Next Article in Journal
Hydrology and Carbon Flux Interconnections in a Hemiboreal Forest: Impacts of Heatwaves in Järvselja, Estonia
Next Article in Special Issue
Degradative Activity of Five Basidiomycota Strains on Sweet Chestnut Wood
Previous Article in Journal
Estimation and Spatial Mapping of Soil Carbon Stock in the Perigi, South Sumatra, Indonesia, Considering Peat Depth Variability
Previous Article in Special Issue
Determinants of Needleleaf and Broadleaf Decomposition Rates Under and Outside the Parent Tree Stand
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 17
Issue 3
10.3390/f17030300
forests-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Domestication Level and Soil Fertility Differentially Alter Soil Carbon Sequestration Potential in Breadfruit (Artocarpus)

by
Lindsey Gohd
1,†,
Louise M. Egerton-Warburton
1,*,†,
Ellinore Porter
2,
Noel Dakar Dickinson
3,4,
Nyree J. C. Zerega
1,5 and
Ray Dybzinski
6
1
Program in Plant Biology and Conservation, Northwestern University, Evanston, IL 60208, USA
2
Department of Biology, Fort Lewis College, Durango, CO 81301, USA
3
National Tropical Botanic Garden Breadfruit Institute, Kalāheo, HI 96741, USA
4
Natural History Museum Denmark, University of Copenhagen, 1307 Copenhagen K, Denmark
5
Negaunee Institute for Plant Conservation, Chicago Botanic Garden, Glencoe, IL 60022, USA
6
School of Environmental Sustainability, Loyola University Chicago, Chicago, IL 60660, USA
*
Author to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Forests 2026, 17(3), 300; https://doi.org/10.3390/f17030300
Submission received: 20 January 2026 / Revised: 22 February 2026 / Accepted: 24 February 2026 / Published: 26 February 2026
(This article belongs to the Special Issue Litter Decomposition and Soil Nutrient Cycling in Forests)
Browse Figures
Versions Notes

Abstract

Plant domestication studies have traditionally focused on morphological factors that are under direct selection, e.g., fruit size, overlooking the consequences of domestication on ecosystem services. We addressed this knowledge gap by documenting for first time the soil carbon (C) sequestration potential in wild relatives and domesticated cultivars of breadfruit (Artocarpus), a long-lived tree crop. We evaluated aggregate-bound and bulk organic C pools in breadfruit wild relatives and domesticates in soils that varied in nitrogen (N) and phosphorus (P) fertility with management practices (fertilizer and mulch). We determined whether C levels were linked to plant domestication, abiotic factors (N, P, pH, and texture), or biotic factors with known links to C accrual (arbuscular mycorrhizal fungi (AMF), and microbial biomass). In low N or N: P soils, increasing breadfruit domestication was associated with reductions in macroaggregate C (by 50%) and bulk C (host determinism); these shifts were associated with AMF hyphal productivity (50% lower than in wild relatives), soil N and P, and microbial biomass. With a high soil N fertility, the levels of aggregate and bulk soil C were similar between wild relatives and domesticates (plasticity). Despite the limited number of cultivars sampled (n = 10) and the different management practices among sites, our findings suggest domestication effects on ecosystem services, especially those modulated by AMF and soil N fertility. The calculated soil C stocks averaged 99.5 Mg C/ha (range 70–122 Mg C/ha), supporting the possibility of C accrual in breadfruit agroforestry.

Graphical Abstract

1. Introduction

The rapid onset of climate change impacts coupled with the growing need for sustainable food production has brought increased attention to agroforestry as a strategy to raise soil organic carbon (SOC) levels without compromising food production [1]. Agroforestry practices using native or non-native forest trees is recognized as a solution for climate mitigation and a substantive sink for atmospheric C [2]. In the tropics, most studies have focused on economically important tree commodities including mango [3], coffee [4], macadamia [5] and cacao [6], and have shown these systems to sequester SOC [2]. The SOC sequestration potential in other perennial crop trees in tropical agroforestry systems is poorly understood. Here, we examined the potential for SOC sequestration in breadfruit (Artocarpus) in the Hawai’i Islands.
Breadfruit (Artocarpus altilis (Parkinson) Fosberg) is a large (up to 30 m tall) and long-lived (80–100 years) tree crop that has played a key role in traditional agricultural systems and food security in Pacific Island communities for centuries [7]. Breadfruit also has a long domestication history. Over thousands of years, human migration across Oceania, coupled with the progressive selection and vegetative propagation of breadfruit cultivars, has resulted in a strong west-to-east domestication gradient (Figure 1A). Domestication has led to substantial phenotypic changes above-ground, notably, a greater leaf area, photosynthetic rate, and larger fruit size, and an absence of seeds (due to triploidy) compared to wild relatives [7,8]. Breadfruit’s closest wild relatives are Artocarpus camansi Blanco, which is endemic to the forests of New Guinea, and A. mariannensis Trécul, a species native to the Mariana Islands and Palau. The earliest breadfruit domestication likely occurred in Melanesia while the more recently derived cultivars (i.e., highly domesticated and seedless) predominate in East Polynesia. In Micronesia, some breadfruit cultivars are A. altilis × A. mariannensis hybrids.
Apart from food security, breadfruit is increasingly recognized as having considerable potential in climate resilience [10]. The trees are readily propagated and fast-growing in marginal habitats, store substantial amounts of C in the aboveground biomass [11], and provide soil erosion control [12]. There has also been a recent resurgence in the cultivation of breadfruit, providing substantial opportunities for economic and environmental sustainability [7]. To the best of our knowledge, the soil C sequestration potential in breadfruit has not been examined.
Practices such as agroforestry using perennial tree crops like breadfruit are expected to produce greater net SOC sequestration than conventional monoculture crops or integrated tree–crop systems [13]. In part, agroforestry practices enhance SOC by stimulating plant growth and litter inputs [14]. The deep, extensive root systems of trees can deliver fine root residues and exudates into SOC pools in deeper soil layers. Further, fine roots may host arbuscular mycorrhizal fungi (AMF; Glomeromycotina), a symbiosis that plays a key role in the transport of nutrients from the soil to host plants in exchange for photosynthates. The AMF symbiosis also stimulates C flow from plants to the soil system to impact SOC stabilization and microbial metabolism [15]. Mechanistically, AMF external hyphae play a key role in stabilization by translocating plant-derived C into the soil and driving soil aggregation, one of the major processes that leads to SOC sequestration [16]. External AMF hyphae enmesh soil particles and fine roots while AMF exudates (glycoproteins) bind soil particles, plant residues, and microbial necromass into macroaggregates (250–2000 µm diameter) [17]. Over time, the organic matter within macroaggregates condenses onto clay particles, leading to the formation of C-rich microaggregates (53–250 µm) in which C is physically protected from microbial oxidation [18]. Thus, AMF communities with a higher hyphal growth can increase aggregation and AMF-derived C inputs into the SOC pool [19].
While these AMF functions are key to C sequestration, the studies, to date, in crop species have emphasized the effect of domestication on intraradical (root) AMF colonization. For example, AMF root colonization is lower in modern crop cultivars in comparison with their wild relatives, including breadfruit [20,21,22]. This outcome may reflect trade-offs between plant C investment in aboveground biomass versus roots and AMF symbioses [20,23] or ploidy- and domestication-associated shifts in root traits that can influence AMF colonization, e.g., architecture, and tissue quality [24,25,26]. In addition, wild crop relatives harbor more diverse and mutualistic AMF communities than their modern or highly domesticated counterparts [21]. Such studies have led to the view that highly domesticated cultivars have a limited ability to recruit or support AMF compared with their wild crop relatives (host determinism [27]).
At the same time, AMF responses to domestication may be modified by abiotic factors and environmental conditions. Shifts in precipitation and temperature as well as soil properties (pH, N, and P) can shape the strength of the plant–AMF association by altering the AMF community composition [28,29], hyphal abundance, and root colonization [23,30]. Collectively, studies have shown that crop wild relatives can maintain AMF root colonization and growth benefits in highly fertile soils whereas domesticated crops can show reduced AMF root colonization and plant growth [20,31,32] because the root systems are adapted for the capture of readily accessible mineral N (host insensitivity [33]).
More recent syntheses have explored how domestication is expected to alter belowground C stabilization pathways. AMF hyphal productivity and turnover (necromass), microbial biomass, and fine root exudates are the biological levers that both actively shape the rhizosphere to mobilize nutrients and represent substantial inputs of organic C into the rhizosphere [34]. Domesticated host plants, with their genetically altered patterns of C allocation to roots (AMF) and root exudates (microbial biomass), may rewrite these ecological roles. Alternatively, soil nutrients, especially N, and management (fertilization and mulch) may increase C accrual by promoting microbial activity [35].
In this study, we used a domestication-gradient test in a perennial tree crop (breadfruit; Figure 1A), examined the linkages between AMF, soil aggregate C, and bulk SOC patterns, and tested whether the effects of high soil N fertility outweigh those of domestication. This approach has not been applied to breadfruit.
We documented the AMF abundance (root, and external hyphae), soil C pools (microbial biomass, labile C, and aggregate C), and bulk SOC in managed breadfruit orchards with low and high soil N fertility since soil N is tightly coupled with plant productivity and terrestrial C cycling [36] and can alter the cost–benefit balance of AMF [15,23,37].
We posed three questions:
  • How are AMF abundance and function in breadfruit influenced by edaphic conditions (abiotic constraints)?
  • To what extent are (a) AMF abundance and (b) functioning (aggregation, and aggregate C) mediated by domestication history (host determinism)?
  • Does breadfruit domestication history and soil N fertility converge to affect AMF functions in soil aggregate C and SOC accrual (host insensitivity)?
We detected reductions in AMF abundance and function (aggregate C) with increasing breadfruit domestication, and variations in bulk SOC levels among cultivars, consistent with host determinism. AMF productivity and soil N and P were significantly corelated with aggregate and bulk soil C levels, while microbial biomass and leaf C were correlates of bulk SOC. In high N soils, however, AMF productivity and levels of aggregate C and bulk SOC were similar among domestication levels, findings more consistent with plasticity than host insensitivity.

