FINE ROOT PRODUCTION AND TURNOVER ALONG A FOREST SUCCESSION AFTER SLASH-AND-BURN AGRICULTURE

turnover rates can vary with the successional gradients of the forest ecosystems created by land use changes. Objective. To evaluate the variations in fine root production and turnover along the successional gradients of tropical secondary forests and compare them with adjacent primary forests. Methodology. A total of 256 cylindrical ingrowth bags (of 8 cm diameter and 30 cm depth) were established in 16 forest stands representing four stages of succession. Four ingrowth bags from each stand were collected at an interval of about three months. Live and dead fine roots (≤2 mm diameter) were separated, washed, oven-dried, and weighed. Results. There was no significant difference in annual fine root productivity between secondary forests and primary forests but the fine root turnover rates were higher in secondary forests than in primary forests. Fine root production rates varied from 1.9 to 2.8 Mg of dry biomass ha -1 yr -1 while fine root turnover over rates ranged from 1.1 to 1.5 yr -1 . Fine root production was higher in the dry season compared to the wet season. Implications. The results on fine root dynamics can be useful in modeling below-ground mechanisms of carbon sequestration in forest ecosystems during succession. Conclusions . Fine root productivity did not vary with forest age but the relative turnover rates were higher in secondary forests than in primary forests. Tree basal area was a significant predictor of fine root production.


INTRODUCTION
Fine root production is a fundamental component of forest productivity because it contributes to 22% -40% of terrestrial net primary production (Lima et al. 2010;Finér et al. 2011a;Luke McCormack et al. 2013;Pandey et al. 2023).Fine root production and turnover is one of the important pathways of carbon and nutrient cycling in forest ecosystems (Nadelhoffer and Raich 1992;Matamala et al. 2003;Jourdan et al. 2008;Germon et al. 2020).Changes in fine root distribution affect mycorrhizal colonization, organic matter, and nutrient allocation to deeper soil profiles (Addo-Danso et al. 2020;Cordeiro et al. 2020;Cusack and Turner 2021).Along with litterfall, fine root turnover provides detrital carbon to soil organisms, which enhances the overall microbial activity in the soil and is a variable to improve soil water holding capacity and ultimately improve soil quality (Metcalfe et al. 2008;Luke McCormack et al. 2013;Zheng et al. 2021;Sánchez-Silva et al. 2022).Fine roots represent a dynamic fraction of belowground biomass and an important nutrient reserve for secondary forest growth, while the turnover is one of the key mechanisms of the forest ecosystem recovery after natural or anthropogenic disturbances (Brunner et al. 2013;Hertel et al. 2013;Freschet et al. 2021).Production and decomposition of fine roots are also related to carbon fluxes from the forest ecosystems since they emit carbon dioxide when they respire and decompose (Chen et al. 2004;Luke McCormack et al. 2013;Sun et al. 2020).Fine root productivity in European broadleaf forests was reported to be 2.5 Mg ha -1 yr -1 while in tropical forests of the Brazilian Amazon, it was 1.6 Mg ha -1 yr -1 (Newman et al. 2020, Martins et al. 2022).However, fine root productivity varies with the changes in forest structural properties and functions.
The structure and function of the tropical forests in the world are in a continuous process of change due to different types of anthropogenic disturbances (Marín-Spiotta et al. 2008;Chazdon 2014;Poorter et al. 2021).Successional forest patches at different stages of growth are the common characteristics of tropical landscapes in southern Mexico because of the recent land use changes (Velasco-Murguía et al. 2021;Jakovac et al. 2022;Aryal et al. 2024).Such successional secondary forest regeneration mostly occurs in the land left fallow after the cessation of farming activities such as slash-andburn agriculture.During the recovery from younger to older ages, these secondary forests undergo through multiple ecological changes.One of the changes during secondary succession is the productivity and turnover of fine roots.Successional changes in species composition, stand structure, above and belowground biomass stocks, and soil fertility can explain the variation in fine root production during secondary vegetation growth (Riutta et al. 2021;Aryal et al. 2022).It is fundamental to understand the successional dynamics of fine root growth and turnover because a significant proportion of total biomass stock and production takes place belowground (Neill 1992;Kochsiek et al. 2013;Martins et al. 2021).
Shifting cultivation is still a common practice in southern Mexico where farmers slash and burn primary as well as secondary forests for milpa, a corn-squash-bean agriculture, for one or few years and they abandon the land for fallowing (Batista et al. 2021;Falkowski et al. 2023).The fallow period of those forests is getting shorter because the pressure for food production is increasing.Many studies have focused on aboveground carbon dynamics but the belowground carbon dynamics of those successional forests has not been studied widely (Girardin et al. 2010;Kochsiek et al. 2013).In this study, we tested the following two hypotheses: i) fine root production dynamics are related to forest growth stages, and ii) aboveground parameters of forest structure like basal area can explain the fine root production variation to some extent in tropical secondary forests.

