Land abandonment as driver of woody vegetation dynamics in Tamaulipan thornscrub at Northeastern Mexico

Introduction

Tropical dry forests are among the most threatened ecosystems in Latin America (García-Millán & Sanchez-Azofeifa, 2018), due to the land-use change for many productive activities among we can include agriculture and cattle expansion (Portillo-Quintero & Sánchez-Azofeifa, 2010). These activities affect directly the dry forest biodiversity and structure since these ecosystems have one of the most endemisms in the globe (Sánchez-Azofeifa et al., 2009).

Vegetation structure is defined as the temporal and spatial distribution of plant species in a particular site; it has been widely used as an indicator of successional changes through time after a process of disturbance in natural ecosystems (Domínguez-Gómez et al., 2013; Kulmatiski & Beard, 2019). Vegetation structure includes vertical and horizontal distribution of vegetation, spatial patterns, and size and age of trees (Oliver & Larson, 1996). These factors are important in ecosystem studies since it can be inferred whether an ecosystem reached the mature successional stage (Domínguez-Gómez et al., 2013). Studies of ecological succession play an important role in the determination of the relative importance of mechanisms that structure plant communities under different environmental conditions (Choung et al., 2020; McCook, 1994; van Andel, Bakker & Grootjans, 1993). Additionally, understanding the forest successional dynamics in forests is a crucial issue for the establishment and design of conservation strategies (Sánchez-Azofeifa et al., 2005; Quesada et al., 2009). According to Falkowski, Chankin & Diemont (2020), several factors affect secondary forest succession, among them: abiotic conditions, the surrounding space, and land-use history (Quesada et al., 2009). The latter has been recognized as one of main drivers of changes in vegetation successional dynamics (Baeza et al., 2007; Quesada et al., 2009; Pequeño-Ledezma et al., 2018). The land-use change is defined as the modification of native vegetation into areas for productive purposes (Alanís-Rodríguez et al., 2015a; Falkowski, Chankin & Diemont, 2020; Pequeño-Ledezma et al., 2018; Zhang et al., 2018) like cattle grazing, agriculture, or mineral extraction.

After an anthropogenic disturbance, such as logging, grazing, or farming, forests follow a revegetation process characterized by changes in the original floristic composition and vegetation structure (Kitenberga et al., 2020; Lohbeck et al., 2020; Tálamo et al., 2020; Zhang et al., 2018). These natural successional processes eventually could restore the attributes of degraded ecosystems in old-growth forests (DeWalt, Maliakal & Denslow, 2003). Nevertheless, despite their importance, the successional trajectories of plant communities (mainly in dry or deciduous forests) after land abandonment are still poorly understood (Bonet & Pausas, 2004; Quesada et al., 2009). Thus, the changes in vegetation structure, composition, and abundance are important to understand the potential driver of long-term land abandonment on the configuration of mature ecosystems of Tamaulipan thornscrub.

Sarmiento-Muñoz et al. (2019) found that the Tamaulipan thornscrub areas used for grazing could affect the natural ecological succession of the ecosystem. Recently, several studies highlighted the structure of regenerated woody communities resulting from secondary succession (Alanís-Rodríguez et al., 2008; González-Rodríguez et al., 2010; Jiménez-Pérez et al., 2009, 2012, 2013; Patiño-Flores et al., 2021), but none of them explored the successional trajectories after land abandonment under a well-defined chronosequence. Although plant diversity has been considered in many studies on Tamaulipan thornscrub and its importance is well known (Alanís-Rodríguez et al., 2015a). It is not yet clear how land abandonment (successional changes) contributes to the changes in woody plant diversity, how the tree and shrub species are replaced by others, or how these replacement contributes to the general structure of Tamaulipan thornscrub.

