mm, rainfall from June to November, and a pronounced dry season. The main vegetation type in the area is TDF with the presence of tall tree species favored by seasonal water bodies, including the outstanding the Armería River [25].

The population of Zacualpan is of Nahua origin, has communal land tenure, and is a community that was declared mining free in 2013 due to the environmental threats posed by a gold, silver, copper, and manganese open-pit mining operation in its territory [41]. As a result of the mining conflict, the community has defended its collective rights to a healthy environment, including the Ojo de Agua spring, which provides water to the community and the cities of Colima and Villa de Álvarez. Landscape components include cultivated fields with and without agroforestry management, Hog Plum (Spondias purpurea) orchards, home gardens, forest and fallow lands, riparian habitats, and an urban settlement with approximately 2000 inhabitants. The community is mainly engaged in rainfed agriculture, focusing on the milpa system, growing maize (Zea mays) and squash (Cucurbita spp.) as main crops, while also maintaining the use of both wild and cultivated native species for subsistence and local trade [42,43].

2.2. Research Design

To conduct our study, we obtained permission from the community authorities and contacted community members to locate and select AFS parcels with milpa cultivation and AFSs with different stages of fallow. The selected parcels were characterized by having been used for non-mechanized rainfed agriculture and having more than two agroforestry practices. The owners of the sampled parcels were interviewed to document the history, use and management of the parcels and the perennial plants occurring on them. The interviews were recorded and transcribed for qualitative analysis. This work summarizes the most meaningful information that is pertinent for the issues analyzed. Ethnobotanical details of in-depth interviews will be published elsewhere.

We sampled the vegetation in active AFSs and fallow AFSs at different stages of regeneration to characterize agroforestry practices and assess their capacity for conserving and recovering the diversity of native perennial flora species. We also evaluated how the use and management of TDFs as AFSs, along with their subsequent natural regeneration, influences species diversity and vegetation structure. In addition, we sampled TDF areas as references to identify regeneration pathways and diversity change trends using species diversity and basal area as indicators [12]. Sampling parcels were selected according to their similarity in topographic and geomorphologic conditions as follows: (1) located on slopes parallel to the Armería River, (2) at elevations between 500 and 650 m a.s.l., and (3) facing north. The sampling was carried out in different phases during the period from October 2021 to December 2022.

Vegetation Sampling

The characterization of the vegetation structure in areas under agroforestry management was carried out following the classification of agroforestry practices and the sampling methods developed by Moreno-Calles et al. [27] and Rendón-Sandoval et al. [28]. In each sampled parcel, we established five 10 m × 10 m squares separated by a minimum distance of 20 m (a total of 500 m² per parcel). The 15 sampled parcels covered a total of 0.75 ha and were classified into the following management categories:

(1) Forest: We included three plots in TDF areas, which according to the community members were protected areas used only for livestock browsing. However, some elders in the community informed us that these forest areas were cleared for agriculture about 50 years ago. Sampling squares were placed randomly in these areas. (2) Active AFS: We sampled three parcels actually cultivated with milpa within which sampling squares were placed to cover agroforestry management practices, including the isolated trees. Fallow AFSs were classified into three regeneration stages according to the information provided by the parcel managers as follows: (3) Recent: Three parcels in recent fallow (2, 5, and 5 years, respectively). (4) Intermediate: Three parcels in intermediate fallow period (7, 10, and 10 years, respectively). (5) Advanced: Three parcels in advanced fallow period (15, 20, and 25 years, respectively). In each parcel of all regeneration categories, three sampling squares covered the detected agroforestry management practices, and two squares were randomly located.

In each sampled parcel, we measured the diameter at breast height (DBH) of plants over 1 cm in diameter and the height of perennial species (trees, shrubs, woody vines, and columnar cacti) rooted within the squares. Herbarium voucher specimens were collected for species identification with the collaboration of expert botanists, and the specimens were deposited in the herbaria of the University of Colima (UCOL) and the National Herbarium of Mexico (MEXU) under the Alana Pacheco-Flores collection numbers. The botanical nomenclature was verified in the TROPICOS database.