2. Materials and Methods

2.1. Study Sites and Species

Our research was conducted in Hawai’i in four study sites across three islands: Maui, Kaua’i, and O’ahu (Figure 1B; Table 1). These sites occur in the atmospherically humid lowlands and experience a tropical climate with annual precipitation ranging from ~1580 to 1980 mm yr−1, most of which is received between October and April. The mean annual temperature ranges from 20.1 °C (min) to 22.9 °C (max) with little variation within or across years [38].
Two sites, McBryde Garden and Kahanu Garden, are part of the National Tropical Botanical Garden (NTBG). McBryde Garden on Kaua’i (MCB; Figure 1C) was established in 1970 to house a subset of core breadfruit species and cultivars selected from NTBG’s broader breadfruit collection. Many breadfruit trees are grown alongside other crop trees as a food forest. The trees are fertilized with Suståne 4-6-4 organic fertilizer (Suståne Corporation, Cannon Falls, MN, USA) and breadfruit leaf compost. Kahanu Garden (KAH; Figure 1D) was established on Maui in 1972 to house an orchard representing accessions of breadfruit (A. altilis, A. altilis × A. mariannensis hybrids), breadnut (A. camansi), and dugdug (A. mariannensis) originating from 34 Pacific islands, the Philippines, Indonesia, and the Seychelles. The 4.86 ha garden is fertilized with Suståne 4-6-4 organic fertilizer, a soil conditioner, EM-1TM (TeraGanix, Alto, TX, USA), and composted breadfruit leaves.
‘Ulutopia (ULU) was established on Kaua’i Island on the campus of the Kaua’i Community College in 2014 (Figure 1E). The 64-tree orchard of Ma’afala breadfruit trees provides opportunities for student research, training, and education. The trees are fertilized using 6-6-6 N:P:K (B. Yamamoto, pers. comm.). The Kāko‘o ‘Ōiwi site (KKO; Figure 1F) in O’ahu is managed by a community-based non-profit that restores sustainable food forests in local communities. The site was established in 2006 and comprises 2.83 ha of Ma’afala breadfruit trees in restored drylands. The trees are fertilized with a mixture of potash and bone meal (N. Reppun, pers. comm.).
The four sites spanned three soil orders, and their classification is presented in Table 1. Soils were dark brown to reddish-brown, finely-textured silty clays and all contained an abundance of macroaggregates (80%–90%) relative to microaggregates (9%–11%; Table S1). Large macroaggregates (2–4 mm diameter) comprised the bulk of the aggregate pool (40%–60%). Aggregate stability was similar among sites (MWD 1.9–2.2 mm; Table S1). Soils were all neutral to slightly acidic but showed key differences in fertility. The high-fertility KAH soils were characterized by a low bulk density, the highest CEC, and significantly higher levels of soil N, bulk N, N:P supply ratio, and total P in comparison to the other three sites (Table 1). The remaining sites were considered less productive due to nutrient limitations, including lower N fertility, N:P supply ratios, and exchange capacities, typical of highly weathered (MCB and ULU) or poorly developed soils (KKO). Weathering resulted in an enrichment in secondary Fe and Mn oxides in all sites, aluminosilicates in KAH [39], and reductions in total Ca in older sites. Soil Ca levels were significantly and positively correlated with total P (rs = 0.593), Fe (rs = 0.620), Mn (rs = 0.755), and Al (rs = 0.604; Spearman-rank correlation, p < 0.05).
To tease apart the effects of soil properties and breadfruit domestication on AMF and soil C pools, we undertook three sets of analyses. First, we explored the effect of soil fertility alone (Question 1). To limit the effects of plant genotype, we sampled and analyzed a limited number of soil samples collected under trees of a commonly planted cultivar (Ma’afala) established in all four study locations (Table 2). Next, we tested the effect of domestication level alone (Question 2). We sampled and analyzed soil under ten different breadfruit cultivars grown in low N soils in MCB (Table 2) to reduce any potentially confounding effects of soil N fertility on AMF abundance. Finally, we tested the interplay of soil N fertility and domestication (Question 3) by sampling and analyzing soil under the ten different breadfruit cultivars that had been cultivated in both MCB (low fertility) and KAH (high fertility; Table 2 and Table S2).
The ten breadfruit cultivars sampled represent the geographic range of breadfruit in Oceania along the west-to-east domestication gradient ([9] Figure 1A). Artocarpus camansi and A. mariannensis represent the closest wild relatives (i.e., non-domesticated) species of breadfruit (D0) and produce small fruits with numerous large seeds. Artocarpus altilis is the most widely distributed domesticated species and includes both seeded (diploid) and seedless (triploid) cultivars, including hybrids. Early diploid domesticates containing a few seeds (D1) are common in Melanesia, diploid domesticates that rarely have seeds (D2) are found in the western Polynesia, and seedless triploid cultivars (D3) are predominant in eastern Polynesia and Micronesia. All cultivars represent clonally propagated material that has preserved the varietal traits of the initial selection, and the trees were large and productive (Table S2). Sample sizes ranged from one (Puou) to four trees per cultivar (Ma’afala) based on the availability of each cultivar within a site (Table 2). As a result, we primarily explore domestication-level effects on soil properties and provide cultivar-level analyses in the Supplementary Materials. All cultivars showed variations in leaf C content but similar leaf C:N levels (Table 2). On the leaf economics spectrum, breadfruit litter is classed as high-quality (low C and high N content) and readily decomposable. However, nutrient returns to the soil and subsequent tree growth may depend on environmental conditions [40].
Table 2. Domestication status, Artocarpus species, geographic origin, and fruit and leaf traits in each cultivar evaluated in the study. For leaf C and C:N, means with standard error are in parentheses.
Table 2. Domestication status, Artocarpus species, geographic origin, and fruit and leaf traits in each cultivar evaluated in the study. For leaf C and C:N, means with standard error are in parentheses.
Status (Ploidy) ‡CultivarArtocarpus sp.Origin (Region) Fruit Size (kg) †Leaf
% C §		Leaf
C:N §		Number of Trees Sampled
KAH ˆMCBULUKKO
D0 (2n)BreadnutA. camansiPapua New Guinea (ME)0.8–1.240.3 (1) ab18 (1)33--
D0 (2n)DugdugA. mariannensisMariana Isl (MI)0.7–3.141.9 (1) a17 (1)32--
D1 (2n)Ulu fitiA. altilisSamoa (WP)1.1–2.841.0 (1) ab15 (2)33--
D2 (2n)Ma’afalaA. altilisSamoa (WP)0.4–1.337.6 (3) c19 (5)3443
D2 (2n)PuouA. altilisCook Isl (EP)0.7–2.541.0 (1) ab15 (2)41--
D3 (3n)OteaA. altilisSociety Isl (EP)1.4–2.539.0 (2) bc15 (1)43--
D3 (3n)‘UluA. altilisHawaii (PO)1.3–5.441.3 (1) ab17 (4)33--
D3 (3n)MeitehidA. altilisPohnpei (MI)~1.241.3 (1) ab16 (3)11--
D3 (3n)WhiteA. altilisSeychelle Isl (IO)1.2–5.040.9 (2) ab17 (1)32--
D3 (3n)MidolabA. altilis ×
A. mariannensis		Palau (MI)0.7–2.039.4 (2) ab15 (2)21--
‡ Domestication status: D0, wild relatives with many large seeds; D1, early domesticates with small seeds; D2, early domesticates with few or no seeds; D3, highly domesticated cultivars with no seeds. Region: ME, Melanesia; MI, Micronesia; WP, West Polynesia; EP, East Polynesia; PO, Polynesia; IO, Indian Ocean. † Fruit size; see Jones et al. 2011 [41]; § This study: leaf C, ANOVA p = 0.003, means within column with the same letter do not differ significantly at p < 0.05 (LSD test); leaf N 2.54 ± 0.64%, ANOVA p = 0.060; leaf C:N ANOVA p = 0.107; ˆ Site: KAH, Kahanu; MCB, McBryde; ULU, Ulutopia; KKO, Kāko’o‘ Ōiwi.

2.2. Soil and Leaf Tissue Sampling

For an individual tree, soil cores (5 cm diam, 15 cm deep) were collected in each of the cardinal and inter-cardinal directions and inside the dripline. These eight cores were then pooled in a Ziploc bag to form a single bulk sample that accounted for spatial variability in soil properties around each tree. This approach resulted in comparison and inferences at the individual tree level. Large materials (e.g., rocks, leaves, and sticks) were removed and soils were gently homogenized. Soil samples were then dried in an Excalibur drier (~66 °C) for three days to meet soil import and transfer requirements between Hawai’i and Illinois (Excalibur, Sacramento, CA, USA). A 100 g sub-sample from each bag was then packaged and shipped to the Chicago Botanic Garden. For each tree, we also collected a 10 cm × 10 cm sub-sample of tissue from a single mature leaf. Each sample was placed in an individual envelope and placed over silica gel to dry. Leaf samples were returned to the Chicago Botanic Garden, dried to constant weight (80 °C), ground, and analyzed for C content by combustion (see Section 2.3 below).

2.3. Soil Analyses

Soil texture was determined using the pipette method [42]. Soil aggregates were extracted from each sample using wet sieving on a set of nested sieves: 2000 µm (large macroaggregates, >2000 µm diameter), 250 µm (small macroaggregates, 250–2000 µm), and 53 µm (microaggregates, 53–250 µm [43]. Aggregate fractions on each sieve were dried (80 °C) for 48 h and weighed, and mean weight diameter (MWD) calculated [44]. All fractions were ground to a fine powder and analyzed for organic N and C by combustion in a Leco TruSpec™ CN Elemental Analyzer (Leco Corp., St. Joseph, MI, USA). Prior to combustion analysis, each sample was tested qualitatively for inorganic carbonates using the hydrochloric acid effervescence test; none of the samples reacted with acid. Aggregate C content was corrected for sand content and expressed as grams OC per kg soil. Dried subsamples of leaf tissue and bulk soil from each site and cultivar were similarly ground and analyzed by combustion for C and N. The resultant data were used to calculate the C:N ratio and infer the C source(s) within aggregate fractions and bulk soil by comparison with breadfruit leaf C:N data (this study) and published microbial C:N values [45].
We used two methods to analyze soil N and P: direct extraction (nutrient pools), and resin bags to obtain integrative measurements of nutrient availability (nutrient fluxes). For nutrient pools, inorganic N and P were extracted from soil using deionized water (1:10 w/v, soil:water; pH 6.8) and by vigorous shaking for 30 min. Extracts were filtered and analyzed colorimetrically on a microplate spectrophotometer (Biotek Epoch, Winooski, VT, USA) using the vanadium (III) reduction method for NO3 [46], phenol-hypochlorite method for NH4 [47], and malachite green method for PO4 [48]. Soil pH was measured in 1:5 soil:water (v/v) using a pH probe (Fisher Scientific, Waltham, MA, USA). Soil total P and Fe were measured by X-ray fluorescence (Hitachi X-Met 800, Shanghai, China). To measure nutrient fluxes, we installed resin bags (each 10 × 10 cm nylon) and filled with 5 g mixed-bed (anion–cation) beads under trees (Supelco AmberLite MB20; Supelco, Philadelphia, PA, USA) in KAH and MCB. Resin bags were placed in two soil core holes (10 cm deep) at opposite cardinal points of an individual breadfruit tree. After two months, the resin bags were collected, shipped to the Chicago Botanic Garden, rinsed to remove soil organic particles, and individually extracted in 100 mL 2 M KCl. Aliquots of each extract were analyzed for NO3, NH4, and PO4 as described previously.

2.4. AMF and Microbial Analyses

Fine roots were manually picked from each soil sample and stained [49], and AMF colonization quantified with the gridline intersect method using a light microscope (200× magnification [50]. We measured the standing AMF hyphal pool (live + dead hyphae) and AMF productivity, i.e., the active (live) AMF hyphae. The standing hyphal pool was measured in soil samples using sodium hexametaphosphate flotation and filtration onto gridded membrane [51]. Fifty-fields of view per sample were viewed and scored for the presence of AMF hyphae and non-AMF (generalist fungi) hyphae using a light microscope (400×), and the hyphal length of AMF and non-AMF hyphae was calculated [52]. We defined AMF hyphae as non-septate or irregularly septate hyphae with characteristic unilateral elbow-like projections; all other hyphae were categorized and counted as generalist soil fungi. AMF productivity was measured using hyphal in-growth bags. At MCB and KAH, hyphal bags (5 × 5 cm, pore diameter ~ 25 µm) filled with sterilized coarse sand (5 g) were buried 15 cm deep in two soil core holes on opposite sides of each tree. After six months, the bags were collected, and hyphae extracted and analyzed as above.
Total microbial biomass C (hereafter, ‘microbial biomass’) was quantified on each soil sample using the chloroform addition–direct extraction technique of dry soils [53]; this approach provides a result comparable to moist soils [54]. Briefly, duplicate aliquots of each soil sample were incubated in 0.5 M K2SO4 buffer. One sub-sample was treated with chloroform and the other received an equivalent volume of 0.5 M K2SO4 buffer (i.e., control). After incubation and filtering, C levels in each extract were measured using the phenol-sulfuric acid method [55]. Levels of K2SO4 extractable C in the controls were considered representative of the labile C fraction, i.e., soil solution containing carbohydrates derived from soil microbes, enzymes, root exudates, and lysates. This is a measure of C that is easily transportable within ecosystems and contributes to the bulk SOC but may not always map directly with ecological lability [56]. Microbial biomass was calculated as the difference between chloroform-fumigated and non-fumigated subsamples [57].