In this study, we explored how the successional stages of four areas of Tamaulipan thornscrub drive the changes in species richness, species diversity (alpha and beta), vertical structure of vegetation, and ecological importance value index. Specifically, we wanted to know: (1) How does species richness, diversity, and vertical structure (A index) change concerning the time of abandonment? (2) Are species similarities among vegetation communities related to land-use abandonment? (3) Which species are the most important within each successional stage? According to this background and questions, we elaborated three main hypotheses: (1) the earlier successional areas would have lower values of species richness, species diversity, and the species height profile index than sites with the longest time of abandonment; (2) the species similarity among sites will be related to the time of abandonment; therefore we would find the lowest species similarity between the sites with early and late years of abandonment; and (3) the species with a higher ecological importance value in early successional stages will not have high importance values in sites with vegetation on late-successional stage.

Materials and Methods

Study area

The study area was located in a Tamaulipan thornscrub in Linares, Nuevo Leon state, Northeastern Mexico (25°09´–24°33´N and 99° 54´–99°07´W; Fig. 1) at an elevation of 350–411 m a.s.l. This area has an arid desert-warm climate with a maximum amount of rainfall during September (INEGI, 1986). The area has an annual average temperature of 23.7 °C and annual cumulative precipitation of 810 mm (Martínez-Adriano et al., 2021). Low hills with gentle slopes are predominant in the area, mainly with Vertisol soils (INEGI, 1986). The four selected areas had similar topographic conditions and only differed in the years of abandonment.

We selected four areas with different time of abandonment after last recorded human disturbance occurs (10, 20, 30, and >30 years). The first three areas were used for cattle grazing, whereas the >30-year area was selected as a control. The native vegetation in the three grazed areas was removed with heavy machinery and replaced with forage grass (Cenchrus ciliaris L.) as food for cattle. Grazing areas were excluded 10, 20, 30, and >30 years before the start of this study and it allowed the natural plant succession. The evaluated areas were located next to plant communities that are growing after livestock activity. The control area (>30 yr) shows Tamaulipan thornscrub vegetation without evidence of disturbance or land-use changes. Local people used the control area exclusively to collect wild fruits and firewood from the ground. This information was corroborated by satellite images and personal interviews. Additionally, we highlighted that EA obtained verbal permission from the owners of the four sites used to conduct this study.

Results

We recorded 27 woody species, belonging to 23 genera and 15 families. The families with highest species number were Fabaceae (11 species), Rubiaceae (two), and Rutaceae (two). While the most representative genera were Acacia (three), Parkinsonia, and Randia (both with two species each; see Table 1). We observed Acacia farnesiana as the unique plant species found in all conditions of abandonment, while species like Cordia boissieri, Celtis pallida, Prosopis laevigata, and Sideroxylon celastrinum were observed in areas ≥20 yr of abandonment. On the other hand, we found that Bernardia myricifolia, Condalia hookeri, Eysenhardtia polystachya, Leucophyllum frutescens, and Yucca filifera only occurring in >30 yr site (Table 1).

Species richness and effective number of species. We found that the >30 yr site showed the highest species richness (21 taxa) and the number of species decreased as the disturbance became more recent (F_{3, 12} = 17.43; p < 0.001). Likewise, for the effective number of species (¹D), we found a similar pattern (F_{3, 12} = 22.76; p < 0.001), where the >30 yr site had the highest number of effective species (¹D = 9.3), followed by the 30 yr (¹D = 4.53), 20 yr (¹D = 2.39), and the lowest was found in the 10 yr site (¹D = 1.28).

Species Height Profile Index (A-index). For this index, the residuals were normally distributed (W = 0.98, p = 0.97) and the variance was not homogeneous (Levene’s test: F = 4.58; p = 0.02). The one-way analysis showed differences among the areas with times of abandonment (F_{3, 6.25} = 21.135; p = 0.0011). The highest altitudinal floors were recorded in the >30 yr area (A = 2.68 ± 0 .18), while the lowest altitudinal floors were observed in the 10 yr area (A = 1.01 ± 0.37).

Similarity among areas with different years of abandonment

We found the highest similarity between areas with closer years of abandonment and decreased with the temporal distance of abandonment. The most similar areas were 10 and 20 yr, followed by 20 and 30 yr (Table 3). The cluster analysis showed three well-defined groups (Cophenetic correlation = 0.95), one formed by 10 and 20 yr sites, and the 30 yr and >30 yr sites in solitary branch each (Table 3; Fig. 2).