2.3. Biodiversity Conservation Capacity Data Analysis

2.3.1. Species Diversity and Composition

We used the effective number of species, or Hill numbers (^qD), as a measure of true diversity [44]. To compare species diversity across management categories, we generated diversity accumulation curves as a function of sample size (rarefaction curves) and sample coverage. We estimated the diversity of order q = 0, 1 and 2, which differ in the importance given to abundance in the calculation of the diversity index. Diversity of order q = 0 is commonly referred to as the total number of species (species richness), that of order q = 1 is associated with the diversity of common species, while that of order q = 2 is commonly referred to as the diversity of dominant species [45]. Diversity plots were estimated using 999 bootstrap iterations with a 95% confidence interval and standardized to 600 individuals and a sampling completeness of 0.82. Analysis was performed using the iNEXT package [46] in R studio. Statistical differences between categories were analyzed with a linear model (LM) in R studio.

Hill numbers were estimated for each component of diversity using the entropart package [47] to obtain the effective number of species of each category as alpha diversity (^qDα), the effective number of distinct communities as beta diversity (^qDβ), and the total regional diversity as gamma diversity (^qDγ) [44]. Evenness was calculated as the ratio of D²/D⁰. To visualize the proportion of shared species between management categories, we used the nVenn software v1.2 [48]. In addition, to know the number of useful plants in agroforestry practices, data on the species uses were obtained from interviews with local people and supplemented with ethnobotanical literature and databases information.

2.3.2. Vegetation Structure

Vegetation structure was characterized using five variables: abundance, basal area (BA), DBH, plant height, and percentage of multi-stemmed individuals. The DBH data were adjusted to a quadratic diameter using the stem inclusion method, which allows the calculation of DBH values for multiple-stemmed plants in tropical forests [49]. We calculated the BA of each species in m² ha⁻¹, the BA by stem diameter classes in cm (1–<3, 3–<5, 5–<10, 10–<20, 20–<30, 30–<50, and >50), and the BA by management category. We calculated the mean height and the percentage of individuals with multiple stems. Linear models (LM) were fitted to assess differences in the structural characteristics between management categories. Additionally, to compare the contribution of individual species within the plant community, importance values were calculated using the relative frequency, density, and dominance scores [50] for each species.

3. Results

3.1. AFS Characterization

Agroforestry Landscape, Plant Management and Agroforestry Practices

Active and fallow AFSs are spatially arranged in patches throughout the agroforestry landscape studied. The parcels are and have been established on slopes that follow the course of the Armería River. Forest patches are located in areas where slopes are more pronounced, and agriculture is difficult to practice. The cultivation of annual plant species has been reduced among community members, which is mainly due to shifts in economic activities, migration, the perception of low profitability in agriculture due to high fertilization costs, little profit in maize trade, and the effects of climate change. However, milpa cultivation tradition persist for household consumption. In some cases, households have another more productive parcel for maize cultivation, and fallow parcels are used for perennial crops, the gathering of wild products, and livestock browsing.

Recently regenerated parcels, called retoñeras (sprouts), are often deforested to prevent regrowth and facilitate land clearing. The owners of these parcels are actually engaged in other economic activities, growing crops for household consumption, and to maintain communal land rights. These parcels must be cleared before trees reach a larger diameter, as community agreements regulate their felling. In parcels with intermediate regeneration, known as monte grueso (thick forest), management focuses on cattle browsing and the harvesting of firewood and other wild products. Grass is encouraged to grow in cleared areas to serve for fodder. Advanced regeneration parcels are unlikely to be replanted, as family activities change and younger generations no longer practice agriculture. As a result, these parcels are used through practices such as vegetation patches and wild products harvesting, which provide both utilitarian value and ownership of the fallow agroforestry system, as they are perceived as more suitable for cultivation than purely forested areas.