2.5. Statistical Analyses

All statistical analyses were implemented in the R software environment (v.4.4.1) using the Vegan package (v.2.7-2) unless otherwise noted [58]. Significance was set to p < 0.05. Data sets were tested for homogeneity of variance using Levene’s test, with most factors meeting the assumption of homogeneity (Table S3). Data sets were also tested for normality using Shapiro–Wilks normality tests and residual diagnostic plots, and those that did not meet the assumptions required for parametric statistical tests were sqrt(x) or sqrt(x + 1) transformed prior to analysis.
To test the effects of abiotic conditions alone (Question 1), we analyzed AMF abundance, aggregation, and aggregate and soil C pools in Ma’afala between all four sites using Welch’s one-way analysis of variance (ANOVA; Table S1), cross-validated using ANOVA with Type II sums of squares (Car package, v. 3.1-3). Significant properties were examined using Fisher’s Least Significant Difference (LSD) tests with Bonferroni corrections The relationship between AMF, bulk SOC, aggregate-associated C, and soil properties was analyzed using the Spearman-rank correlation coefficient (rs; Table S4); only correlation coefficient with p < 0.05 are reported. Next, we tested the effects of domestication alone (Question 2; host determinism) using AMF and soil data from MCB trees only (low fertility). Welch’s one-way ANOVA and LSD tests were used to test the effects of domestication level, cultivar identity, and ploidy level on AMF, microbial, and soil properties (Tables S5 and S6).
Finally, we tested for interactions between soil N fertility and domestication level (Question 3; host insensitivity) using data sets from MCB (low fertility) and KAH (high fertility; Tables S5 and S6). Differences in AMF and soil properties were tested using two-way Scheirer–Ray–Hart ANOVAs (site × domestication level, site × cultivar; site × ploidy, Type II sums of squares; Rcompanion package, v.2.5.2). Significant factors were further examined using Dunn’s test (Fsa package, v.0.10.1), and we examined the relationship between AMF and soil properties using the Spearman-rank correlation coefficients. We were also interested in identifying which soil factor(s) influenced C levels within each pool (aggregate and bulk). Owing to the small sample sizes and large number of predictors, we combined KAH and MCB data sets and used penalizing models to prevent overfitting (Glmnet package, v.4.1-10). Lasso regression (Least Absolute Shrinkage and Selection Operator) was employed to select relevant predictors for each C pool. The tuning parameter λ was selected via 10-fold cross-validation, minimizing the mean squared error. All data were standardized prior to analysis. Important soil properties (predictors) were identified by their beta coefficients in the final model and standardized by multiplying each coefficient by the standard deviation of the corresponding predictor (Table S7). We used rank-based estimates of linear regression to develop models for each C pool using soil properties (predictors) with the largest standardized coefficients (Rfit package, v.0.20.0). Models for bulk SOC included predictors and cultivar identity to identify induced breaks, i.e., major shifts in plant allocation patterns or plant–soil interactions owing to changes in ploidy.

3. Results

3.1. How Are AMF Abundance and Function in Breadfruit Influenced by Edaphic Conditions (Abiotic Constraints)?

In our exploratory study, we compared the AMF abundance and aggregate C pools in Ma’afala in the four study sites (Table 1). Root structures consistent with AMF colonization were detected in Ma’afala root fragments in all sites, with 25%–51% of the root length colonized (Table 3). Vesicles were the most common structures observed (21%–33% colonization) along with intraradical (root) hyphae (16%–24% colonization). Arbuscules were only detected at MCB (1% root colonization). The AMF hyphal pool (live + dead hyphae) averaged 9.4 ± 1.2 m/g soil (mean ± se; range 7.5–13 m/g soil). The levels of AMF root colonization and AMF hyphal pools were similar among sites (Table S1).
The aggregate-occluded C in each size fraction was significantly higher in KAH than in MCB, KKO, or ULU (Table 3); a similar pattern also occurred in bulk SOC (Table S1). Despite these differences, aggregate C:N levels did not differ significantly among sites (Table 3). The values averaged 7.9 ± 0.8 (range 6–9) and were lower than C:N levels in bulk soil (C:N 10.3 ± 1.4) or breadfruit leaf litter C:N (15.9 ± 1.2). In addition, labile C and microbial biomass pools did not differ significantly between sites (Table S1).
Aggregate C levels were significantly and positively correlated with soil nutrients including bulk soil N (large, small macroaggregate C), and, more generally, the soil N: P ratio (large macroaggregate, microaggregate C; Table 4). Interestingly, small macroaggregate C was also positively correlated with soil minerals (Ca and Fe) and negatively correlated with silt content (Table 4). Bulk density had a consistent and negative correlation with aggregate C levels (Table 4). Of the AMF factors assessed, only AMF root colonization was correlated with a large macroaggregate C.

3.2. To What Extent Are (a) AMF Abundance and (b) Functioning (Aggregation, and Aggregate C) Mediated by Domestication History (Host Determinism)?

We tested the effects of domestication using data from 10 cultivars grown in a common garden in a low-N-fertility soil (McBryde Garden; Table 1 and Table 2). Increasing domestication in breadfruit resulted in a reduced root symbiotic partnership with AMF (Figure 2); this decline also maps with the significant effects of ploidy (Tables S5 and S6). The AMF hyphal pool (live + dead hyphae), hyphal productivity (live hyphae), and AMF root colonization by C-rich vesicles (Figure 2A–C) were significantly higher in wild relatives (2n, D0) than domesticated cultivars (3n, D1–3). Notably, the AMF productivity in the most highly domesticated cultivars (D3) was less than half that observed in wild relatives (D0) or early domesticates (D1 and D2). At the same time, domestication promoted significant accumulations of generalist (saprophytic) fungal hyphae in wild relatives (Figure 2D), and the abundance of these fungi was negatively correlated with AMF productivity (rs = −0.292).
Domestication significantly reduced the levels of macroaggregate C, but not microaggregate C (Figure 3A,B; Tables S5 and S6). Analogous to the patterns observed in AMF properties, the levels of small and large macroaggregate C were significantly higher in wild relatives than in domesticated cultivars. Small macroaggregate C levels were positively correlated with resin P (rs = 0.498) while large macroaggregate C was negatively correlated with the levels of root fine endophyte (rs = −0.444). The levels of soil labile C also declined significantly with increasing domestication (Figure 3C). However, there was no statistical relationship between the levels of labile C and aggregate-bound C. Instead, labile C was better correlated with AMF root colonization (rs = 0.330).
Bulk SOC varied significantly between domestication levels and among cultivars but was not influenced by ploidy levels (Tables S5 and S6; Figure S1). Of these responses, the most striking was generated by cultivar identity. On the domestication gradient (Figure 4), bulk SOC was significantly higher in wild relatives (D0) than initial domesticates (D1), and was the lowest in early domesticates (D2). With the transition to highly domesticated (D3) or hybrid cultivars (e.g., Midolab), however, bulk SOC levels progressively increased to become equal to or higher than those in the wild relatives. Models confirmed a breakpoint in bulk SOC between D2 Puou (p = 0.023) and D3 ‘Ulu (p = 0.031) (model R2adj = 0.434, p = 0.041; Table S9). Bulk SOC increased with increasing soil N (rs = 0.523) and was negatively influenced by root fine endophyte abundance (rs = −0.444).

3.3. Does Breadfruit Domestication History and Soil N Fertility Converge to Affect AMF Abundance in Relation to AMF Functions in Soil Aggregation, or Aggregate C and SOC Accrual (Host Insensitivity)?

On average, breadfruit cultivars in KAH (high fertility) demonstrated significantly higher levels of levels of AMF hyphal productivity (Figure 2C), macroaggregate- (Figure 3A,B) and microaggregate-bound C, and bulk SOC than in MCB (low fertility; Tables S5 and S6). The levels of AMF root colonization and hyphal pool were similar between sites (Figure 2A,B), while generalist fungal hyphae were less evident in KAH than MCB (Figure 2D).
Notably, in KAH, the domestication level (or cultivar identity) had no significant effect on AMF root vesicles (Figure 2A) or hyphal productivity (Figure 2B), microbial biomass (Tables S5 and S6), or aggregate C or bulk SOC levels (Figure 3A,B; Tables S5 and S6). In addition, aggregate C:N was similar among domestication levels (7.3 ± 0.5, range 6.8–8.3), as was bulk C:N (13.5 ± 1.1; range 10–15) and leaf C:N (16.9 ± 2.4%, range 15.7–17.4); these values were comparable to levels recorded in MCB (Tables S1 and S5) and Ma’afala in KKO and ULU (Table 3). In fact, only three factors differed significantly among domestication levels in KAH, and these responses overlapped with ploidy effects (Tables S5 and S6). With increasing domestication and the transition from 2n to 3n plants, AMF root colonization (Tables S6 and S11) and labile C declined (Figure 3C), while the abundance of generalist (saprophytic) fungi increased (Figure 2D).
The variations in aggregate C and bulk SOC between KAH and MCB were best explained by a combination of soil fertility, AMF, and microbial predictors (Table 5 and Table S8). Notably, consistent positive predictors of these C pools were AMF productivity, soil N, and soil P (plant available and resin), supporting a tight coupling between soil properties and AMF in C accrual. Additional positive predictors included microbial biomass for large macroaggregate C and bulk SOC. Two significant negative predictors were identified for bulk SOC, namely, leaf C and general fungi (saprophytic) hyphal length, indicating that declines in bulk SOC were potentially associated with the decomposition of leaf C residues.
In general, bulk SOC was significantly positively correlated with small (rs = 0.848) and large macroaggregate C (rs = 0.538). However, it is important to note that aggregate-occluded C accounted for ~48% of the bulk SOC pool, meaning that more than half of C likely resides within the soil matrix as particulate organic matter and/or is associated with the mineral fraction [59].
Interactions between the soil fertility and domestication level had no significant effect on macro- or microaggregate abundance (Tables S5 and S6). On average, aggregate pools comprised 63 ± 8% large macroaggregates, 28 ± 7% small macroaggregates, and 9 ± 1% microaggregates (Table S1), a pattern similar to those reported in other tropical soils [59,60]. While microaggregate abundance was correlated with leaf C levels (rs = −0.423), there was no statistically significant correlation between aggregate abundance in any size class and AMF hyphal length, hyphal productivity, or root colonization (Table S4).

4. Discussion

Domestication studies have traditionally focused on plant morphological factors that are under direct selection, such as fruit size [61], overlooking the consequences of domestication on ecosystem services. Here, we address this knowledge gap by documenting the SOC sequestration potential in breadfruit. Using a widely planted domesticated cultivar (Ma’afala), we identified soil physicochemical properties (bulk density and soil N:P ratio) that covaried with aggregate C levels (Question 1). Using a common garden approach (McBryde Garden), we found reductions in AMF abundance and functioning (aggregation and aggregate C) with increasing breadfruit domestication and variance in bulk SOC levels among cultivars, consistent with host determinism (Question 2). We also noted that the effects of high soil N levels (Kahanu Garden) appeared to exert a larger effect on AMF or SOC than domestication, a finding more in line with plasticity than host insensitivity (Question 3). Despite the small suite of breadfruit cultivars and limited sample (tree) replication, our results support the possibility of plant domestication effects on ecosystem services. In the following sections, we first place our findings within the context of the existing work on domestication and AMF abundance. Next, we examine domestication and cultivar effects on AMF and SOC accrual. Finally, we discuss the implications of our findings with respect to C sequestration practices in tropical agroforestry.

4.1. To What Extent Is AMF Abundance Mediated by Domestication History (Host Determinism)?

Increasing breadfruit domestication was accompanied by a progressive decline in AMF root colonization by C-rich vesicles and hyphal productivity. Similar results for AMF colonization have been noted in previous studies of plant domestication [20]. AMF root colonization and external growth requires a symbiosis established by molecular dialogue [62] and that plants allocate sufficient photosynthates to sustain the mutualism [63]. It follows that domestication in breadfruit may have reduced C allocation to AMF [20,26,64] or diminished the importance of AMF mutualisms for resource acquisition and transfer [21] as occurs in other domesticated crops [64,65]. Our data support both possibilities. For example, a reduced AMF abundance with domestication occurred in both MCB (low N) and KAH (high N), suggesting an inadvertent selection for cultivars with less genetic capacity to host or respond to AMF [65]. In addition, there was no significant difference in foliar N among cultivars in either MCB or KAH, thereby indicating a reduced AMF dependence or responsiveness for plant N acquisition in highly domesticated cultivars [66].

4.2. To What Extent Is AMF Functioning (Aggregation and Aggregate C), Mediated by Domestication History (Host Determinism)?