Table 3:

Bray-Curtis similarity matrix.

	10	20	30	>30
10	1.000	0.708	0.410	0.145
20		1.000	0.599	0.186
30			1.000	0.255
>30				1.000

DOI: 10.7717/peerj.15438/table-3

Note:

The highest values of similarity between paired areas is shown in bold.

Discussion

In this study, we simultaneously explored the changes in species richness, diversity, and importance value index among four sites with different times of abandonment in the Tamaulipan thornscrub.

We found 27 woody species in this study which represents the same number of plant species reported by Jiménez-Pérez et al. (2009) and similar to the species recorded by Alanís-Rodríguez et al. (2013, 2015b) (30 plant species) in the Tamaulipan thornscrub. While Canizales-Velázquez et al. (2009) reported 52 plant species in a piedmont scrub (a value almost two times higher than our results). This means that our study represents at least more than half of the plant species of these previous studies. Additionally, we observed that Fabaceae was the family with the highest number of species. These results agree with previous floristic studies in remnants of Tamaulipan thornscrub, where Fabaceae was the family with a higher number of species (Alanís-Rodríguez et al., 2015b; González-Rodríguez et al., 2010; Jiménez-Pérez et al., 2009; Molina-Guerra et al., 2013). This indicates that the family possibly has species with the potential to adapt to the conditions that occur at sites with different successional stages (e.g., humidity, exposure, radiation).

Our first hypothesis said that the early successional sites would have lower values of species richness, diversity, and A index than sites with earlier-time of abandonment. This prediction is supported by our results since our findings showed that species richness, effective species index, and A index were higher in the >30 yr site, while the number of species decreased when the successional stage was more recent. These results are similar to the general pattern for scrubs found in previous studies where the oldest plant communities have higher plant species richness and diversity (Abella, 2010; Alanís-Rodríguez et al., 2018; Connell, 1978; Ugalde, Granados-Sánchez & Sánchez-González, 2008). Similarly, several studies highlight an increase in species richness and diversity of tree communities from early to late successional stages in tropical deciduous forests (Quesada et al., 2009). However, previous studies found that certain attributes of plant communities (as basal area) are predicted adequately from chronosequences (Chazdon et al., 2007; Pascarella, Aide & Zimmerman, 2004) whereas differences are found concerning species richness and species composition (Chazdon et al., 2007; Johnson & Miyanishi, 2008; Pascarella, Aide & Zimmerman, 2004; Sheil, 2001). Regarding abundance and biomass (volume) we found similar patterns that a study on the Sonoran thornscrub, which also found that older-growth forests have a more complex structure, higher biomass, and individual densities than younger ones (Álvarez-Yépiz et al., 2008). Thus, we corroborated that plant communities that occur in later successional stages within Tamaulipan thornscrub, also could have more complex structure (in both horizontal and vertical stratification), than those that occur in early successional stages. This makes us notice the importance of ecological succession in the dynamics and structure of Tamaulipan thornscrub ecosystems.

Our second hypothesis said that species similarities among sites will be related to the time of abandonment; therefore, we would find the lowest species similarity between the sites with early and late years of abandonment. Our results supported this prediction and were similar to the findings of Alanís-Rodríguez et al. (2018) since the highest similarity occurred among areas with a similar time of abandonment. Additionally, we found the same pattern of similarity to the findings reported by Sánchez-Reyes et al. (2021), where the lowest similarity was observed between the two sites with an extreme time of abandonment (10 and 30 yr in our study, and 4 yr and conserved areas in the study of Sánchez-Reyes et al. (2021)). Kalacska et al. (2004) also found a lower similarity between the early and late stages of successional stages. However, we found the highest similarity among 10 and 20 yr sites, but Connell (1978) attributes this pattern to the possible coexistence of typical species of early, intermediate, and late-successional stages. Thus, our results suggest that the Tamaulipan thornscrub also follows a similar trend to dry forests in tropical regions.