In the active and fallow AFSs sampled, we identified the following agroforestry practices (Figure 2):

(A)

Live fences. This is one of the most common practices because it delimits parcels and prevents livestock from entering. Live fences may include large diameter trees that have been protected for long periods of time, which not only delineate the boundaries of parcels but also provide shade for trails. The components of live fences mostly provide useful products to their managers and have a high resprouting capacity. The most commonly used species were chacalcahuitl (Senegalia macilenta), guamúchil (Pithecellobium dulce), papelillos (Bursera spp.), and zolocuahuitl (Cordia elaeagnoides).

(B)

Vegetation fringes. Borders of native vegetation in different states of conservation are intentionally left in parcels cleared for agriculture, especially in areas with pronounced slopes, which helps to prevent soil erosion. The most common species in these fringes were Crateva palmeri, granjeno (Celtis iguanaea), Coursetia glandulosa, vainilla (Senna atomaria), and San Antonio (Tabernaemontana tomentosa).

(C)

Scattered or isolated trees. These are generally useful species left to grow in the midst of crops, which is usually to provide shade and food. People used to leave tree species that do not shed most of their canopy during the dry season. These are the cases of majahua (Heliocarpus terebinthinaceus) and P. dulce as well as tall trees such as parota (Enterolobium cyclocarpum), and capire (Sideroxylon capiri).

(D)

Vegetation islands. These are small groups of trees composed of one or more tree species such as pitayo (Stenocereus queretaroensis), chacalcahuitl (S. macilenta), and ciruelo (Spondias purpurea).

(E)

Remnants of native vegetation. In this category, we include large areas of vegetation in different stages of conservation that are left to stand within the parcels. These remnants offer a variety of resources, including firewood (S. macilenta) and timber such as tepemezquite (Lysiloma divaricatum) and guásima (Guazuma ulmifolia). They also produce edible forest products such as ilama fruits (Annona macroprophyllata), embiona flower buds (Ledenbergia macrantha), and naznaca, which is a mushroom that grows on Sapranthus violaceus.

(F)

Vegetation patches. These are represented by small orchards of native fruit trees growing together with some wild plant species. Vegetation patches are one of the most common practices on parcels that continue to be used and protected even after the milpa is no longer maintained and the parcels begin to regenerate. The most common fruit trees in these orchards were the ciruelos (S. purpurea), bonetes (Jacaratia mexicana), guamúchil (P. dulce), and pitayos (S. queretaroensis).

(G)

Corral. We found only one corral in an active AFS. It consisted of an array of S. macilenta and huizaches (Vachellia spp.) trees used as a corral for livestock. This practice combines the use of trees for living fences, shade, and forage.

We recorded the highest number of agroforestry practices in active and advanced fallow AFSs. Long-term agroforestry practices such as isolated trees, live fences, vegetation fringes, and vegetation patches were prevalent in regenerating parcels. These practices serve multiple functions including parcel protection, soil erosion control, and the provision of firewood and food. Livestock presence was generally low in all sampled parcels, including vegetation patches, with animals being kept mainly for food and, in some cases, for local trade.

Another plant management practice recorded in both active and fallow AFSs was the removal of species considered by some people to be of little use (H. terebinthinaceaus) or annoying thorny species, such as Celtis iguanaea, C. caudata, and Vachellia spp. Conversely, people protected wild and cultivated tree species by pruning and weeding around them such as C. elaeagnoides, G. ulmifolia, and E. cyclocarpum. In addition, people who do not intend to cultivate their parcels in the near future use fallow parcels to plant or protect trees for future use, such as the fruit trees S. purpurea, J. mexicana, and P. dulce, transforming these parcels into long-term cultivated areas (

Table 1. Management practices recorded in the Zacualpan Nahua community in active and fallow AFSs. Abbreviations: * Agroforestry practices: RV, remnants of native vegetation; LF, live fences; VP, vegetation patches; IT, isolated trees; VF, vegetation fringes; VI, vegetation islands. ** Management: LB, livestock browsing.