While AMF root structures are useful in understanding the effects of domestication on the plant itself, the external hyphae they support are important for the ecosystem functions of soil aggregation and C sequestration [43,51]. The observed decline in breadfruit AMF hyphae with domestication parallels observations that earlier crop varieties support a higher abundance of external AMF hyphae than more recent highly domesticated varieties [64]. However, there was no linear correlation between AMF hyphae (productivity and hyphal pool) and aggregate abundance, and macro- and microaggregate abundances were similar between domestication levels and study sites. In addition, the marginal significance of AMF root colonization on microaggregate abundance points towards only a partial involvement of AMF in soil structure and stabilization, if at all [67,68]. While unexpected, our findings are in line with previous studies in tropical forests [69].
Instead, we found that aggregation was better correlated with the soil mineral fraction, including Fe and silt. In tropical soils, the mineral fraction is large (75%–90% in our study) and can promote mineral–mineral binding and subsequent aggregate formation [59,60]. Additional correlates, such as soil N, and root exudates or rhizodeposits may have stimulated microbial activity and the production of extracellular substrates that stabilize organic residues onto mineral surfaces or bind aggregates [69,70,71]. This is supported by aggregate C:N levels: microbial necromass has a narrow C:N ratio (4–8) [72,73]. Breadfruit macro- and microaggregate C:N ranged from 5 to 9, consistent with inputs from microbial biomass or necromass [45,74] rather than breadfruit litter (C:N 15.9 ± 1.2, mean ± se). These findings reinforce that plant litter was first transformed by microbial assimilation and then stabilized in aggregates [36,75,76].
Domestication also modulated macroaggregate C levels. Our study revealed significantly lower levels of large and small macroaggregate-bound C in domesticated cultivars than in wild relatives, thereby demonstrating that the effects of plant selection for aboveground plant traits can also extend to affect one of the primary contributors to soil C sequestration [70]. Our results also demonstrated the significant and positive effects of AMF productivity and microbial biomass, corroborating the expected contributions of microbes to aggregate C accrual [77], along with soil N. In addition, shifts in large macroaggregate C levels among domestication levels were explained by microbial biomass, a factor that acts as cementing agents to promote aggregation and C stabilization [78,79]. Aboveground, breadfruit cultivars show vast differences in plant structural and functional traits [7]. If similar variations occur below ground, plasticity in the root architecture, morphology, or exudation among breadfruit cultivars may have modulated the C accrual by altering the rhizosphere microbial dynamics, as occurs in other domesticated crops [25,80,81]. For example, cultivars with extensive fine root systems or strong AMF associations may distribute exudates to broadly stimulate the microbial turnover of resources in the rhizosphere [82]. The significant correlation between AMF productivity and labile C levels supports this possibility.
Even so, small macroaggregate C was linked to soil Fe and K levels and declined with increasing silt content. Similar effects have been noted in other tropical systems [59]. Unlike temperate systems, silt and clay content can be poor predictors of the SOC sequestration potential in tropical soils [83]. This discrepancy may be due to soil weathering effects on clay mineralogy and oxyhydroxides (iron, manganese, and aluminum), both of which play key roles in aggregate formation and C stabilization via the development of organo–mineral complexes, i.e., agglomerations of clay, polyvalent metals, and organic matter including microbial necromass [59,68]. Our data shows that soils in all four sites were variously enriched with iron oxides (3%–11% mass) and aluminosilicates (KAH soils [39]). Consequently, there may be a stronger, positive correlation between clay mineralogy, pedogenic oxides, and aggregate-bound C levels in our sites than silt content. In addition, the importance of K may signal the presence of bacteria capable of solubilizing mineral-bound K. However, another parameter, i.e., generalist fungi, also appeared to explain the variations in macroaggregate C and bulk SOC.
The observed increases in generalist fungi with domestication is in line with other reports in other domesticated crops [31,81]. That these fungi showed the highest abundance in the most domesticated cultivars (D3) suggests a greater impact of modern selections on soil fungal communities than the initial domestication events. Their abundance may reflect a responsiveness to functional variations in breadfruit root traits with domestication, including exudation [25,81]. Alternatively, generalist fungi may promote aggregate stabilization via the mechanical entanglement of clay particles while fungal-derived extracellular polymers contribute to aggregate C, as occurs in filamentous soil fungi [84,85]. If so, this process would explain the similar levels of aggregation among domestication levels. However, it does not rule out the possibility that these fungi released hydrolytic and oxidative enzymes to break down aggregate binding agents to access aggregate C [86]. Further, additions of N-rich leaf litter or compost in our study sites may have triggered a fungal demand for C, leading to an increased mineralization of aggregate C (i.e., priming effect [87]).
We also observed a trade-off in abundance between generalist fungi and AMF: as AMF abundance declined with domestication, there was a reciprocal increase in the abundance of generalist fungi. It is tempting to speculate that the increasing abundance of generalist fungi reflects the loss in the bioprotective effects of AMF [88], competition for resources, or changes in the levels of plant-derived secondary metabolites that could variously stimulate or hinder the growth of different fungal groups or taxa [81]. Further, the shifting abundances of generalist fungi, AMF, and, to a lesser extent, microbial biomass among cultivars indicates that domestication has altered the assembly of rhizosphere microbial communities in comparison to their wild counterparts [24,81]. These changes could lead to trade-offs in microbial traits, including the quantity and quality of microbial residues, to influence ecosystem C storage. Clearly, future work should focus on identifying the functional role(s) of these microbial groups in SOC sequestration.

4.3. Can the Combined Effects of Host Domestication and AMF Modulate Bulk SOC (Host Determinism)?

Aligning the cultivars on the west–east domestication gradient illustrated a remarkable shift in bulk SOC levels. Early domestication (D1 and D2) was marked by a progressive reduction in bulk SOC relative to the wild relatives; this outcome parallels the losses in bulk SOC seen in other domesticated crops (e.g., wheat [89]). In contrast, the transition to highly domesticated or hybrid cultivars (D3) was characterized by increasing bulk SOC. These changes occurred in concert with the shift from diploid to triploid status, so that cultivar identity, domestication, and ploidy overlapped along the breadfruit domestication gradient. Disentangling those effects showed that the effects of domestication and cultivar were stronger than ploidy.
Comparisons showed that AMF productivity was a significant correlate of bulk SOC. Here, domestication-driven declines in AMF hyphal productivity were linked to reductions in the bulk SOC level. This significant association supports the sensitivity of AMF to the host genotype [90], and the observed shifts in the AMF community structure towards less mutualistic AMF taxa during breadfruit domestication [21]. These less mutualistic AMF taxa, sometimes referred to as fine endophytes, are characterized by biomass allocation to internal root structures rather than external hyphae [91]. In our study, the negative relationship between fine endophytes and large macroaggregate C is consistent with this possibility. From a functional perspective, AMF hyphae can contribute directly to soil C pools by transferring plant-derived C to microbes in the rhizosphere (e.g., exudates and extracellular enzymes) and via hyphal turnover and necromass accumulation [77,92]. Thus, any domestication-driven shifts in AMF hyphal productivity and AMF-associated biopolymers could be expected to reduce AMF-derived C to organic matter pools and the wider soil microbial community [92,93,94]. Importantly, these findings indicate a possible decoupling between domestication and AMF life history traits that are important for soil C cycling.
Moreover, bulk SOC levels were positively correlated with soil N and microbial biomass. These results are in accordance with earlier works showing linkages between bulk SOC and microbial biomass [95,96] and N [35,36,97], and that rhizosphere bacterial communities are less affected by domestication than fungi [81,90]. In part, our results may reflect increases in microbial C use efficiency and/or microbial biomass or necromass accumulation following inputs of high-quality breadfruit litter (low C and high N [98,99]). Equally, increasing soil N fertility is expected to promote the accumulation of bacterial biomass and necromass [35,97,100], based on tissue stoichiometry (lower C:N in bacteria vs fungi) [45] and the concept of r- and K-selection theory applied to soil microbial communities [101]. Taken together, it is reasonable to suggest that microbial life-history traits and environmental context also contribute to the differences in bulk SOC among cultivars as well as AMF abundance.

4.4. Does Breadfruit Domestication History and Soil N Fertility Converge to Affect AMF Abundance in Relation to AMF Functions in Soil Aggregate C and SOC Accrual (Host Insensitivity)?

While increasing breadfruit domestication levels reduced the abundances of AMF and aggregate and bulk C pools in low-N soils (MCB), such effects were not detected in high-fertility soils (KAH). In fact, AMF abundance and aggregate C pools were largely similar between wild relatives and the domesticated cultivars in KAH. In addition, it appeared that all cultivars may have a similar C sequestration potential in more fertile soils. These findings provide some support for the host insensitivity hypothesis; i.e., more modern cultivars show a reduced capacity to support their recruited AMF. More generally, however, these results are in line with the plasticity (domesticated cultivars [80]) or sensitivity of AMF to high soil fertility (wild relatives [30]). There remains the question of why AMF productivity increased in high-fertility soils.
In soils characterized by higher N availabilities relative to P (KAH), P limitation can trigger increases in plant C allocation to their AMF for external hyphal growth [102] and the subsequent acquisition and transfer of P to the host plant [22,66]. Further, plants in tropical systems experience considerable P limitation because P complexes with Fe and Al (hydroxy) oxides, making P largely unavailable to plants [30,34]. In our study, this is underscored by the significant relationship between total P, Ca, and Fe. Under these conditions, AMF functional traits, including the exudation of enzymes (phosphatases), the acquisition and transfer of P to the host, and the capacity to stimulate the mineralization of organic P by rhizosphere bacteria, can alleviate a plant’s physiological demand for P. Thus, a combination of soil and plant factors may underlie the higher productivity of AMF hyphae in KAH rather than SOC sequestration [15]. This P demand may also be higher in more domesticated cultivars (D2 and D3) owing to their larger stature, bigger leaves, and heavier fruits [7]. Further studies linking domestication-induced changes in AMF traits, including the production of enzymes (phosphatases) and P uptake under natural conditions, are needed to fully evaluate the role(s) of AMF in breadfruit P nutrition or other resources.
None of the AMF properties tested could explain the variations in microaggregate C levels. Despite the small sample sizes in our study, this finding is in general agreement with the results in other tropical forests [103]. In the absence of causal factors, our results can be explained if we consider two non-mutually exclusive mechanisms of microaggregate C stabilization: (i) the transfer of fresh C from residues to mineral surfaces via microbial activity [77] and (ii) the adsorption of microbial residues onto clay-sized mineral particles [104]. The enrichment of microaggregates by microbial residues (C:N 4–9; [45]) both supports this possibility and reinforces the importance of soil microbes to long-term soil C storage, but not AMF.

