Is it possible that the structure of tropical rainforests has recovered 40 years after clear-cutting?

The definition of secondary forest (SF) in general refers to forests that regenerate naturally in deforested areas as a result of succession following logging or subsequent change in land use (Ngo Bieng et al., 2021; Brown & Lugo, 1990). SF corresponds to forest ecosystems developed after the destruction of the original or primary forest (PF) caused by natural and human disturbances (CNCF, 1999), giving rise to the repopulation process (Finegan, 1997) which is also termed forest recovery (Emrich et al., 2000). Different concepts agree that SF is developed mostly without human intervention (Emrich et al., 2000). According to Ngo Bieng et al. (2022), more than a half of the world’s tropical forests are naturally regenerating forests: secondary and degraded. Currently it refers to degraded secondary forests (SDF) that are growing in areas that have suffered from unsustainable human activities, including former deforestation.

New vegetation cover has characteristics within the regeneration process: it is considered that the process follows stages that are successive in time. The successions are dynamic and tend to a state of equilibrium or climax, commonly known as SF, with characteristics similar to the original forest. The different successional stages differ in their floristic composition and structure, with common components and processes they can be grouped and studied as an association.

SF dynamics can be affected by natural disturbances, such as fire, pests, diseases, floods, and by deforestation. It has been described that the first successional stages in the SF are highly vulnerable to change in land use, meaning a loss of some soil or vegetational attributes – such as the floristic composition and structure (Guariguata & Ostertag, 2002) – will negatively influence its operation leading to a possible decrease in productive capacity.

In Bajo Calima, 35,270 hectares were logged between 1960 and 1990 and 23,600 hectares of primary forest (PF) were preserved by the forestry concession (Rodríguez, 1989). Between 1985 and 1991, the structure and diversity at early ages of natural regeneration after clear-cutting in these forests were determined (Ladrach & Mazuera, 1985; Faber-Langendoen, 1992; Forero-Peña & Ordóñez, 1992; Faber-Langendoen & Gentry, 1991). Regarding the rotation time, Mazuera (1985) set a period of 30 years for the SF to equalize the basal area of the PF. Faber-Langendoen (1992) warned that biomass could decline between 15 and 30 years due to the natural death of many pioneer species. Furthermore, Wright (1997) reported that most SFs were being used, to some extent, by settlers. Recently the study carried out by Pacheco-Pascagaza (2020) found that between 2007 and 2018, 33.75% of the forest was significantly modified with an intensity of 4,207 ha per year, almost 10% of the area was deforested and only 2.4% was regenerated.

Currently most of the forests in the study have been exploited with this continuing. In Bajo Calima harvesting is done selectively to extract timber and non-timber resources among other activities (Forero-Peña & Mora-Delgado, 2017). Martínez (2006) reported that the average cutting diameter in the same 35-year-old secondary forests was 10.2 cm at diameter breast height (DBH), and found 447 stumps ha⁻¹ that had been cut down to be marketed as sticks (more than 6 m in length) and tucas (three meters or less in length).

In the forest, species show different growth patterns, some with a long life and slow growth, and others with a short life span – the latter are called pioneer species (Zent, 1995). Finegan (1997) points out that there are two strategies in the colonization stages; the first strategy sees species exhibit high reproductive intensity, whereas the second strategy favours species with better survival and competition capacities.

In Colombia, little work has been carried out regarding the disturbed SF regeneration dynamics, in which the production trends of the basal area (BA), volume (Vol), and biomass (B) are identified. Specifically, in the SF of Buenaventura approaches have focused on describing impacts, such as degradation and changes on floristic composition (Martínez, 2006). The objective of this research was focused on identifying changes in the SF structure and productive variables (BA, Vol, and B) in a chronosequence of 40 years in Bajo Calima, Colombia. The study of SF chronosequences allows a recognition of changes in landscape in the ecosystem, caused over time and are examined in periods to identify the succession (Peña-Claros, 2003).

The chronosequence of the rain tropical SF did not represent a successional dynamic, since the magnitude of changes of the variables BA, Vol and AGB was not significant. Changes have presented without a defined trend during the 40 years evaluated. Although, at the age of 27, a lower productivity and tree D were evident, attributable to the frequency of anthropogenic interventions, the rate of recovery becomes uncertain because they have been affected by human activity at different levels (Norden et al., 2009) and it is possible that the extractions were of such magnitude that affect the retrieval of D and BA (Forero-Peña et al., 2020), which differs from what was found in the Brazilian Amazon where 80% of BA is achieved in 30 years (de Avila et al., 2015).

It was shown that the production values, similar between three ages: 23, 31 and 35, do not represent the productivity they had in the past. The BA of the 31-year-old SF is similar to that of the younger forests, 18 and 20 years (De las Salas, 2002; Faber-Langendoen, 1992; Forero-Peña & Ordóñez, 1992; Mazuera, 1985), and is lower, that other tropical forest in equivalent conditions (Finegan & Sabogal, 1988), in Nicaragua (Ferreira et al., 2002) and Puerto Rico (Aide et al., 1996).

The BA in Bajo Calima, after 23 years of logging, reached 15.8 m² ha⁻¹ and 15.9 m2 ha⁻¹ at 40 years, values that deviate to a great extent from that reported by Peña-Claros (2003) for Bolivian Amazon forests due to clear-cuttings for agriculture, which after two years of abandonment presented a 12.3 m² ha⁻¹ and at 40 years 36.3 m² ha⁻¹, doubling in this case the BA of Bajo Calima.