Category	Agroforestry Practices *	Management **
1 ACTIVE	RV, VP, VI, IT, VF, LF	Milpa, fruit trees, and cultivation of prickle pear (Opuntia spp.). Gathering, promotion, protection and elimination of wild plants. LB of a few goats.
2 ACTIVE	RV, VP, VI, IT, LF	Maize and fruit tree cultivation. Gathering, promotion, protection and elimination of wild plants. LB of a few cows.
3 ACTIVE	RV, VP, VI, IT, VF, LF	Recently cleared cultivation field after 5 years of regeneration. Maize and S. purpurea cultivation. Protection of E. cyclocarpum, Ficus spp., and living fence. LB of a few cows.
4 RECENT REGENERATION	IT, LF	Protection of living fence of P. dulce. Protection of E. cyclocarpum. Elimination of some thorny trees (Vachellia spp). LB of cows.
5 RECENT REGENERATION	LF, VF	Cultivation of P. dulce, protection of C. eleagnoides (living fence) and S. violaceus; and removal of thorny plants (Celtis spp.). Fence to protect land from cow browsing.
6 RECENT REGENERATION	VP, VF, LF	Cultivation of S. purpurea and J. mexicana. Elimination of undesirable wild plants. Protection of isolated E. cyclocarpum trees.
7 INTERMEDIATE REGENERATION	IT, VP, VF	Promotion and gathering of S. macilenta for firewood. Promotion of pasture for fodder. LB of cows.
8 INTERMEDIATE REGENERATION	LF, VP, RV	Promotion and protection of P. dulce. Protection of Ledenbergia macrantha as living fence. Elimination of undesirable trees (H. terebinthinaceus, Celtis spp.). Protection of S. purpurea trees. LB of cows and goats.
9 INTERMEDIATE REGENERATION	LF, VP	Protection of living fence and E. cyclocarpum. LB of goats for consumption and commerce. Gathering of lumber and firewood.
10 ADVANCED REGENERATION	LF, VP, RV, IT	Protection of J. mexicana, Agave sp., and S. queretaroensis. LB of cows and goats.
11 ADVANCED REGENERATION	VP, RV, IT	Protection of S. purpurea and P. dulce. Gathering of plants, lumber and firewood. LB
12 ADVANCED REGENERATION	LF, VP, RV, IT	Protection of P. dulce trees. Gathering of plants, lumber and firewood. LB

).

3.2. Biodiversity Conservation Capacity

3.2.1. Species Richness and Composition

We recorded 1279 individual plants including 119 native perennial species and 5 morphospecies (N = 124), belonging to 43 plant families and 100 genera. Of these, 33 species were endemic to Mexico. The families Fabaceae (29 species), Burseraceae (7), Rubiaceae (7), Malvaceae (6), Anacardiaceae (5), and Cactaceae (5) accounted for 48% of the recorded species richness. Fabaceae had the highest richness of all categories studied. Combined, both active and fallow AFSs harbored 23 endemic species (Figure 3). Of all the recorded species, 12 were included in a conservation category, such as Bursera sarcopoda, B. heteresthes, Bourreria superba, Dalbergia congestiflora, E. cyclocarpum, S. humilis, S. capiri, and Platymiscium lasiocarpum (Table S1).

The forest plots had a richness value (⁰D) of 76 species, accounting for 61% of the of the gamma diversity (124) across all categories studied, while the active AFSs had a richness of 46 species (37%). Within the fallow AFSs, we observed an increase in species richness values in relation to the duration of the regeneration period: recent use parcels had 25 species (20%), intermediate regeneration parcels had 34 species (27%), and advanced regeneration parcels boasted 54 species (44%). Forest plots harbored 35 species exclusive to that category, while active AFSs contained 10 exclusive species (Figure 4A), which were found within management categories such as vegetation fringes, vegetation islands, and remnants of native vegetation (e.g., Bursera fagaroides, F. pringlei, and Morisonia indica). Parcels in advanced regeneration had 15 species exclusive to this category. Accordingly, the highest number of useful species that we recorded occurred in forest areas (59 species), which was followed by parcels in advanced regeneration (44) and active AFSs (40).

