Article pubs.acs.org/JAFC
Cite This: J. Agric. Food Chem. 2018, 66, 9895−9906
Soil Perturbation in Mediterranean Ecosystems Reflected by Differences in Free-Lipid Biomarker Assemblages Pilar Tinoco,† Gonzalo Almendros,*,‡ and Jesús Sanz§ †
Universidad Alfonso X el Sabio, Campus de Villanueva de la Cañada, Avenida Universidad 1, E-28691 Madrid, Spain MNCN, CSIC, Serrano 115B, E-28006 Madrid, Spain § Instituto de Química Orgánica General, CSIC, Juan de la Cierva 3, E-28006 Madrid, Spain Downloaded via UNIV OF LOUISIANA AT LAFAYETTE on September 26, 2018 at 08:09:34 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
‡
ABSTRACT: Environmental information provided by free lipids in soil samples collected from control and disturbed plots (Madrid, Spain) was assessed by comparing molecular assemblages of terpenoids and distribution patterns of alkanes and fatty acids (FAs) analyzed by gas chromatography−mass spectrometry (GC-MS). Wildfires in pine forests led to increased proportions of retene, dehydroabietin, and simonellite. Friedo-oleananes were characteristic in soils under angiosperms, and norambreinolide-type diterpenes were characteristic in soils encroached by Cistus bushes. Steroids were major compounds in pastured sites. Enhanced Shannon’s lipid biodiversity index in disturbed soils compared with in control soils suggested patterns of recent lipids overlapping a preserved original lipid signature. The extent of the environmental impacts was illustrated as Euclidean distances between paired control and disturbed sites calculated using the compounds in alkyl homologous series as descriptors. As expected, reforestation, bush encroachment, wildfires, and cultivation were reflected by changes in the molecular record of lipids in soils. KEYWORDS: alkane, biomarker, terpene, fatty acid, molecular tracer, signature lipid
■
INTRODUCTION The lipid fraction of soil is often considered a valuable source of environmental information, representing a continuous molecular record shedding light on climate change and the intensity of organic-matter (OM) turnover.1−4 In fact, whereas the soil lipid fraction originally consists of a heterogeneous molecular assemblage inherited mainly from plants and microorganisms,5,6 further abiotic or microbial transformation of these biogenic lipids provides additional compounds in the soil.7−11 The dynamics of soil lipids is complex because apart from the above processes, lipid mixtures are subjected to continuous biodegradation of their comparatively labile molecules. Nevertheless, lipid molecules can also be included into organomineral structures in progressively transformed soil organic-matter pools12−14 with improved preservation of compounds associated with small-size-aggregate fractions.15 Other fractions of lipids present in soil, which require chemical treatments for their release in the form of free molecules, may correspond both to cellular constituents in still undegraded biomass and to condensation products incorporated into humic substances. Therefore, the balance between biodegradation and humification processes could be monitored by the minor but diagnostic fraction of lipids temporarily free in the soil, which are the subject of this study. Hence, the surviving lipids that can be directly isolated from soil in the form of free compounds could represent a molecular signature for reconstructing recent and past soil processes, because their occurrence could in most cases depend on the environmental impact on terrestrial ecosystems.16−19 Finally, specific soil lipids are also important because of their roles in soil processes, as antimicrobial agents,20−23 in the alellopathic interactions between higher plants,1,24,25 and in insect−host-plant relationships26 as well as through their effects © 2018 American Chemical Society
on soil physical properties, mainly aggregation and soil water repellence.27−30 In the present study, the aim was an analytical comparison between free soil lipids from relict forests and those from soils in adjacent sites under the same climatic conditions and the same original geological substrate but affected by environmental perturbation processes typical of continental Mediterranean ecosystems in central Spain. Both individual signature compounds and the distribution patterns of the major series of alkyl compounds were analyzed in 16 well-characterized ecosystems in central Spain, consisting of control forests and sites affected by (i) clearing and bush encroachment, (ii) cultivation, (iii) wildfire, and (iv) reforestation with pine. The specific objectives were to assess the extent of the biogeochemical changes undergone by the soils as a result of the environmental perturbation, to test the responses of the biomarker assemblages to environmental changes, and to compare hypothetical new proxies for these perturbations.
