Estimating Carbon Acquisition in a Shade Cocoa Plantation in Southern Bahia, Brazil

Abstract

Cocoa (Theobroma cacao) is one of the world’s most traded commodities. Cocoa grown in agroforestry systems is considered a climate-smart agricultural practice, in part due to the role of shade trees as carbon reservoirs and carbon sinks. In Brazil, most production is concentrated in Bahia state, where traditional cocoa agroforests—locally known as cabrucas—are well known to harbor significant above- and below-ground carbon stocks, although their ability to act as carbon sinks is less well established. By analyzing previously published data on the dynamics of tree assemblages within a 1.7 ha area on a cabruca farm, we estimated an annual carbon increment of 3.46 Mg C ha⁻¹, a value comparable to other shade cocoa plantations elsewhere but more than three times the previous estimate for a cabruca. We discuss the importance of these findings and highlight the potential role of traditional cocoa shade plantations as climate-friendly crops, thus contributing to climate mitigation. It is also essential to highlight the importance of the carbon sequestration and storage services provided by tropical agroforestry systems.

1. Introduction

Agriculture is a major driver of climate change, making the adoption of sustainable, climate-smart farming practices essential. One promising approach is agroforestry, the integration of crops with trees, which can both adapt to and mitigate climate impacts [1]. Cocoa (Theobroma cacao) is one of the world’s most traded commodities [2], and Brazil is a key global cocoa producer [3], 75% of which is produced in Bahia. Cocoa agroforests serve as a notable example of climate-smart agriculture, producing a valuable crop while maintaining substantial organic carbon stocks in both above-ground biomass and soil compartments [4]. In Brazil, a key global cocoa producer, this crop is predominantly grown in Southern Bahia through a traditional agroforestry system called “cabruca”, where cocoa trees are cultivated beneath selectively thinned native forest canopies. These cabrucas maintain a considerable part of the region’s biodiversity [5,6,7] and account for the region’s stored carbon; maintaining these shaded plantations is also expected to significantly mitigate the loss of climatically suitable areas for cocoa cultivation caused by climate change [8,9].

Studies have provided reliable estimates of carbon stocks in cabrucas, including assessments of total organic carbon in soil layers (A, B, and C horizons), which range from 719.2 to 2089.9 Mg ha⁻¹ [10]. The above-ground carbon stocks have also been assessed by different studies, with mean values ranging from 32.5 Mg C ha⁻¹ [4] to 66 Mg C ha⁻¹ [11], and up to 87 Mg C ha⁻¹ [12]. Indeed, cabrucas are estimated to contribute approximately 60% of the total carbon stored in the region’s forest formations, while forest remnants and early secondary forests only retain 32% and 9% of the total stock, respectively [12]. These estimates are crucial not only for demonstrating the significance of these shade cocoa plantations in climate mitigation but also as key information for carbon markets. Indeed, assessments of carbon stocks are currently allowing farmers to trade carbon credits for above-ground stocks in cabrucas, aiming to counteract the current trend of agricultural intensification through avoided deforestation projects [13].

However, cabrucas could also benefit from voluntary carbon markets, depending on their role as carbon sinks—that is, their capacity to sequester atmospheric carbon and convert it into biomass. In fact, the marked values for carbon sequestration are significantly higher than those attained by avoided deforestation [14,15]. Cocoa-based agroforests in Central America, for example, accumulate above-ground carbon at rates of 1.3 to 2.6 Mg C ha⁻¹ per year [4], indicating their potential for a valuable carbon market. Nevertheless, the only data available for cabrucas suggest an estimated carbon acquisition rate of 1.1 Mg C ha⁻¹ per year [4], a value that is below the minimum value reported for similar shade plantations in Central America. However, due to the significant variation in tree density and species composition among different cabrucas, the carbon sequestration potential of these systems is likely to vary as well. Unfortunately, to date, no further studies have addressed this variation or explored the full potential of cabrucas for carbon capture. In this context, our study aims to fill this critical gap by providing an estimate of the annual carbon sequestration rate in these agroforestry systems. To do this, we analyzed data from a published article [16], which monitored changes in the basal area in a cabruca site after an interval of 7 years. To the best of our knowledge, this work [16] is the only study providing such high-quality data on tree growth in cabrucas, thus representing a rare opportunity for us to assess the potential of this system for carbon sequestration.