3.2.2. Species Diversity

The rarefaction–extrapolation curves showed a separation of forest plots from all other management categories in the diversity orders of q = 0, 1, and 2, and there was overlap between active and advanced regeneration parcels as well as between parcels in recent and intermediate regeneration (Figure S1). The corresponding linear models showed that forest plots were the most diverse in order ⁰D and that these plots did not differ from active and advanced regeneration AFSs. We identified significant differences between forest and recent (p = 0.02) and intermediate (p = 0.03) regeneration categories (Figure 4B). The same pattern was found for typical species (¹D) and dominant species (²D). The recent regeneration category had the lowest diversity (⁰D), while the intermediate category was the less diverse in typical species (¹Dα) as well in dominant species (²Dα) and evenness (0.15) (

Table 2. Alpha diversity and evenness for active and fallow AFSs and TDF areas in Zacualpan, Colima, Mexico.

Diversity	Active	Recent	Intermediate	Advanced	Forest
0Dα	46	25	34	54	76
1Dα	21.85	12.1	9.4	25.8	43.1
2Dα	13.05	7.09	5.01	14.85	26.11
Evenness factor	0.28	0.28	0.15	0.28	0.34

). In terms of the effective number of communities, the ⁰Dβ value was 2.62 and for ²Dβ, it was 1.68.

3.2.3. Vegetation Structure

Of the 1279 individual plants we recorded, 283 were in forest. However, we found the highest abundance in parcels undergoing intermediate (335) and advanced (293) stages of regeneration, which was followed by those in active (208) and recent (160) regeneration parcels. The most frequently recorded species, accounting for 53% of all individuals, were C. glandulosa, H. terebinthinaceus, S. atomaria, S. macilenta, and V. farnesiana. About 61% of the measured individuals were in the 1–10 cm DBH class (Figure 5A) with 670 (44%) of the individual plants being multi-stemmed trees. We found the highest mean value of qDBH (18.5 cm) and the highest percentage of multi-stemmed trees (60%) in the active AFSs compared to the forest category, where these values were 11.4 and 45%, respectively. The latter value was the lowest percentage of multi-stemmed trees recorded (Table S2).

The cumulative BA of all sampled parcels was 43.9 m²/ha with a mean of 2.9 m²/ha. Ten species together accounted for 60% of the total BA: E. cyclocarpum, P. dulce, C. glandulosa, S. purpurea, C. elaeagnoides, L. macrantha, S. queretaroensis, J. mexicana, S. macilenta, and H. terebinthinaceus. Notably, the forest category had the lowest BA recorded (6.1 m²/ha), which was followed by the intermediate (6.2 m²/ha), recent (8 m²/ha), and advanced (10 m²/ha) regeneration categories with the active AFSs having the highest BA values (13.5 m²/ha). We found no significant differences in BA among management categories (p = 0.6). However, we observed that although the 1–10 cm qDBH class contained the highest number of individuals, it contributed the least to the total BA. Conversely, trees in the qDBH class ≥ 20 cm had a greater contribution in active, recent, and advanced regeneration parcels (Figure 5B).

In the forest category, the trees with the highest recorded BA were Cochlospermum vitifolium, Ledenbergia macrantha, and Ficus cotinifolia, while in the active and fallow AFSs, they were E. cyclocarpum, J. mexicana, P. dulce, and S. capiri. In contrast, the mean tree height was consistent across all categories at 5 m (SD = 0.7). The tallest species in the forest category were Astronium graveolens (17 m), Lonchocarpus rugosus (16.5 m), and Ipomoea sp. (11.6 m), while in the AFS category, they were E. cyclocarpum (16.5 m), F. petiolaris (12.5 m), and F. cotinifolia (10.5 m).