5. Conclusions, Limitations, and Outlook for Agroforestry

In breadfruit and other domesticated crop species, the footprint of human selection is most readily evidenced through phenotypic changes in tree, leaf, and fruit size [61,105]. Our study captured additional traits that are variously modified by domestication, including the levels of AMF productivity, generalist fungi, macroaggregate C, and bulk SOC. In addition, plant cultivar identity influenced bulk SOC, while soil N fertility, AMF productivity, and microbial biomass were correlates of aggregate and bulk C accrual.
A key question is whether these traits and properties contributed to building substantive C stocks. Calculating soil C stocks from our data shows that the upper 15 cm soil layer in breadfruit orchards stored, on average, 99.5 Mg C ha (range 98–103 Mg C ha−1). These levels compare favorably to the mean stocks of 126 Mg C ha−1 found in the meta-analyses of C stocks in agroforestry systems, including cacao, rubber, and mango [1,2,86], and support a role for breadfruit in the provision of ecosystem services.
There are some caveats. First, we sampled a small number of cultivars relative to the immense variety of known breadfruit cultivars. In addition, the replication in some cultivars was limited due to the availability of material (Puou), and our analyses were completed on a per-tree basis. Even so, there were signals that breadfruit in high-N soils (KAH) created larger soil C pools than those in low-N soils (MCB), and the effects of soil N levels appeared to be more important that the influence of domestication on SOC pools. Second, the analysis of microbial biomass was conducted using dry soils, an approach necessitated by soil import and quarantine requirements. While this approach may have under-estimated the microbial biomass, the experimental evidence suggests that the relative difference in microbial biomass in using dry versus moist soils was likely small [54]. However, future analyses using DNA-based procedures (e.g., qPCR) may provide a greater resolution of bacterial versus fungal contributions to microbial biomass. Finally, we used a single sampling point whereas seasonal or inter-annual samplings in concert with breadfruit phenology may better predict the importance of AMF abundance and microbial traits [102].
The positive effect of soil N fertility on soil C pools point to the importance of soil nutrient management. Acquisitive species that are typically fast-growing, including Artocarpus, generally show a higher realized growth in moist mild climates and fertile soils. This suggests a need to consider the environmental context [40] for growing breadfruit. For example, adjusting the type and rate of mulches, organic additives, and mineral fertilizers to improve soil N, P, and K fertility may be important in site management, especially in soils with a high mineral content (Fe and Al oxides). All sites received mineral fertilizers, although at different rates, and only KAH and MCB were amended with composted breadfruit leaves. The additions of leaf litter or compost may have enriched soil N pools (KAH) or were readily cycled by the microbial biomass (MCB). Even so, a more coordinated approach that tests soil fertility and adjusts fertilizer applications (type and rate) and mulching is needed to better understand how to harness breadfruit for soil C accrual. Our results also show the potential for matching the cultivar to the soil conditions; e.g., highly domesticated cultivars provided the greatest C accrual in high-N soils.
While our results are a first step in understanding the role of breadfruit in soil C accrual, we focused only on AMF and microbial contributions to C sequestration. What is missing is an estimate of C in belowground plant compartments, including fine root biomass and root exudation patterns with depth, and consideration of leaching [1]. For example, soil labile C was influenced by breadfruit domestication and linked to AMF productivity. The extent to which root exudation contributed to the levels of labile C and interacted with microbial communities was beyond the scope of our study. However, such studies could further inform cultural practices with breadfruit that seek to enhance soil C, thereby contributing to both food security and climate change mitigation.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/f17030300/s1, Figure S1. Influence of site and breadfruit domestication level on bulk SOC; Table S1. Statistical comparisons of AMF, microbial, and soil properties in Ma’afala across sites; Table S2. Collection information for individual breadfruit trees; Table S3. Summary of Levene and Shapiro–Wilks’ tests across soil, AMF, and microbial properties; Table S4. Correlations between aggregate and bulk soil OC and soil, plant, and AMF properties in (a) McBryde Garden only, and (b) McBryde and Kahanu Gardens; Table S5. Mean values of soil, AMF, and plant factors and statistical summary (p values) of the effect of cultivar, domestication. and ploidy on soil factors in Kahanu and McBryde Gardens; Table S6. Statistical summary (p values) of the effect of cultivar, domestication level, and ploidy soil properties in Kahanu and McBryde Gardens; Table S7. Summary of standardized LASSO coefficients; Table S8: Summaries of rank regressions for macroaggregate Cand bulk SOC pools. Table S9. Results of rank regression model to identify cultivars that differed significantly in bulk SOC levels in McBryde Garden; Table S10. Effect of geographic origin on leaf, AMF, and soil factors; Table S11. Raw data of parameters across sites and breadfruit cultivars.

Author Contributions

Conceptualization, L.M.E.-W. and L.G.; methodology, L.G., E.P., N.D.D. and L.M.E.-W.; formal analysis, L.M.E.-W., L.G. and R.D.; investigation, L.G., E.P., L.M.E.-W. and N.D.D.; resources, N.J.C.Z., R.D. and L.M.E.-W.; writing—original draft preparation, L.M.E.-W. and L.G.; writing—review and editing, N.J.C.Z., R.D. and N.D.D. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by research and travel grants from The Graduate Program in Plant Biology and Conservation at Northwestern University, and the Research Experiences for Undergraduates (REU) program, Plant Conservation from Genes to Ecosystems, at the Chicago Botanic Garden (NSF DBI-1757800, DBI-2149888).

Data Availability Statement

Summary data are available at https://doi.org/10.5281/zenodo.17859655; full data sets available upon request.

Acknowledgments

We thank the staff at The Breadfruit Institute for the logistic support, Brian Yamamoto (Kaua’i Community College) and Nicholas Reppun (Kāko‘o ‘Ōiwi) for their assistance in soil and leaf sampling, and Elijah McDavid and Steven Finkelman for laboratory assistance. We extend our sincere thanks to the three anonymous reviewers for their insights and comments that significantly improved the manuscript. Soil samples were shipped from Hawai’i to the Chicago Botanic Garden, Illinois, in accordance with the APHIS Permit to Receive Soil (P330-18-00040).