2. Material and Methods

The study area is located at Fazenda Retiro, in the municipality of Ilhéus, (14°43′14″ S, 39°09′31″ W), Bahia, Brazil (Figure 1). It is located within the Atlantic Forest domain, more specifically in Southern Bahia, which harbors the world’s record of local tree species richness [17], nearly 25% being endemic [18]. The regional climate has a tropical climate, classified as Af according to Köppen, with precipitation levels up to 2000 mm annually and no clearly defined dry season, although occasionally one to three months may receive less than 100 mm of rain [19]. The average annual temperature is between 24 and 25 °C, with an annual variation of 7–8 °C, and the coldest months are June, July, and August.

The property comprises a 10 ha cabruca established between 1973 and 1974, when native forest was thinned and the largest and most resilient trees were selected as shade for cocoa, which includes multiple varieties of cacao trees. Within this cabruca, 34 plots of 20 × 25 m were distributed in four contiguous parallel strips, totaling a sampling area of 1.7 ha. The plots were randomly selected within the same type of environment, avoiding steep slopes and low-lying terrain prone to flooding [16]. As a result, the selected sites shared similar conditions in terms of moisture, temperature, and soil type. Excluding cocoa trees, all individual trees with a DBH (diameter at breast height) of 10 cm or more were measured, had their height estimated, marked, and identified at the species or morphospecies level. Two assessments were conducted, the first occurring in one in 1998 with a total of 120 individual trees from 62 species (all native species) and the second assessment occurring 7 years after in 2005, when it was possible to estimate an annual mortality rate of 2.16%, a recruitment rate of 0.81%, and an average rate of diameter growth for living individuals of 0.7 ± 0.6 cm per year, ranging from −0.1 to 2.6 cm per year [16].

To estimate carbon acquisition, for each measured species, we used the diameter at breast height (DBH) of each individual tree over a 7-year period. We employed the BIOMASS package [20], which calculates the above-ground biomass and carbon of each tree using equation 7 of Chave (2014). This equation considers AGB_est = exp [−1.803 − 0.976E + 0.976 ln(p) + 2.673 ln(D) − 0.0299 [ln(D)]²], in which AGB_est is the above-ground biomass, E is a measure of environmental stress, and D is the diameter [21]. This method determines the wood density of each tree based on its species or genus from a global database. Site coordinates were used to estimate the “E” parameter of equation 7, which considers the impact of climate on the allometric relationship between tree height and diameter [21]. To calculate the errors associated with the biomass and carbon estimation for the tree sets in the 1998 and 2005 periods, we used Monte Carlo simulation procedures and error propagation [20]. All analyses were implemented in the R environment [22].

We subtracted the estimated biomass and carbon of 2005 from the results obtained for 1998 to estimate the total biomass and carbon increment after 7 years. Subsequently, we divided the result by 7 to estimate the annual biomass increment of per hectare (for a more detailed description of the analytical procedure, see the Markdown Document in the Supplementary Materials). In addition to the calculations for the entire community, we also performed separate analyses for the subgroups of individuals belonging to fast-growing and slow-growing species. We classified plant species as slow-growth, i.e., those that are capable of establishing in forest shade and that normally have greater wood density and greater longevity, or as fast-growth, those that present opposite characteristics such as shade intolerance, lower wood density and longevity [23,24]. For 7% of the recorded species that did not have a well-defined pattern (intermediate characteristics between the two strategies), no group was assigned. Our classifications were based on a comprehensive literature review and consultations with experts who have over two decades of experience studying the region’s flora.

3. Results and Discussion

In the 1998 survey, the mean estimated above-ground biomass within the 1.7 ha area was 425.79 ± 32.58 Mg C. By 2005, this estimate had increased to 506.58 ± 39.28 Mg C (Table S1). Consequently, the carbon estimates for these years were 200.14 ± 15.35 Mg C and 237.55 ± 17.15 Mg C, respectively (Figure 2a). This corresponds to an annual carbon increment of 3.14 Mg C ha⁻¹ (Figure 2b; Table S1), which is three times higher than the current estimate for cabrucas and greater than the average reported for other cocoa agroforestry systems (2.6 Mg C ha⁻¹) in Latin America [4]. Additionally, this rate falls within the range of annual carbon sequestration reported for well-preserved, advanced successional forests in Southeastern Brazil, averaging 2.4 Mg C ha⁻¹, with values ranging from 0.2 to 7 Mg C ha⁻¹ [25]. However, the composition and structure of trees vary among different cabrucas, and these characteristics are key determinants of their potential for carbon sequestration and storage [26,27,28]. For instance, in older cabrucas, such as those sampled in this study, all recorded tree species were native to Brazil, with a high proportion of slow-growing, large-sized trees [29]. Slow-growing species dominated these agroforestry systems, accounting for 78.81% of the individuals and 75.86% of the species (Table S1), with an annual carbon accumulation rate of 2.85 Mg C ha⁻¹ (Figure 2b). In contrast, fast-growing species only represented 16% of the individuals and 17% of the species (Table S1), with an annual accumulation rate of 0.38 Mg C ha⁻¹, corresponding to only 10% of the total annual carbon increment in the agroforest (Figure 2b). These species significantly contribute to biomass accumulation, with per capita increases exceeding 0.15 Mg/ha/year, and include Eryotheca sp., Ficus clusiifolia, Ficus gomelleira, Ficus trigonata, Licaria bahiana, and Micropholis crassipedicellata (Table S2). However, over time, even poorly managed shade plantations tend to shift shade tree composition, favoring the proliferation of fast-growing species [30]. While these pioneer species establish readily and can significantly increase biomass in the short term, they do not contribute as much as those late successional species to long-term carbon storage [31,32,33], thereby reducing the long-term carbon sequestration capacity of cabrucas. The recruitment of new individuals resulted in an only modest increase in carbon stocks, with an annual rate of 0.55 Mg C ha⁻¹, occurring infrequently and accounting for just 6.78% of the total number of individuals. (Table S1). Similarly, the decrease in carbon increment due to tree mortality was minimal, as mortality only affected 4.24% of the individuals and had an annual rate of 0.11 Mg C ha⁻¹. Therefore, the observed carbon accumulation in this ecosystem was primarily driven by the growth of existing individuals, predominantly composed of slow-growing species (Figure 2).

Therefore, considering the variation in structural and compositional features among cabrucas, our estimate is only preliminary at most. Nevertheless, for the sake of an exercise, if we assume that such a number is somehow representative of a mean or even the maximum potential amount of carbon sequestration for this land use, we could roughly estimate that the 656,200 ha covered by cabrucas within Southern Bahia [34] could be fixing ~2.5 Tg C annually. This value represents more than ⅕ (21.4%) of the total estimated above-ground carbon currently hosted by mature stands in the region [12]. Even considering the previous low estimate of 1.1 Mg C ha⁻¹ [4], the annual sequestration of 0.72 Tg C would be still significant, particularly when we consider the high rate with which this biome is losing forest cover, with Bahia as a major state leading this tragedy [35]. In fact, studies have estimated a decay of 2.3% in carbon gains even in well-preserved forests along the Atlantic Forest Biome, an estimated average loss of 0.13 Mg C ha⁻¹ year⁻¹ [25]. Such a looming scenario is likely to get worse considering the increasing temperatures predicted by climate change, because carbon gains are often lower under warmer and drier environmental conditions [36].

4. Conclusions

Our study highlights that cabrucas play a significant role as carbon sinks on a regional scale, although such potential varies across different plantations due to obvious differences in vegetation structure. These findings contribute to the growing body of evidence on the importance of this land-use system in enhancing regional climate resilience, acting as key carbon reservoirs, and promoting low-carbon-footprint cultivation practices. These results also highlight the importance of tropical agroforestry systems in providing carbon sequestration and storage services. In light of the global relevance of carbon sequestration, the need for more sustainable production models, and the emergence of carbon markets, studies assessing such aspects of cocoa agroforests—including cabrucas—are becoming increasingly essential.