The top 15 species with the highest importance value percentages were C. glandulosa (38.4%), H. terebintinaceus (35.2%), S. atomaria (32.7%), S. macilenta (31.3%), V. farnesiana (25.8%), C. palmeri (21.2%), P. dulce (20.8%), S. purpurea (20.8%), V. campechiana (20.6%), Ledenbergia macrantha (20%), S. queretaroensis (19.7%), Conzattia multiflora (18.7%), C. iguanaea (18.2%), E. cyclocarpum (16.7%), and Guazuma ulmifolia (14.2%); however, the species with the highest importance values differed in each management category. These species are mostly multipurpose plants, and they were present in at least three of the five management categories (Table S1).

4. Discussion

4.1. Agroforestry Practices Characterization

Due to the communal land tenure system in Zacualpan, the existing trees found in the AFSs often reflect the historical agroforestry practices and motivations of previous generations who passed on the AFS parcels to the current owners. As a result, the management of the vegetation structure can be considered communal and intergenerational. This fact reinforces the importance of communal land tenure and the transmission of traditional ecological knowledge (TEK) for biodiversity conservation. Furthermore, the agroforestry practices found in this study were similar to those documented in other AFSs of México [27,28,30], reflecting similarities in biocultural heritage in the use and management of agrobiodiversity [51].

4.2. Biodiversity Conservation Capacity

Species Composition and Diversity

Our results showed that the parcels with the active agriculture and fallow AFSs that we surveyed in Zacualpan together harbored a total of 89 species, representing almost 71% of the species richness (⁰D) of the perennial native flora recorded in the area. Moreover, we observed a progressive increase in species diversity in order q = 0 as succession progressed in the fallow management categories. These systems formed a mosaic that complemented the gamma diversity of the TDF in the study area and provided habitat for 23 endemic and 12 species included in a conservation category. Through the different agroforestry practices, the AFSs represented a reservoir of species diversity, BA, and a source of biocultural resources in crop fields, during active agriculture and fallow periods.

Across all categories, Fabaceae, Burseraceae, Rubiaceae, Malvaceae, Anacardiaceae, and Cactaceae were the families with the highest species richness, which is consistent with other studies conducted in the TDF [25,52,53]. In addition, studies focusing on vegetation succession also have found that species of the family Fabaceae were dominant during the early stages [54,55,56], which is attributed to their functional traits for drought tolerance [10]. In particular, numerous species of the Fabaceae have biocultural value in the study region [43]. Burseraceae was the second family with the highest species richness with seven species from the genus Bursera, which are often used in live fences due to their high regrowth capacity. Within the Cactaceae family, species such as Pachycereus pecten-aboriginum and Stenocereus queretaroensis are maintained and tolerated in cultivated plots, which is characteristic of AFSs in semiarid environments [27,28].

The gradual increase in species diversity with succession that we observed is an indicator of vegetation recovery [12,54,57]. The forest category was the most diverse in all diversity orders, resembling active AFSs and those parcels in advanced regeneration, both of which had similar evenness and number of useful species. Since q = 0 can include rare or low abundance species, this similarity in richness could be related to their presence in some agroforestry practices such as remnants of native vegetation and vegetation fringes. It remains to be seen how viable the populations of these rare or low abundance plant species can be when maintained in AFSs [58], which is an essential objective of future studies.

The finding that active AFSs were the second most diverse category, followed by a decline in species diversity in the recent category, could be attributed to the following reasons. (A) First, active AFSs maintained a greater number of management practices due to the vegetation benefits they provide to meet human needs, as agriculture is the primary activity of landowners. (B) Second, recent regeneration categories are frequently deforested, and their owners have shifted their focus to other economic activities over a longer period of time, resulting in the erosion of TEK. As a result, only the most useful agroforestry practices for them are maintained.