There were no differences between SF of 23, 31, and 35 years, but the chronosequence showed highly significant differences between the SF of 23, 31 and 35 years old and the PF of Bajo Calima. In this disturbed group, the BA, Vol and AGB corresponded to 45%, 29% and 22% with respect to the PF, and 94%, 50% and 49% with respect to MSF without intervention. While in Bolivia it was reported that at 25 years the SF reached 70% of the mature forest BA (Finegan & Sabogal, 1988), and in the Amazon 80% of the BA after 30 years (de Avila et al., 2015). In Asian forests it took more time: in Indonesia to reach 82% AGB and 74% AGB took 55 years (Brearley et al., 2004), and in Singapore after 56 years only 58% of the AGB of the neighbouring PF was recovered (Chua et al., 2013) showing that the recovery of the forest structure is slow in these areas.

Poorter et al. (2021) analyzed 12 attributes during the secondary succession of abandoned and slightly disturbed tropical forests, concluding that they have the potential to regrow. After 20 years, the forest attributes reached 78% (33% to 100%) of their old growth values, the recovery to 90% of the old growth values for the structure and diversity of species was from 2.5 to 6 decades, while for the biomass and the composition of species it was slower (>12 decades). Particularly, Zambiazi et al. (2021) found that in some SF located in the Brazilian Atlantic, abandoned from migratory agriculture, and probably not intervened, they managed to accumulate more volume of wood in fast-growing species.

The characterization of PF in Bajo Calima (Rodríguez, 1989) reported an AGB of 190.7 Mg ha⁻¹ (vegetation greater than or equal to 3 cm in diameter), considerably less than that of the PF studied, even more so because this measurement only took into account those with a diameter greater than 10 cm.

As for the results of the AGB of this research, they contrast with those reported by Saldarriaga et al. (1988) in a SF in the Alto Río Negro basin of the Colombian-Venezuelan Amazon, where they reported linear increases up to 40 years, similar to the values reported in SF in Puerto Rico (Aide et al., 1996) and Nicaragua (Ferreira et al., 2002), where production and density increased with the age of SF. In Chocó, Colombia, under extremely high rainfall, production also increased after just 10 years (Quinto Mosquera & Moreno Hurtado, 2011).

The evolution of the SF after 23 and up to 35 years after clear-cutting does not show significant changes in BA, Vol, and AGB variables, which are the basis of the forest recovery or resilience indicators. It is noteworthy that in this rain tropical zone, there are no limitations due to precipitation, on the contrary rainfall exceeds 7,000 mm annual rainfall, with 255 days of rain per year. Under these conditions, the effect of water on the regenerative process of the forest was not evident, as proposed by Poorter et al. (2016) in their modeling study, considering available water as one of the favorable variables for the process.

The AGB of MSF was lower than PF, attributed to the death of many pioneer species. While the AGB accumulated at 23 years (56.5 Mg ha⁻¹) in the SF is lower than that of other tropical forests, in Costa Rica a 20-year-old SF accumulated 65 Mg ha⁻¹ of AGB being in the range reported by Chacón et al. (2007) and Yan et al. (2006).

In tropical eastern Africa (Uganda) (Osazuwa-Peters et al., 2015), as in Bajo Calima, neither the density nor the BA nor the biomass recovered after more than 40 years of logging, revealing the fragility of the tropical SF.

The homogeneity for the diametric distribution was at 23, 27 and 31 years; without successional changes, it seems that regeneration had been continuous (De Lima et al., 2017). This distribution compared to that reported by Forero-Peña & Ordóñez (1992) is similar to that of a 20-year-old forest, but with fewer trees in each size class, a condition that coincides with that reported by Martínez (2006) for a 35-year-old SF subjected to exploitation and with that found by Vallejos (1996) for the Ecuadorian Amazon. Louman et al. (2001) affirms that young SF frequently correspond to approximately coetaneous structure, while the PF, intervened or not, as well as mature secondary ones, present disetaneous structure as is the case of the SF of this research, but with a small diameter.

In the Bajo Calima chronosequence, the resilient capacity of the SF is not evident in the dynamics of its biomass since it does not increase. It is reasoned that this is due to the harvesting carried out by local communities of trees that exceed 10 cm in diameter, affecting successional trajectories, leading to a deleterious effect on forest development (Aide et al., 2013). Thus, recovery in this area is influenced by the context, idiosyncrasy (Martínez, 2006; Forero-Peña & Mora-Delgado, 2017; Norden et al., 2015), the intensity of use (Martínez, 2006; Hüller et al., 2011; Aguiar et al., 2019), and partly by weather conditions.

The structure and productive variables in the SF, after clear-cutting with periods that go beyond 20 and up to 40 years present their own dynamics, which are affected by the inhabitants of the region, without a tendency to reach maturity and far from the productive characteristics of the preceding PF.

The magnitude of the changes of the different variables in the evaluated ages was not significant. In general, the structure of the SF at different ages shows a certain similarity. The results for the ages studied do not meet the characteristics of a sequence, the change is not ordered or unidirectional in progression.

In the secondary forests of Bajo Calima after clear-cutting, it was evident that the resilience process is conditioned to selective extraction, where the forest production variables do not present the same recovery trend.