Study site and experimental plots
The study was conducted in a chronosequence of tropical secondary forests regenerated after the cessation of slash-and-burn agriculture.Secondary forests (SF) of 5, 10, and 20 years of age and oldgrowth primary forests (PF) were sampled.Slash and burn agriculture in the region comprise of clearing forests, cultivating milpa (corn-squashbean) for a few years and leaving the land fallow for forest recovery.Sampling sites were located in ejidos El Carmen II and Cristóbal Colon of Calakmul municipality, in the state of Campeche, the southern part of the Yucatan Peninsula, Mexico (Fig. 1).The region is characterized by a subhumid tropical climate (García 1973) with an average precipitation of 1000-1500 mm per year (with major portions of the rainfall from July to October) and mean annual temperature of 22 -26 o C (García Gil et al. 2002).Rendzic leptosols and vertisols are the dominating soil types ( Bautista et al. 2011).The dominant forest type in the region is semi-evergreen tropical forest (Rzedowski 1981;Pérez-Salicrup 2004), of which large portions have been converted to slash-and-burn agriculture that created a mosaic of agricultural lands mixed with secondary forests in various stages of development.A total of 16 sampling plots (four plots × four successional categories) were established to evaluate fine root production and turnover (Fig. 1).Successional categories include 1) younger secondary forests of five years old, 2) medium secondary forests of 10 years, 3) advanced secondary forests of 20 years, and 4) old-growth primary forests.The most common tree species of these successional forests are listed in Table 1 in the order of abundance.The average stand height corresponding to 5-, 10-, and 20-years SF, and PF are found in the range of 3 -6, 6 -10, 10 -15, and 20 -25 meters respectively.The average diameter at breast height (DBH) of all the inventoried trees were 3.2, 4.6, 5.4, and 7.5 cm respectively for 8.4,8.4,and 5.0 thousand trees (>1 cm DBH) per hectare were recorded respectively for 5-, 10-, and 20-years SF and PF.Coccoloba acapulcensis Successional categories: 5 yrs.SF -younger secondary forests of five years old, 10 yrs.SF-medium secondary forests of 10 years, 20 yrs.SF -advanced secondary forests of 20 years, and PF -old-growth primary forests.

Fine root production and turnover
A total of 256 ingrowth bags (16 bags per plot) of 8 cm diameter were installed in the soil to a depth of 30 cm (Fig. 2).Sixteen ingrowth bags were placed in a row at an interval of about 0.5 m taking in to account the minimum distance of 0.5 m from tree trunks in each plot.A metal core (probe) of 8 cm diameter was used to dig the hole and take the soil out.All fine and coarse roots were removed from the soil.The root-free soils were reused to fill up the nylon bags (0.4 mm mesh size) and placed again in the same hole.Bags were filled with the soil after eliminating all living and dead roots from the same hole maintaining approximately the original layer, structure, and soil density (Jourdan et al. 2008).
The initial root samples collected from soil cores were transported to the laboratory, washed, and separated into fine (≤2 mm diameter) and coarse root (>2 mm) fragments using different mesh-sized sieves and Vernier calipers (Majdi et al. 2005).The fine root biomass samples were oven-dried for 48 hours at 65 º C and weighed with the ±0.001 g balance.The sixteen samples of initial fine root biomass weights were then averaged and calculated as megagrams of dry biomass per hectare for each plot.A total of 64 ingrowth bags (four bags from each plot) were collected at three-month intervals up to a year.Live and dead roots were separated using forceps and lenses looking at the differences in the characteristics like color, turgidity, flexibility, and water floating.Fine roots (≤2 mm diameter) were then washed, oven-dried, and weighed.
Fine root production (Mg C ha -1 ) was estimated as the sum of dead and live root biomass from the positive increments method (eq. 1) (Neill 1992).In the positive increments method, we summed up all the positive increments of biomass and necromass between two or more successive sampling dates and converted them to annual production (Jourdan et al. 2008).Mean standing live fine root biomass is the average of live root biomass obtained from different collections during the whole experimental period (eq.2).Fine root turnover rate (yr -1 ) was estimated as the ratio between positive increments and mean standing live fine root biomass (eq. 3) ------------------------eq. 1

PI = (∑ P − P )
Where PI is the positive increments of fine roots cumulative over the sampling period (g m -2 yr -1 ), Pi is the fine root production after i period of regrowth (where i equals 1, 2 …n), D is the number of days throughout the study period after installation of ingrowth cores, MSB is the mean standing biomass (g m -2 ) of fine root over the sampling period, LB1, LB2, LB, and LBn indicate the live root biomass of each sampling period, and TR denotes the turnover rate.

Tree measurements
All the trees with at least 1 cm diameter at breast height (DBH, 1.3 m above ground level) were measured in a 400 m 2 plot, while trees with 5 cm DBH or more were measured in a 1000 m 2 plot.Smaller and shorter trees that did not reach a diameter of 1 cm at a height of 1.3 m from the ground level were discarded.DBH values were used to calculate basal areas of individual trees and summed to plot level basal area as m 2 ha -1 .Tree counts were converted to tree density (numbers of trees per hectare).The coarse root biomass of the trees was estimated using published allometric equations (Cairns et al. 1997).The floristic list of the species can be found in Aryal et al. (2024).

Statistical analysis
Fine root production and turnover data obtained in grams from 8 cm diameter and 30 cm long cylindrical ingrowth bags were transformed to megagrams per hectare (Mg ha -1 ).Adjustments to 365 days were made to present the annual productivity where the duration of the experiment differed from the exact one-year period.Four samples of fine root production and turnover from each experimental plot were averaged to obtain the plot-level mean at each collection period.A total of 16 plots (four plots × four forest successional categories) were considered in statistical analysis.The data were tested for normality using the Shapiro-Wilk test.Since all the datasets met or were found near to fulfilling the ANOVA assumptions, no transformations were made.Oneway ANOVA was used to test the significant differences between forest successional categories.The repeated measures ANOVA was applied to test the significant differences in fine root production and turnover between forest growth stages (age classes) throughout the sampling period.Tukey´s post hoc test was performed to compare means between forest successional categories.Monthly rainfall data obtained from nearby meteorological stations were plotted in the graph to align with the seasonal variation of fine root production.Multiple regressions as well as simple linear regression analyses were performed to identify the significant predicting variables of fine root production such as basal area and tree density.

Initial fine-root biomass
Initial fine root biomass ranged from 2.7 to 3.9 Mg ha -1 which did not change significantly among forest successional categories (Table 1).It is the amount of dry fine root biomass obtained from soil cores before installing the ingrowth bags.Total root biomass estimated by using an allometric equation showed a gradual increase with forest age.
The values ranged from 7 to 51 Mg ha -1 with primary forest being the highest.Initial fine root biomass contributed 8 to 38 percent of total estimated coarse root biomass which was higher in younger secondary forests and lower in primary forests.The basal area of the trees of ≥1 cm diameter at breast height (DBH) also increased gradually with increasing forest age which ranged from 8 to 33 m 2 per hectare (Table 1).

Fine root production and turnover
Primary forests showed a higher fine root production compared to secondary forests with a slightly increasing trend with forest age.However, there were no significant differences between forest successional classes in fine root turnover.
The average fine root production rate ranged from 1.9 to 2.8 Mg of biomass ha -1 yr -1 while turnover over rate ranged from 1.1 to 1.5 yr -1 (Fig. 3).Fine root production reached about 60 -70% of initial fine root biomass in primary as well as secondary forests.The proportion of fine root turnover rate to fine root production rate ranged from 0.41 to 0.68 with the highest value in the young secondary forest of 5 years and the lowest value in primary forests.This showed that relative fine root turnover is faster in younger forests compared to older ones.Among different estimations of fine root production, long-term core values indicate the amount of fine root biomass produced in the final ingrowth bag samples which was of 13 months duration in our experiment.Although the numerical mean value seems to be higher in primary forests, fine root biomass from long-term cores did not change significantly (Table 2).Mean standing biomass is the average of live fine root biomass from all different collections throughout the sampling period.Primary forests showed a significantly higher mean standing biomass compared to secondary forests.This showed that the fine root mortality can be slower in mature forests which were also reflected by lower turnover rates.

Seasonal variation of fine root production
After the establishment of ingrowth bags in February, fine root production increased in May and decreased in August.August was a rainy month and probably the fine root mortality during summer rainy months was higher.Fine root production again increased in December, when the amount of rainfall decreased before reaching to dry season (Fig. 4).Then after, the production trends varied among forest successional stages.Primary forest and younger secondary forest of 10 years showed a slightly decreasing trend while secondary forest of 5 and 20 years showed a slight increase in fine root production.
In a multiple regression analysis, considering tree density and basal area as independent variables and fine root production as a dependent variable, we found that only basal area was the significant predictor of the variation in fine root production (β = 0.5, ρ<0.05).Tree density was not considered a significant predicting variable (Table 3).The linear relationship between basal area and fine root production showed an increasing trend of fine root production with the increase in basal area of the trees of ≥1 cm DBH (Fig. 5).Because of the abundance of outliers in the graph, we assume that other variables also affect fine root production in such forest ecosystems which we did not evaluate in this study.

DISCUSSION
Our results demonstrated that the ratio of fine root biomass to estimated total root biomass decreased gradually with forest age.This trend can be explained by the fact that carbon allocation in coarse root increases with the increase in tree size, aboveground biomass, and forest successional age (Zhou et al. 2020).At the same time, the relative fine root turnover rate (ratio between fine root turnover and initial fine root biomass stock) also decreased with forest age providing the basis to accept the hypothesis that younger secondary forests are faster in belowground carbon turnover compared to older secondary and primary forests (Puglielli et al. 2021;Titlyanova and Shibareva 2022).However, other factors such as soil fertility, plant species composition, and the changes in aboveground net primary productivity can determine fine root production and turnover (Becknell et al. 2021;Ma et al. 2022).Younger secondary forests were dominated by soft-wood density species, while older secondary and primary forests were dominated by dense-wood species (Aryal et al. 2024).Aboveground litter production and decomposition rates were affected by leaf traits that varied with forest successional ages in these plots (Sánchez-Silva et al. 2018).These variations of tree species composition and related changes in plant traits can explain the changes in belowground functioning, including fine root productivity and turnover.A shift from more acquisitive traits to conservative traits on root characteristics can explain our results of lower turnover during older phases of forest succession (Hogan et al. 2023).
The changes in soil fertility, nutrient availability, and relative abundance of nitrogen-fixing plants along the forest succession also determine the variation in fine root productivity (Figueiredo et al. 2024).Fine root production rates in tropical forests obtained from different published studies varied from 1.8 to 13.6 Mg ha -1 yr -1 while the stock varied between 0.9 and 24.3 Mg ha -1 (Castellanos et al. 2001;Finér et al. 2011a;Girardin et al. 2013;Huaraca Huasco et al. 2021).Our results of 1.9 to 2.8 Mg per hectare per year of fine root production and 2.7 to 3.9 Mg ha -1 of initial fine root biomass stocks are found within the reported ranges.Castellanos et al (2001) in a study in tropical dry forests of Jalisco, Mexico reported a fine root production rate of 1.8 Mg ha -1 yr -1 to a depth of 10 cm and stated that burning in slash and burn agriculture decreased the fine root production.The fine-root turnover rate reported in a semi-arid tropical forest of India was 0.65 to 0.97yr -1 and was reported to be important to maintain site productivity (Jha and Mohapatra 2010).Although some studies reported an increasing trend of fine root biomass along the successional gradient with the increase in forest age, our results showed that the difference among successional categories was statistically non-significant (Yang et al. 2010;Hogan et al. 2023).Seasonal climatic variations also affect fine root production rates (Gill and Jackson 2000;Sánchez-Silva et al. 2022).Some studies reported higher fine root production during the wet season (Chen et al. 2004;Morales Ruiz et al. 2021) but higher fine root production was observed in drier months compared to wet months in our study which was similar to the findings reported in Amazonian forests (Lima et al. 2010).This variation can be explained by the fact that plants shift carbon allocation towards roots where photosynthates are used to increase water uptake when water is a limiting resource, resulting in plants with higher fine root production in the dry season (Metcalfe et al. 2008;Umaña et al. 2021).
Aboveground forest structural parameters like tree size, tree density, basal area, and species composition can be related to fine root production and turnover (Finér et al. 2011a;Barbhuiya et al. 2012;Weemstra et al. 2020;Zeng et al. 2020).The basal area of the trees (≥1 cm DBH) explained about 50% of the variation in fine root production in our study.In a global compilation report, it is mentioned that tree basal area predicted 49% of fine root biomass (Finér et al. 2011b).However, the correlation between fine root production and aboveground net primary productivity varies with forest type, site conditions, and nutrient distribution at different soil horizons (Nadelhoffer and Raich 1992;Ding et al. 2021).
Some studies mention that fine root dynamics depend on soil properties like texture and available nutrients (Hertel et al. 2013).In an analysis of some soil parameters, soils in our experimental plots are found relatively homogenous (Aryal et al. 2015), thus, we assume that the variation in fine root production caused by soil variation is not significant.Some of the reports mentioned that understory vegetation can contribute to 20 -30% of the fine root biomass (Finér et al. 2011b).However, we could not separate the contribution of tree and understory vegetation in fine root production and turnover due to logistical limitations.We recommend evaluating the role of understory vegetation in future studies of this type.

Implications of plant succession in future estimates of carbon dynamics in tropical environments
Earlier studies showed that the aboveground carbon and species composition recovery period of these secondary forests was more than 100 years after the cessation of slash-and-burn agriculture especially due to the successional shift of softwood species to hardwood species (Aryal et al. 2024).
The responses of such changes can also be noted in the fine root dynamics of the vegetation.It is interesting to note that the results of the fine root production in our study did not show any statistical difference between secondary and primary forests indicating that fine root production of secondary forests gets near to primary forests from the early stages of succession (Sánchez-Silva et al. 2022).However, the relative turnover rates were found higher in secondary forests than in primary forests.This showed that the growth and dieback of the fine roots are faster in secondary forests which, in turn, helps secondary forests to recover their soil organic carbon earlier.However, the durability of belowground biomass carbon can be shorter in secondary forests compared to primary forests.Our results can be considered as one of the important bases regarding fine root production and turnover for modeling carbon dynamics of the tropical secondary and primary forests in the region.If combined with aboveground litterfall monitoring, annual live biomass increments rates, and soil organic carbon estimations (Aryal et al. 2017(Aryal et al. , 2022;;Sánchez-Silva et al. 2018), this study provides strong evidence regarding the successional changes in tropical forest carbon dynamics useful for carbon measurement, reporting and verification (MRV) as proposed in reduction of emissions from deforestation and degradation plus conservation (REDD+) strategies.
Here, we did not analyze the effect of fine root turnover on soil organic carbon storage but a study in the tropical forests of Panama reported that fine root production and turnover contributed to organic matter addition in surface soil but this effect was minimal in the deeper soil horizons, which can be explained by the decline of fine root turnover rates with soil depth (Cusack et al. 2021;Cordeiro et al. 2020).Fine root biomass production and turnover contributed to increased particulate and readily oxidizable organic carbon in the soil, resulting in total soil organic carbon in a subtropical forest (Xiang et al. 2022).Fine root turnover makes a greater contribution to organic carbon accumulation in soil aggregates than aboveground litterfall during the course of forest succession because of the distribution of root biomass in the soil profile (Shi et al. 2023).

CONCLUSIONS
Fine root productivity in tropical secondary forests reached near to primary forest level from the early stages of succession (~5 years).However, fine root turnover rates were faster in secondary forests than in primary forests accepting the hypothesis of faster carbon cycling in younger phases of forest growth than in older phases.Aboveground forest structural parameters like tree basal area explained about 50% of the variation in fine root productivity.Fine root production was higher in the dry season than in the wet season in this study.The results of this study can be used for belowground carbon balance modeling of the successional forests.The information also contributes to generating a baseline for future studies on below-ground carbon dynamics in the succession of tree species from the conversion of an agricultural system to forests in tropical environments.Campeche for their guidance during experimental design and setup.Landowners from two communities of Calakmul municipality: Cristobal Colon and El Carmen II permitted to establish experimental plots in their field.We also thank Demetrio Alvarez, Antonio Ramirez, Armando Velazquez, Victoria Hernandez, Beatriz Peña, and Alan Pacheco for their help in the field and laboratory work.

Funding. Mexican National Council of Humanities
Sciences and Technologies (CONAHCyT) provided the scholarship funding (No. 316709;2011-2014) to the first author.
Conflict of interest statement.The authors declare that there is no conflict of interest.
Compliance with ethical standards.Due to the nature of this work, this research does not require approval from an ethical committee.

Figure 1 .
Figure 1.Location of the study site and distribution sampling plots according to the forest age class.CC = Cristobal Colon, EC = El Carmen II.SF = secondary forests and PF = primary forests (source: original from authors).

Figure 2 .
Figure 2. Collected ingrowth bag (left) and fine root separation (dead and live) from the soil (right) (source: original from authors).

Figure 3 .
Figure 3. Mean (upper 95% confidence interval) values of fine root production and turnover (Mg C ha -1 yr -1 ) among four stages of forest growth.SF = advanced secondary forest, PF = primary forest.

Figure 4 .
Figure 4. Fine root production (Mg C ha -1 yr -1 ) at different months during the period of sample collections (from February to March).SF = secondary forests, PF = primary forests.Vertical bars represent the 95% confidence intervals.The upper graph shows the monthly rainfall pattern during the study period, recorded in Zoh Laguna metrological station, about 30 km from the study plots.

Figure 5 .
Figure 5. Relationship between fine root production and basal area of the trees with ≥1 cm DBH, measured in tropical secondary forests of Calakmul, Campeche, Mexico.

Table 1 . Average (± 95% confidence interval) initial fine root biomass, estimated total root biomass, and basal area of trees ≥1 cm DBH measured before the establishment of ingrowth bags at different stages of forest growth in Calakmul, Campeche, Mexico. Forest age class Initial fine root biomass, Mg ha -1 , estimated by soil coring method Total root biomass, Mg ha -1 , estimated by using allometric equations
ADifferent letters followed by the numbers indicate significant differences among forest successional classes.SF = secondary forests, PF = primary forests.

Table 2 . Average (± 95% confidence interval) fine root production estimated from long-term cores, positive increments, and mean standing biomass at different stages of forest growth in Calakmul, Campeche, Mexico. Different letters in the exponent denote significant differences among forest age classes.
Different letters followed by the numbers indicate significant differences among forest age classes.SF = secondary forests, PF = primary forests.