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Albrecht, A.; Kandji, S.T. Carbon sequestration in tropical agroforestry systems. Agric. Ecosyst. Environ. 2003, 99, 15–27. [Google Scholar] [CrossRef]
  2. Ma, Z.; Chen, H.Y.H.; Bork, E.W.; Carlyle, C.N.; Chang, S.X.; Fortin, J. Carbon accumulation in agroforestry systems is affected by tree species diversity, age and regional climate: A global meta-analysis. Glob. Ecol. Biogeogr. 2020, 29, 1817–1828. [Google Scholar] [CrossRef]
  3. Haider, S.M.A.; Ahmad, I.; Majeed, K.; Hussain, M. Soil organic carbon sequestration potential and its sustainability comparison between mango-based agroforestry and cropland seeking soil fertility parameters under climate resilience. Water Air Soil Pollut. 2024, 235, 682. [Google Scholar] [CrossRef]
  4. Noponen, M.R.A.; Healey, J.R.; Soto, G.; Haggar, J.P. Sink or source—The potential of coffee agroforestry systems to sequester atmospheric CO2 into soil organic carbon. Agric. Ecosyst. Environ. 2013, 175, 60–68. [Google Scholar] [CrossRef]
  5. Murphy, T.; Jones, G.; Vanclay, J.; Glencross, K. Preliminary carbon sequestration modelling for the Australian macadamia industry. Agrofor. Syst. 2012, 87, 689–698. [Google Scholar] [CrossRef]
  6. Monroe, P.H.M.; Gama-Rodrigues, E.F.; Gama-Rodrigues, A.C.; Marques, J.R.B. Soil carbon stocks and origin under different cacao agroforestry systems in Southern Bahia, Brazil. Agric. Ecosyst. Environ. 2016, 221, 99–108. [Google Scholar] [CrossRef]
  7. Lincoln, N.K.; Cho, A.; Dow, G.; Radovich, T. Early growth of breadfruit in a variety × environment trial. Agron. J. 2019, 111, 3020–3027. [Google Scholar] [CrossRef]
  8. Dow, G.J.; Lincoln, N.K.; Autufuga, D.; Paull, R. From forests to farming: Identification of photosynthetic limitations in breadfruit across diverse environments. Tree Physiol. 2024, 45, tpaf007. [Google Scholar] [CrossRef]
  9. Zerega, N.J.; Ragone, D.; Motley, T.J. Complex origins of breadfruit (Artocarpus altilis, Moraceae): Implications for human migrations in Oceania. Am. J. Bot. 2004, 91, 760–766. [Google Scholar] [CrossRef]
  10. Mausio, K.; Miura, T.; Lincoln, N.K. Cultivation potential projections of breadfruit (Artocarpus altilis) under climate change scenarios using an empirically validated suitability model calibrated in Hawai’i. PLoS ONE 2020, 15, e0228552. [Google Scholar] [CrossRef]
  11. Livingston, C.; Lincoln, N.K. Determining allometry and carbon sequestration potential of breadfruit (Artocarpus altilis) as a climate-smart staple in Hawai‘i. Sustainability 2023, 15, 15682. [Google Scholar] [CrossRef]
  12. Langston, B.; Lincoln, N. The role of breadfruit in biocultural restoration and sustainability in Hawai‘i. Sustainability 2018, 10, 3965. [Google Scholar] [CrossRef]
  13. Lorenz, K.; Lal, R. Soil organic carbon sequestration in agroforestry systems. A review. Agron. Sustain. Dev. 2014, 34, 443–454. [Google Scholar] [CrossRef]
  14. Cardinael, R.; Guenet, B.; Chevallier, T.; Dupraz, C.; Cozzi, T.; Chenu, C. High organic inputs explain shallow and deep SOC storage in a long-term agroforestry system–combining experimental and modeling approaches. Biogeosciences 2018, 15, 297–317. [Google Scholar] [CrossRef]
  15. Basiru, S.; Hijri, M. Trade-off between soil organic carbon sequestration and plant nutrient uptake in arbuscular mycorrhizal symbiosis. Fungal Biol. Rev. 2024, 49, 100381. [Google Scholar] [CrossRef]
  16. Leifheit, E.F.; Veresoglou, S.D.; Lehmann, A.; Morris, E.K.; Rillig, M.C. Multiple factors influence the role of arbuscular mycorrhizal fungi in soil aggregation—A meta-analysis. Plant Soil 2013, 374, 523–537. [Google Scholar] [CrossRef]
  17. Agnihotri, R.; Sharma, M.P.; Prakash, A.; Ramesh, A.; Bhattacharjya, S.; Patra, A.K.; Manna, M.C.; Kurganova, I.; Kuzyakov, Y. Glycoproteins of arbuscular mycorrhiza for soil carbon sequestration: Review of mechanisms and controls. Sci. Total Environ. 2022, 806, 150571. [Google Scholar] [CrossRef]
  18. Basile-Doelsch, I.; Balesdent, J.; Pellerin, S. Reviews and syntheses: The mechanisms underlying carbon storage in soil. Biogeosciences 2020, 17, 5223–5242. [Google Scholar] [CrossRef]
  19. Horsch, C.C.A.; Antunes, P.M.; Fahey, C.; Grandy, A.S.; Kallenbach, C.M. Trait-based assembly of arbuscular mycorrhizal fungal communities determines soil carbon formation and retention. New Phytol. 2023, 239, 311–324. [Google Scholar] [CrossRef]
  20. Martin-Robles, N.; Lehmann, A.; Seco, E.; Aroca, R.; Rillig, M.C.; Milla, R. Impacts of domestication on the arbuscular mycorrhizal symbiosis of 27 crop species. New Phytol. 2018, 218, 322–334. [Google Scholar] [CrossRef]
  21. Xing, X.; Koch, A.M.; Jones, A.M.; Ragone, D.; Murch, S.; Hart, M.M. Mutualism breakdown in breadfruit domestication. Proc. R. Soc. Ser. B Biol. Sci. 2012, 279, 1122–1130. [Google Scholar] [CrossRef]
  22. Giovannini, L.; Sbrana, C.; Giovannetti, M.; Avio, L.; Lanubile, A.; Marocco, A.; Turrini, A. Diverse mycorrhizal maize inbred lines differentially modulate mycelial traits and the expression of plant and fungal phosphate transporters. Sci. Rep. 2022, 12, 21279. [Google Scholar] [CrossRef]
  23. Dai, M.; Hamel, C.; Bainard, L.D.; Arnaud, M.S.; Grant, C.A.; Lupwayi, N.Z.; Malhi, S.S.; Lemke, R. Negative and positive contributions of arbuscular mycorrhizal fungal taxa to wheat production and nutrient uptake efficiency in organic and conventional systems in the Canadian prairie. Soil Biol. Biochem. 2014, 74, 156–166. [Google Scholar] [CrossRef]
  24. He, X.; Hochholdinger, F.; Chen, X.; Peng, Y. Crop domestication and improvement reshape root traits and the structure and function of their associated microbiome. Front. Agric. Sci. Eng. 2024, 12, 27–36. [Google Scholar] [CrossRef]
  25. Rinehart, B.; Borras, L.; Salmeron, M.; McNear, D.H.; Poffenbarger, H. Commercial maize hybrids have smaller root systems after 80 Years of breeding. Rhizosphere 2024, 30, 100915. [Google Scholar] [CrossRef]
  26. Isaac, M.E.; Nimmo, V.; Gaudin, A.C.M.; Leptin, A.; Schmidt, J.E.; Kallenbach, C.M.; Martin, A.; Entz, M.; Carkner, M.; Rajcan, I.; et al. Crop domestication, root trait syndromes, and soil nutrient acquisition in organic agroecosystems: A systematic review. Front. Sustain. Food Syst. 2021, 5, 716480. [Google Scholar] [CrossRef]
  27. Porter, S.S.; Sachs, J.L. Agriculture and the Disruption of Plant-Microbial Symbiosis. Trends Ecol. Evol. 2020, 35, 426–439. [Google Scholar] [CrossRef]
  28. Garo, G.; Van Geel, M.; Eshetu, F.; Swennen, R.; Honnay, O.; Vancampenhout, K. Arbuscular mycorrhizal fungus communities and their response to soil phosphorous differ between wild and domesticated enset (Ensete ventricosum) in Southern Ethiopia. Rhizosphere 2021, 20, 100444. [Google Scholar] [CrossRef]
  29. Sangabriel-Conde, W.; Maldonado-Mendoza, I.E.; Mancera-López, M.E.; Cordero-Ramírez, J.D.; Trejo-Aguilar, D.; Negrete-Yankelevich, S. Glomeromycota associated with Mexican native maize landraces in Los Tuxtlas, Mexico. Appl. Soil Ecol. 2015, 87, 63–71. [Google Scholar] [CrossRef]
  30. Ma, X.; Geng, Q.; Zhang, H.; Bian, C.; Chen, H.Y.H.; Jiang, D.; Xu, X. Global negative effects of nutrient enrichment on arbuscular mycorrhizal fungi, plant diversity and ecosystem multifunctionality. New Phytol. 2021, 229, 2957–2969. [Google Scholar] [CrossRef]
  31. Abdelhalim, T.; Jannoura, R.; Joergensen, R.G. Arbuscular mycorrhizal dependency and phosphorus responsiveness of released, landrace and wild Sudanese sorghum genotypes. Arch. Agron. Soil Sci. 2019, 66, 706–716. [Google Scholar] [CrossRef]
  32. Elliott, A.J.; Daniell, T.J.; Cameron, D.D.; Field, K.J. A commercial arbuscular mycorrhizal inoculum increases root colonization across wheat cultivars but does not increase assimilation of mycorrhiza-acquired nutrients. Plants People Planet 2021, 3, 588–599. [Google Scholar] [CrossRef]
  33. Schmidt, J.E.; Bowles, T.M.; Gaudin, A.C. Using ancient traits to convert soil health into crop yield: Impact of selection on maize root and rhizosphere function. Front. Plant Sci. 2016, 7, 373. [Google Scholar] [CrossRef] [PubMed]
  34. Cesco, S.; Alzate Zuluaga, M.Y.; Cavani, L.; Borruso, L.; Laudicina, V.A.; Mazzetto, F.; Mimmo, T.; Pii, Y.; Terzano, R.; Astolfi, S. Roots and the rhizosphere: A perspective on the hidden engine of regenerative, antifragile, and digitally enabled agriculture. Farming Syst. 2026, 4, 100199. [Google Scholar] [CrossRef]
  35. Zhang, X.; Yang, M.; Liu, Y.; Ye, J.; Tangyu, J.; Gao, J.; Liu, W.; Fan, Y. Soil Microbes Mediate Productivity Differences Between Natural and Plantation Forests. Plants 2025, 15, 98. [Google Scholar] [CrossRef] [PubMed]
  36. Tang, B.; Rocci, K.S.; Lehmann, A.; Rillig, M.C. Nitrogen increases soil organic carbon accrual and alters its functionality. Glob. Change Biol. 2023, 29, 1971–1983. [Google Scholar] [CrossRef]
  37. Ma, Y.; Zhang, H.; Wang, D.; Guo, X.; Yang, T.; Xiang, X.; Walder, F.; Chu, H. Differential Responses of Arbuscular Mycorrhizal Fungal Communities to Long-Term Fertilization in the Wheat Rhizosphere and Root Endosphere. Appl. Environ. Microbiol. 2021, 87, e0034921. [Google Scholar] [CrossRef]
  38. Giambelluca, T.W.; Chen, Q.; Frazier, A.G.; Price, J.P.; Chen, Y.-L.; Chu, P.-S.; Eischeid, J.K.; Delparte, D.M. Online Rainfall Atlas of Hawai‘i. Bull. Am. Meteorol. Soc. 2013, 94, 313–316. [Google Scholar] [CrossRef]
  39. Deenik, J.; McClellan, A. Soils of Hawaii. Univ. Hawaii Soil Crop Manag. 2007, 12. Available online: http://hdl.handle.net/10125/12463 (accessed on 23 February 2026).
  40. Augusto, L.; Borelle, R.; Boča, A.; Bon, L.; Orazio, C.; Arias-González, A.; Bakker, M.R.; Gartzia-Bengoetxea, N.; Auge, H.; Bernier, F.; et al. Widespread slow growth of acquisitive tree species. Nature 2025, 640, 395–401. [Google Scholar] [CrossRef]
  41. Jones, A.M.P.; Ragone, D.; Aiona, K.; Lane, W.A.; Murch, S.J. Nutritional and morphological diversity of breadfruit (Artocarpus, Moraceae): Identification of elite cultivars for food security. J. Food Compos. Anal. 2011, 24, 1091–1102. [Google Scholar] [CrossRef]
  42. Miller, W.P.; Miller, D.M. A micro-pipette method for soil mechanical analysis. Commun. Soil Sci. Plant Anal. 1987, 18, 1–15. [Google Scholar] [CrossRef]
  43. Tisdall, J.M.; Oades, J.M. Organic matter and water-stable aggregates in soils. J. Soil Sci. 1982, 33, 141–163. [Google Scholar] [CrossRef]
  44. Kemper, W.D.; Rosenau, R.C. Aggregate stability and size distribution. Methods Soil Anal. 1986, 5, 425–442. [Google Scholar]
  45. Cleveland, C.C.; Liptzin, D. C:N:P stoichiometry in soil: Is there a “Redfield ratio” for the microbial biomass? Biogeochemistry 2007, 85, 235–252. [Google Scholar] [CrossRef]
  46. Doane, T.A.; Horwáth, W.R. Spectrophotometric determination of nitrate with a single reagent. Anal. Lett. 2003, 36, 2713–2722. [Google Scholar] [CrossRef]
  47. Weatherburn, M.W. Phenol-hypochlorite reaction for determination of ammonia. Anal. Chem. 1967, 39, 971–974. [Google Scholar] [CrossRef]
  48. Baykov, A.A.; Evtushenko, O.A.; Avaeva, S.M. A malachite green procedure for orthophosphate determination and its use in alkaline phosphatase-based enzyme immunoassay. Anal. Biochem. 1988, 171, 266–270. [Google Scholar] [CrossRef]
  49. Koske, R.; Gemma, J. A modified procedure for staining roots to detect VA mycorrhizas. Mycol. Res. 1989, 92, 486–488. [Google Scholar] [CrossRef]
  50. Giovannetti, M.; Mosse, B. An evaluation of techniques for measuring vesicular arbuscular mycorrhizal infection in roots. New Phytol. 1980, 84, 489–500. [Google Scholar] [CrossRef]
  51. Jakobsen, I.; Abbott, L.; Robson, A. External hyphae of vesicular-arbuscular mycorrhizal fungi associated with Trifolium subterraneum L. 1. Spread of hyphae and phosphorus inflow into roots. New Phytol. 1992, 120, 371–380. [Google Scholar] [CrossRef]
  52. Newman, E. A method of estimating the total length of root in a sample. J. Appl. Ecol. 1966, 3, 139–145. [Google Scholar] [CrossRef]
  53. Vance, E.D.; Brookes, P.C.; Jenkinson, D.S. An extraction method for measuring soil microbial biomass C. Soil Biol. Biochem. 1987, 19, 703–707. [Google Scholar] [CrossRef]