■
MATERIALS AND METHODS
Sampling. Soil samples from eight contrasting continental Mediterranean forest ecosystems in Madrid (central Spain) representative for sclerophyllic (oak), mesophilic (chestnut and ash), and pine forests were collected. In addition to the control sites, another set consisting of eight altered neighbor soils with the same climatic, geologic, and topographic features were sampled.31 A two-character labeling code was used to refer to the sampling sites: the first letter (see below) was the code for the soil series; this was followed by odd Received: Revised: Accepted: Published: 9895
March 22, 2018 August 24, 2018 September 4, 2018 September 4, 2018 DOI: 10.1021/acs.jafc.8b01483 J. Agric. Food Chem. 2018, 66, 9895−9906
Article
Journal of Agricultural and Food Chemistry Table 1. Soil Classification and General Analytical Characteristics of Sampled Soils in Central Spain universal transverse mercator (UTM)
label
ecosystem
soil type32
R1
relict evergreen oak forest
R2
reforested pine forest
R3
relict evergreen oak forest
R4
reforested pine forest
C1
relict oak forest
C2
cultivated wheat field
C3
relict ash forest
C4
pastured site
B1
relict chestnut forest
B2
secondary bush
B3
control evergreen oak forest
B4
secondary bush
F1
pine forest
F2
burned pine forest (highintensity fire) pine forest
Eutric Cambisol (loamic, humic) Eutric Cambisol (loamic, humic) Cambic Folic Phaeozem (loamic, humic) Eutric Cambisol (loamic, humic) Eutric Folic Cambisol (loamic, humic) Rendzic Leptosol (loamic) Eutric Cambisol (loamic, humic) Eutric Cambisol (loamic, humic) Eutric Folic Cambisol (loamic, humic) Dystric Cambisol (loamic, humic) Eutric Cambisol (loamic, humic) Eutric Cambisol (loamic, humic) Calcaric Cambisol (loamic, humic) Calcaric Cambisol (loamic, humic) Cambic Umbrisol (loamic, hyperdystric) Hyperdystric Cambisol (loamic, humic)
F3
F4
burned pine forest (mainly affecting forest canopy)
parent rock
vegetation
4488-411
granite
4488-411
granite
Quercus ilex subsp. ballota in addition to Rosmarinus off icinalis, Daphne gnidium, and Cistus ladanifer Pinus pinea in addition to Cistus ladanifer
4490-403
granite
Quercus pyrenaica in addition to Prunus spinosa and Rosa canina
4490-403
granite
Pinus pinaster in addition to Cistus ladanifer
4472-476
limestone
Quercus ilex subsp. ballota
4472-476
limestone
Triticum aestivum
4526-451
granite
Fraxinus angustifolia in addition to Rosa canina and Paeonia coriacea
4526-451
granite
4465-371
granite
Poa bulbosa, Trifolium dubium, and Trifolium campestre in addition to Fraxinus angustifolia, Rosa canina, Paeonia coriacea, and Micropyrum tenellum Castanea sativa, Genista sp., Rosa canina, Vicia sp., and Poaceae
4465-371
granite
Cytisus scoparius, Genista sp., Retama sphaerocarpa, and Lavandula stoechas
4493-401
granite
4493-401
granite
4459-450
limestone
Quercus ilex subsp. ballota, Cistus ladanifer, Rosmarinus of ficinalis, and Daphne gnidium Cistus ladanifer, Daphne gnidium, Retama sphaerocarpa, and residual forest of Quercus ilex subsp. ballota Pinus halepensis, Eryngium campestre, and Reseda sp.
4459-450
limestone
Pinus halepensis, Eryngium campestre, and Reseda sp.
4553-452
gneiss
Pinus sylvestris, Erica sp., Pteridium aquilinum, and Cytisus scoparius
4553-452
gneiss
Pinus sylvestris, Erica sp., Pteridium aquilinum, and Cytisus scoparius
Table 2. Analytical Characteristics of Soil Samples (0−10 cm Depth) sample R1 R2 R3 R4 C1 C2 C3 C4 B1 B2 B3 B4 F1 F2 F3 F4 LSDe
altitude (m asl)
slope (%)
pH (H2O)
clay (g kg−1)
textural type (USDA)
Ca (g kg−1)
C/N
lipidb g kg−1
CECc (cmolc kg−1)
Sd (cmolc kg−1)
880 855 1150 1240 870 870 950 950 840 825 1015 990 630 624 1580 1615
5 0 15 15 0 0 2 2 20 5 15 20 0 8 20 15
7.5 6.0 5.9 6.2 7.9 8.4 6.8 6.0 6.2 6.2 7.1 6.5 6.9 8.7 4.6 5.7 0.1
101 30 158 55 27 228 113 126 134 45 51 86 120 125 77 64 30
sandy loam loamy sand sandy loam sandy loam silt loam silt loam sandy loam sandy loam loam loamy sand sandy loam sandy loam silt loam sandy loam sandy loam sandy loam
170 67 39 65 95 17 92 49 61 57 44 88 213 39 69 64 5
16.5 24.7 11.3 29.6 15.3 12.7 13.3 13.1 15.5 13.0 14.2 17.1 14.5 12.3 14.7 8.2 1.4
1.45 0.59 0.17 0.54 0.29 0.10 0.49 0.18 0.08 0.13 0.13 0.37 9.50 0.05 0.42 0.14 0.02
47.1 12.2 17.2 19.0 41.2 21.2 21.8 12.6 17.1 10.6 15.0 25.0 70.6 13.5 23.5 41.2 1.8
33.9 7.6 10.4 13.3 41.2 21.2 20.8 10.2 11.1 4.7 7.5 16.8 49.7 13.5 2.1 4.0 5.1
a Total oxidizable soil C. bSoxhlet extraction with petroleum ether. cCation-exchange capacity. dSum of exchangeable bases (Na+ + K+ + Ca2+ + Mg2+). eLeast-significant difference based on adjacent spatial replicates.
and the soil material (the whole O horizon) was collected with a spade. In order to obtain representative samples averaging the possible spatial variability in the plots, composite samples were taken from each plot. Each individual sample was prepared by mixing three soil subsamples of ca. 1 L from the points of a virtual triangle of ca. 100 m per side. The samples were air-dried; litter and root fragments were picked out by
numbers in the case of the relict ecosystems and even numbers for the perturbed ones. The distances between sampling points in the paired soils were always C20/C20/