4.3. Vegetation Structure

Compared to the results of studies carried out in other forests in advanced successional stages in the Mexican Pacific coast [16,54], the BA values that we recorded were lower. However, the BA values that we measured in active and in regenerating AFSs in the Zacualpan community represent the human management of plant species over time. Live fences, isolated trees, vegetation patches, and vegetation fringes were the management categories that preserved the highest number of trees in the qDBH class ≥ 20 cm such as E. cyclocarpum, J. mexicana, P. dulce, and S. capiri, all of which are multiple-use plants.

We did not find a trend indicating the increase in BA relative to the fallow period as has been found in other successional studies [12,57]. The parcels with active agriculture and those in recent and intermediate regeneration stages showed the lowest species diversity but accumulated BA. The increase in BA in early and intermediate successional stages has been reported in previous studies and related to the presence in parcels of vegetation remnants as well as the resprouting capacity of pioneer species [10,20]. Compared to forest sites, the AFSs we studied were those with a higher percentage of multi-stemmed plant species. Other studies have reported a large proportion of multiple-stemmed species in high managed sites due to the presence of abundant species, mainly from the family Fabaceae, which have a high coppicing capacity [15].

Therefore, we must consider that the BA is useful as an indicator of regeneration only in the absence of continuous timber and firewood extraction, or intensive slash-and-burn shifting cultivation, as reported by other researchers [11]. Decreases in density and BA reported by other studies have been associated with selective timber extraction [21] and browsing, which are practices that can alter successional pathways and slow the recovery of the TDF through soil compaction, the loss of forest understory, trampling browsing, and inhibition of the recruitment of new individuals [20]. The extraction of timber and firewood is common in the study area, and it is especially practiced in places in the regeneration categories we sampled; however, one of the main consequences of selective harvesting of forest plant resources is the local loss of species and density, favoring species that are resilient to harvesting [21].

In all the management categories we sampled, the species with the highest importance values were C. glandulosa, H. terebinthinaceus, S. atomaria, S. macilenta, and V. farnesiana. Instead, the latter species differed only in their abundance, which was higher in the recent and intermediate regeneration parcels, which is in agreement with the significant differences in typical and dominant species that we observed in these categories as well as in the low values of evenness. Also, the abundance of this species decreased as succession progressed at the same time that other species became dominant. The change in vegetation composition through the successional process goes from an initial dominance of pioneer species—species that are more efficient in resource acquisition—to the gradual dominance of more advanced successional stage species that conserve resources [10,57].

4.4. Conservation and Restoration Implications of Vegetation Succession Management

Vegetation succession management is a set of practices and knowledge that can have beneficial consequences for the conservation of flora and vegetation characteristics over time. In communal land tenure areas, the management of vegetation succession is essential for the protection and promotion of many plant species if a sustainable use is maintained [22]. In the case of the AFSs in Zacualpan, the succession will not progress to full forest regeneration in most cases because the parcels we studied will continue to have an agricultural purpose, but at least for some time, the management of vegetation succession will preserve biodiversity and BA. Considering that Zacualpan is part of a Biological Corridor that connects several protected natural areas, increasing the quality of the landscape matrix with successional forest patches at different stages of regeneration and under different agroforestry practices could increase ecological connectivity and provide habitat for numerous plant and animal species along with other groups of organisms [58].

In our study, the management practices that could promote forest recovery that we recorded were as follows:

(1)

Leaving tree stumps. The way in which trees are felled down allows the persistence of some tree stumps with high coppicing capacity. Local managers recognize this practice as a way to prevent soil erosion and also as a way to promote the rapid recovery of fallen trees.

(2)

Maintaining stands of abundant species with high coppicing capacity. These species are often and can be used as fuelwood and good quality poles [21]. In Zacualpan, the endemic species Senegalia macilenta is recognized as having these characteristics.

(3)

Species protection. In active and in fallow AFSs, the management aimed at protecting plants of interest and eliminating plant species that may pause or arrest vegetation succession—as previously reported for V. farnesiana and Mimosa arenosa [55]—could significantly promote vegetation recovery.

(4)

Promotion of key species through agroforestry practices. The promotion of key species in AFSs favors ecological connectivity by providing habitat and food for dispersing animals that form nucleation clusters [33,43,58,59].

(5)

To continue the practice of fallow. Maintaining landscapes with patches at different successional stages can ensure the presence of habitats used by a diversity of species, soil fertility, and sources of propagules for the recovery of the used land.

(6)

Maintaining long-term fallows parcels. These parcels can be enriched by planting trees and to recover soil fertility and species diversity.

(7)

Maintaining high-quality matrix. It is possible and recommendable to maintain forest patches with 30% of cover in dispersed patches and 10% of cover in continuous vegetation areas, which can be achieved through the various agroforestry practices [58].

(8)

Ensuring the maintenance of forest patches within the agricultural landscape matrix. This action would promote the protection of species and propagules as well as the preservation of key functions for the recovery of ecosystems.

(9)

Promoting the traditional way of managing the milpa system and agrobiodiversity. The milpa system allows for obtaining the benefits of an ancient polyculture and favors the conservation of local biodiversity through agroforestry practices.

(10)

Encouraging the conservation of local ecological knowledge and biocultural memory. Local or traditional ecological knowledge (LEK or TEK) encompasses accumulated knowledge, practices, and beliefs that are culturally transmitted across generations and relate to human interactions with nature. Thus, local knowledge and biocultural memory are critical for protecting, promoting, conserving and restoring biodiversity [60,61].

(11)

Conserving communal land tenure. Community agreements on vegetation management as well as the inheritance of agroforestry practices can promote the conservation of biodiversity.

5. Conclusions

The highest number of agroforestry practices was observed in active and advanced AFSs. In regenerating parcels, long-term agroforestry practices such as isolated trees, live fences, vegetation fringes, and vegetation patches were prevalent. Thus, communal land tenure plays a critical role in the intergenerational management of vegetation structure. These practices provide multiple benefits including parcel protection, soil erosion control, and the provision of food, firewood, and medicine while maintaining ecosystem regulatory contributions.

Active and fallow AFSs together harbored 89 species, representing nearly 71% of the species richness (⁰D) of the perennial native flora recorded in the area. In fallow AFSs, species diversity in order q = 0 increased as succession progressed, representing an indicator of regeneration. Conversely, BA did not show this pattern. BA may be an indicator of intergenerational tree management. These systems formed a mosaic that complemented the gamma diversity of the TDF in the study area and provided habitat for 23 endemic and 12 protected species. The diversity of management practices applied by the members of the community promoted a better quality in the landscape matrix, which could improve the connectivity between fragmented and conserved areas within the Nevado de Colima-Sierra de Manantlán Biological Corridor.

Some pertinent actions for biodiversity conservation based on the site we studied but that can be extended to other regions include the following: (1) leaving tree stumps; (2) maintaining stands of abundant species with high coppicing capacity [21]; (3) protecting species, including those directly beneficial for humans and those with a critical role in vegetation recovery; (4) promoting key species through agroforestry practices; (5) continuing to practice fallow; (6) maintaining and enriching long-term fallows; (7) maintaining high-quality matrix with patches 30% of vegetation cover and 10% of cover in continuous vegetation areas [58]; (8) ensuring the maintenance of forest patches within the agricultural landscape matrix; (9) promoting the traditional way of managing the milpa system and agrobiodiversity; (10) encouraging the conservation of local ecological knowledge and biocultural memory; and (11) conserving communal land tenure. Community agreements on vegetation management as well as the inheritance of agroforestry practices can promote the conservation of biodiversity.