  54. Schroeder, J.; Kammann, L.; Helfrich, M.; Tebbe, C.C.; Poeplau, C. Impact of common sample pre-treatments on key soil microbial properties. Soil Biol. Biochem. 2021, 160, 108321. [Google Scholar] [CrossRef]
  55. DuBois, M.; Gilles, K.A.; Hamilton, J.K.; Rebers, P.A.; Smith, F. Colorimetric method for determination of sugars and related substances. Anal. Chem. 1956, 28, 350–356. [Google Scholar] [CrossRef]
  56. Neff, J.C.; Asner, G.P. Dissolved organic carbon in terrestrial ecosystems: Synthesis and a model. Ecosystems 2001, 4, 29–48. [Google Scholar] [CrossRef]
  57. Horwath, W.R.; Paul, E.A.; Harris, D.; Norton, J.; Jagger, L.; Horton, K.A. Defining a realistic control for the chloroform fumigation- incubation method using microscopic counting and 14C-substrates. Can. J. Soil Sci. 1996, 76, 459–467. [Google Scholar] [CrossRef]
  58. Oksanen, J.; Simpson, G.; Blanchet, F.; Kindt, R.; Legendre, P.; Minchin, P.; O’Hara, R.; Solymos, P.; Stevens, M.; Szoecs, E.; et al. Vegan: Community Ecology Package. R Package Version 2.8-0. Available online: https://vegandevs.github.io/vegan/ (accessed on 21 February 2026).
  59. Kirsten, M.; Mikutta, R.; Vogel, C.; Thompson, A.; Mueller, C.W.; Kimaro, D.N.; Bergsma, H.L.T.; Feger, K.-H.; Kalbitz, K. Iron oxides and aluminous clays selectively control soil carbon storage and stability in the humid tropics. Sci. Rep. 2021, 11, 5076. [Google Scholar] [CrossRef]
  60. Six, J.; Feller, C.; Denef, K.; Ogle, S.; de Moraes Sa, J.C.; Albrecht, A. Soil organic matter, biota and aggregation in temperate and tropical soils-Effects of no-tillage. Agronomie 2002, 22, 755–775. [Google Scholar] [CrossRef]
  61. Alam, O.; Purugganan, M.D. Domestication and the evolution of crops: Variable syndromes, complex genetic architectures, and ecological entanglements. Plant Cell 2024, 36, 1227–1241. [Google Scholar] [CrossRef]
  62. Guigard, L.; Jobert, L.; Busset, N.; Moulin, L.; Czernic, P. Symbiotic compatibility between rice cultivars and arbuscular mycorrhizal fungi genotypes affects rice growth and mycorrhiza-induced resistance. Front. Plant Sci. 2023, 14, 1278990. [Google Scholar] [CrossRef] [PubMed]
  63. Sugiura, Y.; Akiyama, R.; Tanaka, S.; Yano, K.; Kameoka, H.; Marui, S.; Saito, M.; Kawaguchi, M.; Akiyama, K.; Saito, K. Myristate can be used as a carbon and energy source for the asymbiotic growth of arbuscular mycorrhizal fungi. Proc. Natl. Acad. Sci. USA 2020, 117, 25779–25788. [Google Scholar] [CrossRef] [PubMed]
  64. Sawers, R.J.; Ramírez-Flores, M.R.; Olalde-Portugal, V.; Paszkowski, U. The impact of domestication and crop improvement on arbuscular mycorrhizal symbiosis in cereals: Insights from genetics and genomics. New Phytol. 2018, 220, 1135–1140. [Google Scholar] [CrossRef] [PubMed]
  65. Essiane-Ondo, O.; Zerbib, J.; Gianinazzi, S.; Wipf, D. Wheat landraces with low mycorrhizing ability at field respond differently to inoculation with artificial or indigenous arbuscular mycorrhizal fungal communities. Symbiosis 2019, 78, 229–240. [Google Scholar] [CrossRef]
  66. Sawers, R.J.H.; Svane, S.F.; Quan, C.; Grønlund, M.; Wozniak, B.; Gebreselassie, M.-N.; González-Muñoz, E.; Chávez Montes, R.A.; Baxter, I.; Goudet, J.; et al. Phosphorus acquisition efficiency in arbuscular mycorrhizal maize is correlated with the abundance of root-external hyphae and the accumulation of transcripts encoding PHT1 phosphate transporters. New Phytol. 2017, 214, 632–643. [Google Scholar] [CrossRef]
  67. Wu, Q.-S.; Cao, M.-Q.; Zou, Y.-N.; He, X.-h. Direct and indirect effects of glomalin, mycorrhizal hyphae and roots on aggregate stability in rhizosphere of trifoliate orange. Sci. Rep. 2014, 4, 5823. [Google Scholar] [CrossRef]
  68. Totsche, K.U.; Amelung, W.; Gerzabek, M.H.; Guggenberger, G.; Klumpp, E.; Knief, C.; Lehndorff, E.; Mikutta, R.; Peth, S.; Prechtel, A. Microaggregates in soils. J. Plant Nutr. Soil Sci. 2018, 181, 104–136. [Google Scholar] [CrossRef]
  69. Tripathi, S.K.; Kushwaha, C.P.; Singh, K.P. Tropical forest and savanna ecosystems show differential impact of N and P additions on soil organic matter and aggregate structure. Glob. Change Biol. 2008, 14, 2572–2581. [Google Scholar] [CrossRef]
  70. Six, J.; Elliott, E.T.; Paustian, K. Soil macroaggregate turnover and microaggregate formation: A mechanism for C sequestration under no-tillage agriculture. Soil Biol. Biochem. 2000, 32, 2099–2103. [Google Scholar] [CrossRef]
  71. Kopittke, P.M.; Hernandez-Soriano, M.C.; Dalal, R.C.; Finn, D.; Menzies, N.W.; Hoeschen, C.; Mueller, C.W. Nitrogen-rich microbial products provide new organo-mineral associations for the stabilization of soil organic matter. Glob. Change Biol. 2018, 24, 1762–1770. [Google Scholar] [CrossRef]
  72. Kögel-Knabner, I. The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter: Fourteen years on. Soil Biol. Biochem. 2017, 105, A3–A8. [Google Scholar] [CrossRef]
  73. Mooshammer, M.; Wanek, W.; Hämmerle, I.; Fuchslueger, L.; Hofhansl, F.; Knoltsch, A.; Schnecker, J.; Takriti, M.; Watzka, M.; Wild, B.; et al. Adjustment of microbial nitrogen use efficiency to carbon:nitrogen imbalances regulates soil nitrogen cycling. Nat. Commun. 2014, 5, 3694. [Google Scholar] [CrossRef] [PubMed]
  74. Rosinger, C.; Keiblinger, K.M.; Rousk, J.; Sandén, H. Shifts in microbial stoichiometry upon nutrient addition do not capture growth-limiting nutrients for soil microorganisms in two subtropical soils. Biogeochemistry 2022, 159, 33–43. [Google Scholar] [CrossRef]
  75. Zhao, R.; Kuzyakov, Y.; Zhang, H.; Wang, Z.; Li, T.; Shao, L.; Jiang, L.; Wang, R.; Li, M.; Jianxin Sun, O.; et al. Labile carbon inputs offset nitrogen-induced soil aggregate destabilization via enhanced growth of saprophytic fungi in a meadow steppe. Geoderma 2024, 443, 116841. [Google Scholar] [CrossRef]
  76. Khan, K.S.; Mack, R.; Castillo, X.; Kaiser, M.; Joergensen, R.G. Microbial biomass, fungal and bacterial residues, and their relationships to the soil organic matter C/N/P/S ratios. Geoderma 2016, 271, 115–123. [Google Scholar] [CrossRef]
  77. Six, J.; Frey, S.D.; Thiet, R.K.; Batten, K. Bacterial and fungal contributions to carbon sequestration in agroecosystems. Soil Sci. Soc. Am. J. 2006, 70, 555–569. [Google Scholar] [CrossRef]
  78. Denef, K.; Six, J. Clay mineralogy determines the importance of biological versus abiotic processes for macroaggregate formation and stabilization. Eur. J. Soil Sci. 2005, 56, 469–479. [Google Scholar] [CrossRef]
  79. Matus, F.J. Fine silt and clay content is the main factor defining maximal C and N accumulations in soils: A meta-analysis. Sci. Rep. 2021, 11, 6438. [Google Scholar] [CrossRef]
  80. Perez-Jaramillo, J.E.; Mendes, R.; Raaijmakers, J.M. Impact of plant domestication on rhizosphere microbiome assembly and functions. Plant Mol. Biol. 2016, 90, 635–644. [Google Scholar] [CrossRef]
  81. Yue, H.; Yue, W.; Jiao, S.; Kim, H.; Lee, Y.H.; Wei, G.; Song, W.; Shu, D. Plant domestication shapes rhizosphere microbiome assembly and metabolic functions. Microbiome 2023, 11, 70. [Google Scholar] [CrossRef]
  82. Junaidi, J.; Kallenbach, C.M.; Byrne, P.F.; Fonte, S.J. Root traits and root biomass allocation impact how wheat genotypes respond to organic amendments and earthworms. PLoS ONE 2018, 13, e0200646. [Google Scholar] [CrossRef] [PubMed]
  83. Bruun, T.B.; Elberling, B.; Christensen, B.T. Lability of soil organic carbon in tropical soils with different clay minerals. Soil Biol. Biochem. 2010, 42, 888–895. [Google Scholar] [CrossRef]
  84. Lehmann, A.; Zheng, W.; Ryo, M.; Soutschek, K.; Roy, J.; Rongstock, R.; Maaß, S.; Rillig, M.C. Fungal traits important for soil aggregation. Front. Microbiol. 2020, 10, 2904. [Google Scholar] [CrossRef] [PubMed]
  85. de Goede, S.P.; Hannula, S.E.; Jansen, B.; Morriën, E. Fungal-mediated soil aggregation as a mechanism for carbon stabilization. ISME J. 2025, 19, wraf074. [Google Scholar] [CrossRef]
  86. Shi, L.; Feng, W.; Xu, J.; Kuzyakov, Y. Agroforestry systems: Meta-analysis of soil carbon stocks, sequestration processes, and future potentials. Land Degrad. Dev. 2018, 29, 3886–3897. [Google Scholar] [CrossRef]
  87. Fontaine, S.; Hénault, C.; Aamor, A.; Bdioui, N.; Bloor, J.; Maire, V.; Mary, B.; Revaillot, S.; Maron, P.-A. Fungi mediate long term sequestration of carbon and nitrogen in soil through their priming effect. Soil Biol. Biochem. 2011, 43, 86–96. [Google Scholar] [CrossRef]
  88. Veresoglou, S.D.; Rillig, M.C. Suppression of fungal and nematode plant pathogens through arbuscular mycorrhizal fungi. Biol. Lett. 2012, 8, 214–217. [Google Scholar] [CrossRef]
  89. Van de Broek, M.; Ghiasi, S.; Decock, C.; Hund, A.; Abiven, S.; Friedli, C.; Werner, R.A.; Six, J. The soil organic carbon stabilization potential of old and new wheat cultivars: A 13CO2-labeling study. Biogeosciences 2020, 17, 2971–2986. [Google Scholar] [CrossRef]
  90. Leff, J.W.; Lynch, R.C.; Kane, N.C.; Fierer, N. Plant domestication and the assembly of bacterial and fungal communities associated with strains of the common sunflower, Helianthus annuus. New Phytol. 2017, 214, 412–423. [Google Scholar] [CrossRef]
  91. Hart, M.M.; Reader, R.J. Taxonomic basis for variation in the colonization strategy of arbuscular mycorrhizal fungi. New Phytol. 2002, 153, 335–344. [Google Scholar] [CrossRef]
  92. Hawkins, H.-J.; Cargill, R.I.M.; Van Nuland, M.E.; Hagen, S.C.; Field, K.J.; Sheldrake, M.; Soudzilovskaia, N.A.; Kiers, E.T. Mycorrhizal mycelium as a global carbon pool. Curr. Biol. 2023, 33, R560–R573. [Google Scholar] [CrossRef] [PubMed]
  93. Buckeridge, K.M.; Creamer, C.; Whitaker, J. Deconstructing the microbial necromass continuum to inform soil carbon sequestration. Funct. Ecol. 2022, 36, 1396–1410. [Google Scholar] [CrossRef]
  94. Liang, C.; Amelung, W.; Lehmann, J.; Kästner, M. Quantitative assessment of microbial necromass contribution to soil organic matter. Glob. Change Biol. 2019, 25, 3578–3590. [Google Scholar] [CrossRef] [PubMed]
  95. Wang, C.; Qu, L.; Yang, L.; Liu, D.; Morrissey, E.; Miao, R.; Liu, Z.; Wang, Q.; Fang, Y.; Bai, E. Large-scale importance of microbial carbon use efficiency and necromass to soil organic carbon. Glob. Change Biol. 2021, 27, 2039–2048. [Google Scholar] [CrossRef]
  96. Yan, Z.; Zhou, J.; Liu, C.; Jia, R.; Mganga, K.Z.; Yang, L.; Yang, Y.; Peixoto, L.; Zang, H.; Zeng, Z. Legume-based crop diversification reinforces soil health and carbon storage driven by microbial biomass and aggregates. Soil Tillage Res. 2023, 234, 105848. [Google Scholar] [CrossRef]
  97. Hu, J.; Huang, C.; Zhou, S.; Liu, X.; Dijkstra, F.A. Nitrogen addition increases microbial necromass in croplands and bacterial necromass in forests: A global meta-analysis. Soil Biol. Biochem. 2022, 165, 108500. [Google Scholar] [CrossRef]
  98. Manzoni, S.; Taylor, P.; Richter, A.; Porporato, A.; Ågren, G.I. Environmental and stoichiometric controls on microbial carbon-use efficiency in soils. New Phytol. 2012, 196, 79–91. [Google Scholar] [CrossRef]
  99. Yang, Y.; Gunina, A.; Cheng, H.; Liu, L.; Wang, B.; Dou, Y.; Wang, Y.; Liang, C.; An, S.; Chang, S.X. Unlocking mechanisms for soil organic matter accumulation: Carbon use efficiency and microbial necromass as the keys. Glob. Change Biol. 2025, 31, e70033. [Google Scholar] [CrossRef]
  100. Zhou, R.; Liu, Y.; Dungait, J.A.; Kumar, A.; Wang, J.; Tiemann, L.K.; Zhang, F.; Kuzyakov, Y.; Tian, J. Microbial necromass in cropland soils: A global meta-analysis of management effects. Glob. Change Biol. 2023, 29, 1998–2014. [Google Scholar] [CrossRef]
  101. Yang, L.; Canarini, A.; Zhang, W.; Lang, M.; Chen, Y.; Cui, Z.; Kuzyakov, Y.; Richter, A.; Chen, X.; Zhang, F.; et al. Microbial life-history strategies mediate microbial carbon pump efficacy in response to N management depending on stoichiometry of microbial demand. Glob. Change Biol. 2024, 30, e17311. [Google Scholar] [CrossRef]
  102. Egerton-Warburton, L.M.; Johnson, N.C.; Allen, E.B. Mycorrhizal community dynamics following nitrogen fertilization: A cross-site test in five grasslands. Ecol. Monogr. 2007, 77, 527–544. [Google Scholar] [CrossRef]