Our third hypothesis said that species with a high ecological importance value in early successional stages will not have high importance values in fragments with vegetation in the late-successional stage. In this sense, we found that Fabaceae species have high representativeness in all stages of abandonment, like recent studies findings at Tamaulipan thornscrub (Alanís-Rodríguez et al., 2018; Pequeño-Ledezma et al., 2018). It has been documented that Fabaceae species play an important role as pioneer species and can be found in the initial stages of ecological succession due to their capacity to fix atmospheric nitrogen and resistance to drought (Jia et al., 2020; Zhou et al., 2019). These adaptations could give to this plant family the ability to maintain high abundances in disturbed areas, mainly in the early successional stages. Contrary, in late-successional stages, Fabaceae species are generally less important, because these species require open sites with abundant light.

At the species level, we found that Acacia farnesiana was the woody species with the highest abundance and IVI values in the first three successional stages (10, 20, and 30 years of abandonment), but in the >30 yr area Havardia pallens and Diospyros texana become the species with highest IVI values. These results are similar to the findings reported in several studies from Tamaulipan thornscrub (Alanís-Rodríguez et al., 2018; González-Rodríguez et al., 2010; Pequeño-Ledezma et al., 2017, 2018), where A. farnesiana (along with other native species of Tamaulipan thornscrub) was recorded as species with higher dominance in the four areas of Tamaulipan thornscrub. This could indicate that A. farnesiana could play an important role in the early stages of the successional process at Tamaulipan thornscrub vegetation. Thus, A. farnesiana could have an essential function in the regeneration of ecosystems that have disturbance such as grazing.

In addition to all aforementioned, we highlight that Tamaulipan thornscrub areas have a relatively fast successional recovery (~30 years). However, Quesada et al. (2009) said that regeneration rates in dry forests show relatively lower regeneration rates than we think because in almost secondary forests the coppicing from remainings (stumps and roots) after disturbance theoretically allows these forests to reach maturity faster than the other forests (Ewel, 1977; Murphy & Lugo, 1986). In the case of this study, the areas have a well-defined history of use for cattle grazing (according to owners), therefore their rapid recovery is basically due to the introduction of pioneer species by some seed vector of surrounding matrix of vegetation. Then, we recommend that future studies also explore the distance of the seed sources from the surrounding matrix to the studied sites and the intensity of use before abandonment (Guariguata & Ostertag, 2001); services of seed vectors, which are variables that have an important role in plant dynamics in degraded sites (Ramos-Robles et al., 2018); and changes in the functional characteristics of woody species within fragments (Guariguata & Ostertag, 2001). Thus, we highlight the importance of secondary forests for Tamaulipan thornscrub, as they harbor a higher woody plant diversity, which in turn could have the capacity to support the associated fauna that occurs in places with intact vegetation.

Conclusions

With the resulting data, we observed the successional trajectory of the Tamaulipan thornscrub fragments. We highlighted that the advance of the successional stages in post-grazing sites of Tamaulipan thornscrub (latest successional stages) generates communities with higher species richness and a higher complex structure than those communities with earliest time of abandonment. The highest similarity among sites with different time of abandonment was observed between sites with closer time of abandonment. On the other hand, a lower similarity was shown between the sites with extremes time of abandonment. Acacia farnesiana was the most important and abundant species in the first three successional stages, so we propose it as good pioneer species for restoration programs on Tamaulipan thornscrub ecosystems. We conclude that Tamaulipan thornscrub shows a similar trend of ecological succession to other dry forests and the time of abandonment has a high mediation on plant dynamics in the Tamaulipan thornscrub. Also, we stand out the importance of secondary forests for Tamaulipan thornscrub communities, as they harbor a higher woody plant diversity, which in turn could have the capacity to support the associated fauna that occurs in sites with intact vegetation. Finally, we recommended for future successional studies include aspects of regeneration speed, the proximity of mature vegetation within the matrix where these studied fragments are found, and the interactions of plants with their seed dispersers.

Supplemental Information