Introduction

Soil carbon losses in ecosystems and climate change are problems that currently influence agricultural production. Worldwide agricultural soils have lost between 30 and 75 % of SOC stocks from 30 to 40 Mg C ha^-1 (^{La et al., 2007}), which has contributed to the enrichment of greenhouse gases and to the global warming of the atmosphere, provoking what is known as “Climate Change,” a phenomenon that is currently occurring, causing not only natural disasters (extreme droughts and floods) but also major losses in agricultural production (^{Muñoz-Rojas et al., 2017}).

Soil natural coverage and agroforestal systems increase SOC, which helps improve soil aggregates, reduce erosion, diminish carbon and nitrogen losses, and improve its accumulation (^{Chen et al., 2017}); likewise, tropical soil destined to pasture cultivation (Pennisetum purpureum) has proven to be a way to increase SOC stocks as its exploitation increases (^{Lok et al., 2013}). However, the inadequate use and management of soil, aside from contributing to the greenhouse effect, provoke problems related to sustainability due to the degradation of SOC, which negatively acts in its physical and chemicals properties and in its biodiversity as well.

SOC losses are closely related with agro-productive properties, they negatively influence other properties like volume density, dispersion factor, biological activity and the decrease of its productivity. In these soils, ^{Hernández-Jiménez et al. (2013}, ²⁰¹⁴⁾, determined that highly cultivated soils have lost between 50 and 55 % of their agricultural productivity. As well, the size of particle fractions decreases as time passes in cultivated soils (^{Schiedung et al., 2017}).

Deforestation and agro-ecosystem establishment alter SOC stocks as occurred in the humid tropic of the Amazon; in oxisol soils, the ground leveling of the natural rainforest to be turned into pasture resulted in a decrease of SOC in the 20 cm of surface, two years after pasture establishment, it changed from 90.0 to 68.8 t C/ha, but due to the entries of SOC coming from the pasture during a period of eight years, it made return back to 96 t C/ha (^{Cerri et al., 1996}). In dark-reddish latosol soils (rhodic ferrarsol), firstly, they lead to SOC when inappropriate management practices were applied, but later it recovers between 0.3 and 1.91 Mg C ha^-1 year^-1 when agricultural practices with little soil preparation were used (Battle-Bayer et al., 2010). On the other hand, carbon stocks increase one year and a half after deforestation, although this carbon was rapidly mineralized and poorly contributes to SOC stocks of five years after deforestation (^{Fujisaki et al., 2017}).

In Mexico, changes in SOC due to the modification of land use has been reported (^{González-Molina et al., 2014}), as well as changes according to the agricultural use of soilsof the coastal plains of Nayarit (^{Murray et al., 2012}) and SOC losses or gains according to natural coverages and of sugarcane crops in the basin of the Mololoa river, Nayarit (^{Bojorquez et al., 2015}).

In the case of Cuba, it is known that climate change causes the increase of mean temperatures by 0.9 °C in plains (^{Planos et al., 2013}), which was part of what caused the increase of pH in red ferralitic (RF) soils and leachated red ferralitics (LRF) soils of what is known as “Llanura Roja de la Habanda” that comprises Mayabeque and Artemisa provinces (^{Morales & Hernández, 2011}; ^{Hernández-Jiménez et al., 2014}; ^{Cánepa et al., 2015}). In addition, based on SOC losses due to the change of soil use, only the losses of LRF soil ecosystem of this region have been reported, where cultivated soils have lost between 50 and 55 % of SOC for the layer of 0-20 cm of the superior thickness of soil (^{Hernández-Jiménez et al., 2014}). Based on the above, the objective of this work was to evaluate gains or losses of SOC in LRF soils of the province of Mayabeque, under different soil use.

Material and Methods

The zone of study was located in western Cuba, it presents a sub-humid tropical climate with an annual precipitation between 1,300 and 1.,500 mm, an average temperature of 24.5° C and a flat area; the origin material was limestone from the Miocene epoch, with development of LRF soil (^{Hernández-Jímenez et al., 2017}). For this work, previous research works on the diagnosis of the change of properties of LRF soils in Mayabeque province served as starting point (^{Hernández-Jiménez et al., 2013}, ²⁰¹⁴), seven profiles of soils with different coverages were selected, one sampling was performed in triplicate in the period comprised from February to May, 2017.

Results and Discussion

In Figure 1, the results of the changes (gains or losses) in SOC of the studied grove ecosystem were presented, under different soil management.

Figure 1 Gain or losses in SOC in LRF soils under groves and cultivation areas, in Mayabeque province, Cuba. 

As observed, for the ficus grove, at 0-10 cm in depth, in 2010, SOC was of 54 Mg ha^-1, which increased up to 65 Mg ha^-1 seven years later, having a gain of 11. This occurs at depths of 0-20 and 0-30 as well, where it increased between 16 and 33 Mg ha^-1, respectively.

As for the case of mango groves (Mangifera indica), the one of 14 years sequestered more organic carbon than the one of seven years, similar situation was observed for the three evaluated depths, being the sequestration between 6 and 15 Mg ha^-1. In the secondary forest grove, organic carbon gains were observed to increase between 9 and 15 Mg ha^-1 in a period of six years.

Comparing among ficus (Ficus sp.) grove and Arboretum grove sown more than 100 years ago, the highest sequestration over time was observed to occur in ecosystems with ficus groves (Ficus sp.) (33 Mgha^-1), this result could be related to the decomposition of ficus leaves, which may occur in a lesser time in relation to trees of the secondary forest. In this regard, studies performed by ^{Cuevas (2014)}, cited by ^{Cuevas et al. (2014)} stated that the carbon content in fallen leaves depends on the decomposition degree of their elements and the decomposition rate, in turn, was determined by their chemical and physical composition, as well as by the climatic conditions of the place.

Likewise, the study for the cultivation areas (Figure 1) showed that, for the area that was under scrub vegetation for a period of 13 years, SOC losses of 7, 2 and 10 Mg ha^-1 occurred for the depths of 0-10, 10-20 and 20-30 cm respectively, caused by the change from scrub soil to intensive cultivation.

A similar behavior was obtained in the medicinal plant garden, formerly vegetable garden, where the losses were lesser regarding the other two profiles. However, in the area reconverted 12 years ago from pasture to intensive cultivations of maize, sorghum, beans, and tomato under irrigation and application of mineral fertilizers, the losses were higher, -13, -21 and -27 in order, for each of the evaluated depths, respectively. Variance and tandard deviation of the results of soil profiles with different coverage were presented in Table 1.

Regarding the rate of SOC gains and losses for years (Table 2), it was observed that the balance is positive for each one of the depths, stored annually for the layer of 0-30 cm in depth, a gain of 4.71 Mg ha^-1 for the ficus grove, in the case of mango groves, the sequestration was between 1.07-1.71 and for Arboretum of 2.50 Mg ha^-1.

For cultivation areas, SOC losses were generally observed. Only in the medicinal plant garden, gains were obtained at 0-10 and 0-20 cm in depth but not at 0-30 cm in depth where there were SOC losses, this could be related to the presences of organic compost (cow manure) applied on soil at the depth of 0-20 cm in doses of 1 kg m² in each cultivation cycle, for which their decomposition process contributes to the preservation of soil carbon stocks at the depth of 0-20 cm.

Results showed that soils under groves gained SOC in all of their variants, being the soil of ficus grove the one with the highest gain, followed by the soil of secondary forest from INIFAT, which is a lighten grove, and then followed by the mango plantation, while in cultivated soils, there were generally losses in SOC stocks.

SOC gain in forests has been pointed out in many research works, ^{Luis-Mejía et al. (2007)} reported that carbon accumulation of carbon to in soil by deforestation, was of 62 and 18 %, at the depths of 0-5 and 5-10 cm, respectively, in twenty-years-old reforestations. Data adjustment by apparent density and the use of a quadratic model indicated that the average mass of incorporation was of 11.2 and 2.30 t C ha^-1 20 years after, and accumulation rates of 0.561 and 0.11 t C ha^-1 year^-1 at the same depths.

In European mountainous lands, the conversion from pastures to groves has been observed to decrease organic carbon stocks from 110 t C ha^-1 to 81 t C ha^-1 for 40 years, but with the establishment of groves 90 years later, it increased up to 174 t C ha^-1 (^{Hunziker et al., 2017}).

In temperate climate as well, in forest areas of the Monarch butterfly in Michoacán, México, ^{Pérez-Ramírez et al. (2013)}, in their research, showed differences in SOC content found under different types of vegetation and grove conditions. These authors demonstrated that conserved sacred fir round patches of land have 153Mg ha^-1 of SOC in average, while the exploited and altered ones have 95 and 125 Mg ha^-1, respectively. Results showed that the average of SOC in conserved forests of pines-oaks is 103 Mg ha^-1, while the exploited and altered ones have 39 and 13 Mg ha^-1, respectively. Conserved forests stored between 40-80 % of SOC in soil A horizons. SOC content should be considered for determining the impact of forest management and of any other preservation policies.

These results occurred for tropical climates as well, for LRF soils in Cuba, where this work demonstrated that the intensive and continuous agricultural activity causes losses in SOC content resulting in the most affected agricultural layer (0-20 cm) by anthropogenic activity. Similarly, Hernández-Jiménez et al. (2007) established changes in SOC under different coverages (groves, pastures, and cultivation areas), determining a higher SOC content in groves, followed by pastures and the lowest SOC contents recorded in continuous cultivation areas.

^{Bojórquez et al. (2015)} on the analysis of the changes in SOC for pine forests, in pastures, in avocado cultivations and in cultivated lands with sugarcane, located in the basin of Mololoa river, Nayarit, found similar results to the ones obtained in this study, where the stable coverages of forest and pasture generated SOC gains, being pasture the one that recorded the highest quantity, followed by oak forest, pine forest and finally, avocado orchard. The coverage of cultivation with sugarcane, where burning for harvesting and re-burning of residues were practiced generated SOC losses.

The profile 5 was found in a cultivation area of medicinal plants where agro-ecological practices are performed like the application of cow-manure-based organic fertilizer, which contributes to decrease SOC losses. In reference to the above-mentioned, studies realized by ^{Acevedo et al. (2015)}, found that in the organic management, there was a higher production of SOC and a higher quantity soil carbon sequestration(SCS), as well as an apparent lesser soil density in comparison with the conventional management.

Conclusions

In LRF soils from the Mayabeque province, there were SOC gains in quantity as well as in the annual rate in the groves, being higher in ficus groves and in the secondary forest (Arboretum), while in cultivated soils, there were SOC losses, although these losses were not present where agro-ecology was practiced with application of organic compost.