  103. Reichenbach, M.; Fiener, P.; Garland, G.; Griepentrog, M.; Six, J.; Doetterl, S. The role of geochemistry in organic carbon stabilization against microbial decomposition in tropical rainforest soils. Soil 2021, 7, 453–475. [Google Scholar] [CrossRef]
  104. Lehmann, J.; Kinyangi, J.; Solomon, D. Organic matter stabilization in soil microaggregates: Implications from spatial heterogeneity of organic carbon contents and carbon forms. Biogeochemistry 2007, 85, 45–57. [Google Scholar] [CrossRef]
  105. Milla, R. Phenotypic evolution of agricultural crops. Funct. Ecol. 2023, 37, 976–988. [Google Scholar] [CrossRef]
Forests 17 00300 g001
Figure 1. (A) Map of Oceania indicating the major cultural regions (Melanesia, Micronesia, and Polynesia) and south-east Asian origin of Artocarpus wild relatives (A. camansi, and A. mariannensis). Yellow circles indicate the geographic location of each species used in the study, with small white circles denoting the location of cultivars and their domestication level (0, 1, 2, and 3; see Table 1). Solid line indicates the west-to-east domestication pathway; dotted lines denote the geographic pathways of hybridization among Artocarpus species. (B) Map of Hawai’i Islands showing the location of each study site (•). Breadfruit trees growing in (C) McBryde Garden; (D) Kahanu Garden; (E) ‘Ulutopia; and (F) Kāko‘o‘Ōiwi. Image credits: (A) adapted from [9]; (B) U.S. Geological Service; (C) National Tropical Botanic Garden; (D) N. Dickinson; and (E,F) L. Gohd.
Figure 1. (A) Map of Oceania indicating the major cultural regions (Melanesia, Micronesia, and Polynesia) and south-east Asian origin of Artocarpus wild relatives (A. camansi, and A. mariannensis). Yellow circles indicate the geographic location of each species used in the study, with small white circles denoting the location of cultivars and their domestication level (0, 1, 2, and 3; see Table 1). Solid line indicates the west-to-east domestication pathway; dotted lines denote the geographic pathways of hybridization among Artocarpus species. (B) Map of Hawai’i Islands showing the location of each study site (•). Breadfruit trees growing in (C) McBryde Garden; (D) Kahanu Garden; (E) ‘Ulutopia; and (F) Kāko‘o‘Ōiwi. Image credits: (A) adapted from [9]; (B) U.S. Geological Service; (C) National Tropical Botanic Garden; (D) N. Dickinson; and (E,F) L. Gohd.
Forests 17 00300 g001
Forests 17 00300 g002
Figure 2. Effects of breadfruit domestication level and planting site on (A) AMF root colonization by vesicles; (B) AMF hyphal pool, (C) AMF productivity; and (D) abundance of general fungi (non-AMF). Vertical bars indicate the standard error of the mean. For each site and AMF property, domestication levels with the same letter do not differ significant at p < 0.05 (Dunn test); ns, not significant (p > 0.05). MCB, McBryde Garden; KAH, Kahanu Garden.
Figure 2. Effects of breadfruit domestication level and planting site on (A) AMF root colonization by vesicles; (B) AMF hyphal pool, (C) AMF productivity; and (D) abundance of general fungi (non-AMF). Vertical bars indicate the standard error of the mean. For each site and AMF property, domestication levels with the same letter do not differ significant at p < 0.05 (Dunn test); ns, not significant (p > 0.05). MCB, McBryde Garden; KAH, Kahanu Garden.
Forests 17 00300 g002
Forests 17 00300 g003
Figure 3. Effects of breadfruit domestication level and planting site on (A) large macroaggregate; (B) small macroaggregate C levels; and (C) labile C levels. Vertical bars indicate the standard error of the mean. For each soil property and site, domestication levels with the same letter do not differ significant at p < 0.05 (LSD test); ns, not significant (p > 0.05). MCB, McBryde Garden; KAH, Kahanu Garden.
Figure 3. Effects of breadfruit domestication level and planting site on (A) large macroaggregate; (B) small macroaggregate C levels; and (C) labile C levels. Vertical bars indicate the standard error of the mean. For each soil property and site, domestication levels with the same letter do not differ significant at p < 0.05 (LSD test); ns, not significant (p > 0.05). MCB, McBryde Garden; KAH, Kahanu Garden.
Forests 17 00300 g003
Forests 17 00300 g004
Figure 4. Effects of breadfruit cultivar on bulk SOC levels in McBryde Garden. Cultivars ranked, left to right, from wild relatives to highly domesticated and hybrid (D3h) cultivars. Cultivar abbreviations: Dug, Dugdug; Kap, Kapiak; Ulf, Ulu fiti; Maa, Ma’afala; Puo, Puou; Ulu, ‘Ulu; Ote, Otea; Mei, Meitehid; Mid, Midolab. Vertical bars indicate the standard error of the mean. Mean bulk SOC levels with the same letter do not differ significant at p < 0.05 (LSD test). MCB, McBryde Garden; KAH, Kahanu Garden. Asterisks denote cultivars identified as significantly different in bulk SOC by rank regressions across cultivars.
Figure 4. Effects of breadfruit cultivar on bulk SOC levels in McBryde Garden. Cultivars ranked, left to right, from wild relatives to highly domesticated and hybrid (D3h) cultivars. Cultivar abbreviations: Dug, Dugdug; Kap, Kapiak; Ulf, Ulu fiti; Maa, Ma’afala; Puo, Puou; Ulu, ‘Ulu; Ote, Otea; Mei, Meitehid; Mid, Midolab. Vertical bars indicate the standard error of the mean. Mean bulk SOC levels with the same letter do not differ significant at p < 0.05 (LSD test). MCB, McBryde Garden; KAH, Kahanu Garden. Asterisks denote cultivars identified as significantly different in bulk SOC by rank regressions across cultivars.
Forests 17 00300 g004
Table 1. Geographic location, mean annual rainfall, and soil properties in each study site. Data presented as mean with standard error in parentheses. Significant p values highlighted in bold. Within rows, values with the same letter do not differ significantly at p < 0.05 (Fisher LSD test).
Table 1. Geographic location, mean annual rainfall, and soil properties in each study site. Data presented as mean with standard error in parentheses. Significant p values highlighted in bold. Within rows, values with the same letter do not differ significantly at p < 0.05 (Fisher LSD test).
Site (Abbreviation)
PropertyKahanu
(KAH)		McBryde
(MCB)		Ulutopia
(ULU)		Kāko’o ‘Ōiwi
(KKO)		p
IslandMauiKauiKauiO’ahu-
Latitude (N)20.79821.88621.96821.434-
Longitude (W)156.037159.493159.417157.811-
Elevation (m above sea level)1513110539-
Rainfall (mm per annum) §1980169317831580-
Soil order ‡AndisolOxisolOxisolInceptisol-
Soil series ‡HanaLihuePuhiHanalei-
Sand (%)5 (5) a6 (5) a6 (5) a6 (5) a0.107
Clay (%)86 (10) a85 (10) a81 (9) a85 (11) a0.968
Silt (%)9(10) a9 (10) a11 (9) a9 (10) a0.435
Bulk density (g/cm3)0.62 (0.08) b1.31 (0.09) a1.25 (0.08) a1.14 (0.09) a0.001
pH6.33 (0.12) b7.01 (0.14) a5.58 (0.33) c7.07 (0.33) ab0.003
CEC (cmol/kg) ‡34.617.512.513.9-
Bulk soil N (%)1.19 (0.18) a0.47 (0.21) b0.38 (0.40) b0.83 (0.51) ab0.023
Plant-available N (µg/g) †48 (5) a22 (6) b17 (13) b11 (14) b<0.001
Plant-available P (µg/g)16 (2) a22 (3) a12 (5) a15 (6) a0.395
N:P ratio4.1 (0.6) a1.2 (0.6) b1.7 (1.4) b0.7 (1) ab<0.001
Total P (µg/g)547 (133) b533 (111) c512 (21) bc831 (155) a0.021
Fe (%)4.36 (1) a3.59 (0.6) ab0.99 (0.1) b3.47 (0.6) ab0.035
Mn (%)0.149 (0.02) a0.051 (0.08) b0.021 (0.01) b0.146 (0.02) a0.011
Ca (%)0.739 (0.1) b0.311 (0.2) c0.485 (0.15) c2.719 (0.6) a<0.001
K (%)0.21 (0.08) a0.25 (0.02) a0.17 (1) a0.14 (0.04) a0.238
§ Dept of Land and Water Resources, State of Hawaii, and Western Regional Climate Center (https://wrcc.dri.edu; accessed 15 February 2026). ‡ Data from https://casoilresource.lawr.ucdavis.edu, accessed 15 February 2026. † Sum of NH4 and NO3.
Table 3. Levels of aggregate and bulk soil C pools, and AMF and microbial biomass in soils collected under Ma’afala trees in each study site. Values represent the mean with standard error in parentheses. Significant p values denoted in bold. Within rows, values with the same letter do not differ significantly at p < 0.05 (Fisher LSD test).
Table 3. Levels of aggregate and bulk soil C pools, and AMF and microbial biomass in soils collected under Ma’afala trees in each study site. Values represent the mean with standard error in parentheses. Significant p values denoted in bold. Within rows, values with the same letter do not differ significantly at p < 0.05 (Fisher LSD test).
Site ˆ
PropertyKAH ˆMCBULUKKOp
Bulk SOC (%)11.46 (1.5) a4.47 (1.3) b5.48 (0.58) b5.09 (0.86) b<0.001
Labile C (mg/kg)252 (31) a217 (35) a196 (84) a302 (98 a)0.411
Large macroaggregate OC (g/kg soil)7.40 (0.4) a2.33 (0.4) b3.18 (0.3) b2.51 (0.4) b<0.001
Small macroaggregate OC (g/kg soil)6.84 (0.3) a2.29(0.3) b2.31 (0.3) b3.55 (0.3) b<0.001
Microaggregate OC (g/kg soil)6.85 (0.4) a2.33 (0.4) b2.95 (0.4) b2.51 (0.4) b<0.001
Aggregate C:N8.7 (2) a8.3 (3) a6.5 (1) a4.3 (2) a0.246
Bulk soil C:N13 (6) a8 (3) a15 (6) a7 (2) a0.131
AMF hyphal pool (m/g soil)8.55 (0.8) a12.6 (0.9) a9.6 (2.0) a12.9 (2.4) a0.624
AMF root colonization (%), total49.8 (9) a50.1 (8) a46.9 (9) a59.8 (8) a0.857
AMF root colonization (%) by vesicles27 (21) a22 (23) a16 (56) a29 (65) a0.171
Microbial biomass C (mg/g)2.28 (0.02) a1.79 (0.84) a3.15 (0.61) a2.92 (0.23) a0.366
ˆ Site: KAH, Kahanu; MCB, McBryde; ULU, Ulutopia; KKO, Kāko’o‘ Ōiwi.
Table 4. Summary of significant Spearman-rank correlation coefficients among soil factors, large and small macroaggregate C, microaggregate C, and bulk SOC levels in Ma’afala cultivars across sites. ns, not significant (p > 0.05); * significant at p < 0.05.
Table 4. Summary of significant Spearman-rank correlation coefficients among soil factors, large and small macroaggregate C, microaggregate C, and bulk SOC levels in Ma’afala cultivars across sites. ns, not significant (p > 0.05); * significant at p < 0.05.
Carbon Pool
PropertyLarge MacroaggregateSmall MacroaggregateMicro-
Aggregate		Bulk SOC
Bulk density (g/cm3)−0.732 *−0.871 *−0.675 *−0.773 *
Soil P (µg/g)−0.591 *nsns−0.538 *
N:P ratio0.625 *ns0.568 *0.661 *
Bulk N (%)0.547 *0.535 *nsns
Soil K content (%)ns0.553 *nsns
Soil Fe content (%)ns0.563 *nsns
AMF root colonization (%)0.722 *nsnsns
Silt (%)ns−0.752 *nsns
Table 5. Summary of rank regression models for each C pool showing significant predictors.
Table 5. Summary of rank regression models for each C pool showing significant predictors.
C PoolPredictorEstimateS.E.t-ValuepModel R2p
S Macroaggregate(Intercept)0.6191.8220.3400.7360.4030.001
AMF productivity0.4360.2191.9870.053
Soil N0.0740.0174.393<0.001
L Macroaggregate(Intercept)−1.5281.609−0.9500.3480.552<0.001
AMF productivity0.6060.1993.0480.004
Microbial biomass0.9270.3942.3520.024
Soil N0.0760.0145.409<0.001
Resin P0.0250.0102.3780.022
Bulk SOC(Intercept)21.4946.6133.2500.0020.645<0.001
AMF productivity1.0410.1905.4740.000
Microbial biomass1.3280.3793.5000.001
Soil N0.0680.0144.7010.000
Soil P0.0530.0202.5820.013
Resin N−0.0130.005−2.7940.008
Resin P0.0580.0115.4820.000
Generalist fungal hyphae−0.5060.218−2.3190.025
Leaf C−0.5340.157−3.3930.001
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Gohd, L.; Egerton-Warburton, L.M.; Porter, E.; Dickinson, N.D.; Zerega, N.J.C.; Dybzinski, R. Domestication Level and Soil Fertility Differentially Alter Soil Carbon Sequestration Potential in Breadfruit (Artocarpus). Forests 2026, 17, 300. https://doi.org/10.3390/f17030300

AMA Style

Gohd L, Egerton-Warburton LM, Porter E, Dickinson ND, Zerega NJC, Dybzinski R. Domestication Level and Soil Fertility Differentially Alter Soil Carbon Sequestration Potential in Breadfruit (Artocarpus). Forests. 2026; 17(3):300. https://doi.org/10.3390/f17030300

Chicago/Turabian Style

Gohd, Lindsey, Louise M. Egerton-Warburton, Ellinore Porter, Noel Dakar Dickinson, Nyree J. C. Zerega, and Ray Dybzinski. 2026. "Domestication Level and Soil Fertility Differentially Alter Soil Carbon Sequestration Potential in Breadfruit (Artocarpus)" Forests 17, no. 3: 300. https://doi.org/10.3390/f17030300

APA Style

Gohd, L., Egerton-Warburton, L. M., Porter, E., Dickinson, N. D., Zerega, N. J. C., & Dybzinski, R. (2026). Domestication Level and Soil Fertility Differentially Alter Soil Carbon Sequestration Potential in Breadfruit (Artocarpus). Forests, 17(3), 300. https://doi.org/10.3390/f17030300

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop