Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
1
Using plant traits to understand the contribution of biodiversity effects to 1
community productivity in an agricultural system 2
3
Nadine Engbersen1, Laura Stefan1, Rob W. Brooker2, Christian Schöb1 4
1 Department of Environmental Systems Science, Swiss Federal Institute of Technology, ETH Zurich, 5
Universitätstrasse 2, 8092 Zurich, Switzerland. 6
2 The James Hutton Institute, Craigiebuckler, Aberdeen, AB15 8QH, Scotland, UK 7
Summary 8
Increasing biodiversity generally enhances productivity through selection and complementarity 9
effects not only in natural but also in agricultural systems. However, explaining why diversity 10
enhances productivity remains a central goal in agricultural science. 11
In a field experiment, we constructed monocultures, 2- and 4-species mixtures from eight crop 12
species with and without fertilizer and both in temperate Switzerland and semi-arid Spain. We 13
measured environmental factors and plant traits and related these in structural equation models 14
to selection and complementarity effects to explain yield differences between monocultures and 15
mixtures. 16
Increased crop diversity increased yield in Switzerland. This positive biodiversity effect was 17
driven to almost same extents by selection and complementarity effects, which increased with 18
plant height and C:N ratio, respectively. Also, ecological processes driving yield increases from 19
monocultures to mixtures differed from those responsible for yield increases through the 20
diversification of mixtures. 21
While selection effects were mainly driven by one species, complementarity effects were linked 22
to higher productivity per unit N. Yield increases due to mixture diversification were driven 23
only by complementarity and were not mediated through the measured traits, suggesting that 24
ecological processes beyond those measured in this study were responsible for positive diversity 25
effects on yield beyond 2-species mixtures. 26
27
Key words: agroecology, biodiversity–productivity, complementarity and selection effects, 28
intercropping, plant traits 29
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
2
1. Introduction 30
Plant primary productivity increases with higher species diversity (e.g. Tilman et al. (2001); Cardinale 31
et al. (2006)). While the majority of research in this area has been done in perennial systems (Tilman et 32
al. 2001; Weisser et al. 2017; Huang et al. 2018), recent studies have demonstrated similar effects in 33
annual systems (Li et al. 2014; Brooker et al. 2015; Stomph et al. 2020). Intercropping of annual crops, 34
where at least two crop species are grown in close proximity at the same time, is therefore a promising 35
application of agroecological concepts. Intercropping can make use of species complementarity and 36
beneficial above- and belowground facilitation and niche differentiation, which can ultimately lead to 37
overyielding, i.e. increased yield in mixture compared with the average of the monocultures 38
(Vandermeer 1989). 39
Two main mechanisms have been proposed to explain positive biodiversity–productivity 40
relationships. First, sampling or selection effects (hereafter: selection effects) encompass the greater 41
probability that more diverse communities include highly productive species or functional groups, which 42
then account for the majority of productivity (Huston 1997; Tilman et al. 1997). Enhanced ecosystem 43
functioning in diverse agroecosystems can be driven by selection effects (i.e. communities with more 44
species are more likely to host a high-performing species). For instance, in China, a hotspot of 45
intercropping, a recent meta-analysis has shown that 10% of all yield gain of intercropping compared to 46
sole cropping was due to selection effects (Li et al. 2020). 47
The second mechanism is complementarity through resource partitioning or facilitation. Resource 48
partitioning involves more diverse communities containing species with contrasting demands on 49
resources, which leads to a more complete exploitation of available resources in diverse plant 50
communities compared with monocultures and hence increased productivity (Tilman et al. 1997; Loreau 51
and Hector 2001). Facilitation involves plants altering their environment in a way that is beneficial to at 52
least one co-occurring species (Brooker et al. 2008). However, knowledge about the precise mechanisms 53
that lead to overyielding in crop mixtures and how environmental conditions can alter these mechanisms 54
still remains incomplete (Duchene et al. 2017). 55
Resource partitioning can occur across spatial, temporal or chemical gradients. For example, 56
different rooting depths allow plants to take up water or nutrients from different soil layers, thus limiting 57
competition. While this has been observed for water uptake (Miyazawa et al. 2009), the evidence of 58
resource partitioning for soil nutrients as a driver of biodiversity effects is less clear (Von Felten et al. 59
2012; Jesch et al. 2018). Aboveground, diverse communities have been observed to harbor more diverse 60
canopy growth forms allowing for a more complete use of photosynthetically active radiation (PAR) 61
(Spehn et al. 2002; Fridley 2003), leading in some cases to yield advantages in mixtures compared with 62
monocultures (Bedoussac and Justes 2009). 63
Facilitation involves plants altering their environment in a way that is beneficial to at least one 64
co-occurring species (Brooker et al. 2008). Facilitation among plants can happen via either the 65
enrichment of resource pools by one plant for neighboring plants or the mediation of physical stress. 66
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
3
The facilitative benefits of legume neighbors for enhancing soil N are well known, for example for 67
cereals intercropped with legumes (Jensen 1996; Hauggaard-Nielsen et al. 2001; Spehn et al. 2002; 68
Temperton et al. 2007). Other below-ground facilitative mechanisms shown to occur in intercrops 69
include enrichment of resource pools by hydraulic lift, which not only facilitates water uptake (Sekiya 70
and Yano 2004), but can also enhance nutrient mobilization and lead to increased nutrient status of the 71
intercrop (Sun et al. 2013). Other evidence of facilitation mediating physical stress in crop systems come 72
from studies of barley variety mixtures, where a denser canopy structure, shading of the soil surface, 73
and thus reduced evaporation were observed to decrease the soil temperature due (Cooper et al. 1987). 74
Also, plant species can alleviate the microclimate for their neighbors by mediating wind, heat or 75
photoinhibition (Wright et al. 2017). 76
However, despite these examples of different resource partitioning and facilitation mechanisms 77
occurring in crop systems, knowledge about the precise mechanisms that lead to overyielding in crop 78
mixtures, and how environmental conditions can alter these mechanisms, still remains incomplete 79
(Duchene et al. 2017). While there is abundant evidence for the presence of facilitation and 80
complementarity in diverse agricultural systems, the presence of these processes alone does not 81
guarantee overyielding. 82
Here, we applied ecological knowledge and methods to agriculture by assessing to what extent 83
complementarity and selection effects drive yield gains in crop mixtures compared to crop monocultures 84
and how these effects are related to plant functional traits and environmental factors. By linking 85
frequently-used plant traits to yield gains in mixtures, we can improve our ability to predict optimal 86
species combinations which can help to promote sustainable agricultural production through 87
intercropping. We used four plant traits indicative of resource use: leaf dry matter content (LDMC), 88
specific leaf area (SLA), carbon to nitrogen ratio (C:N ratio) and plant height and two environmental 89
factors: PAR and volumetric soil water content (VWC). We used a structural equation model (SEM) to 90
assess whether overyielding is driven by selection or complementarity effects or by a combination of 91
both. Furthermore, we used this hierarchical model to understand the context–dependence of the 92
complementarity and selection effects and how they are linked to plant functional traits. 93
94
2. Materials and Methods 95
2.1. Site description 96
The study was carried out in two outdoor experimental gardens in Zürich, Switzerland and Torrejón el 97
Rubio, Cáceres, Spain. The Swiss site was located at an altitude of 508 m a.s.l. (47°23’45.3” N 98
8°33’03.6” E), the Spanish site at 290 m a.s.l. (39°48’47.9” N 6°00’00.9” W). The regional climate in 99
Switzerland was classified according to Köppen-Geiger (Kottek et al. 2006) as warm temperate, fully 100
humid with warm summers while in Spain it was classified as warm temperate, summer dry with hot 101
summers. Precipitation in Zürich during the growing season 2018 (April – August) was 572 mm, and in 102
Spain 218 mm (February – June). Daily average hours of sunshine during the growing season were 5.8h 103
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
4
in Zürich and 8.4h in Cáceres. Temperatures between the two sites were similar: average of daily mean, 104
minimum and maximum temperatures were 14.0 °C, 9.3 °C and 18.6°C in Zürich and 14.6 °C, 9.6°C 105
and 19.6°C in Cáceres. All climatic data are from the Deutsche Wetterdienst (www.dwd.de). 106
The experimental garden at each location covered 98 m2, divided into 392 square plots of 0.25 m2 107
and 40 cm depth. However, plots were open at the bottom and allowed root growth beyond 40 cm. In 108
Switzerland, the plots were arranged in 14 beds of 7 × 1 m, with two rows of 14 plots, resulting in 28 109
plots per bed. In Spain, the plots were arranged in 14 beds of 10 × 1 m, with two rows of 20 plots, 110
resulting in 40 plots per bed. The plots were filled with local, unenriched agricultural soil. Soil structure 111
and composition therefore differed between the sites. In Switzerland, soil was composed of 45% sand, 112
45% silt, 10% clay and contained 0.19% nitrogen, 3.39% carbon, and 333 mg total P/kg with a mean 113
pH of 7.25. Spanish soils consisted of 78% sand, 20% silt, 2% clay and contained 0.05% nitrogen, 0.5% 114
carbon, 254 mg total P/kg with a mean pH of 6.3. 115
The experimental gardens were irrigated throughout the growing season with the aim of 116
maintaining the differences in precipitation between the two sites but assuring survival of the crops 117
during drought periods. In Switzerland, the dry threshold was set to 50% of field capacity, with a target 118
of 90% of field capacity, while in Spain the thresholds were 17% and 25% of field capacity, respectively. 119
Automated irrigation was configured such that irrigation would start if the dry threshold was reached, 120
and irrigate until the target threshold was reached. 121
At each site, half of the beds were chosen randomly to be fertilized with N-P-K (1-1.7-1) while 122
the other half served as unfertilized controls. Fertilizer was applied three times: 50 kg/ha just before 123
sowing, another 50 kg/ha when wheat was tillering and 20 kg/ha when wheat was flowering. 124
2.2. Crop species and cultivars 125
At each site, experimental communities were constructed with eight annual crop species of agricultural 126
interest. The crop species belonged to four different phylogenetic groups: Triticum aestivum (wheat, C3 127
grass) and Avena sativa (oat, C3 grass), Lens culinaris (lentil, legume) and Lupinus angustifolius (lupin, 128
legume), Linum usitatissimum (flax, herb [superrosids]) and Camelina sativa (false flax, herb 129
[superrosids]), Chenopodium quinoa (quinoa, herb [superasterids]) and Coriandrum sativum (coriander, 130
herb [superasterids]). The four phylogenetic groups were based on their phylogenetic distances: Cereals 131
diverged from the other groups 160 million years ago (mya); superasterid herbs diverged from 132
superrosid herbs including legumes 117 mya and finally, legumes diverged from the other superrosid 133
herbs 106 mya (TimeTree). Phylogenetic distance was chosen as a criterion for functional similarity as 134
it is often positively correlated with functional diversity and acts as a proxy to assess the impacts of 135
species diversity on ecosystem functions (Mouquet et al. 2015). At both sites, we grew commercial 136
cultivars typically used for organic farming in Switzerland (Table S1). While these were bred for a Swiss 137
climate, their cultivation in Spain demonstrated the ability of these cultivars to adapt to a climate change-138
type scenario, with conditions considerably warmer and drier than in Switzerland. 139
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
5
2.3. Experimental crop communities 140
Experimental crop communities at both sites consisted of eight different monocultures, 24 different 2-141
species mixtures consisting of two different phylogenetic groups and 16 different 4-species mixtures 142
consisting of four different phylogenetic groups. Every combination of 2-species mixture with two 143
species from different phylogenetic groups and every possible 4-species mixture with species from four 144
different phylogenetic groups were planted. Each experimental community was replicated two times. 145
Plots were randomized within each country and fertilizer treatment. Each monoculture and mixture 146
community consisted of one, two or four crop species planted in four rows. The row order of the species 147
was chosen randomly. Sowing densities differed among phylogenetic groups and were based on current 148
cultivation practice (Table S1). Sowing was done by hand in early February 2018 in Spain and early 149
April 2018 in Switzerland. 150
2.4. Data collection 151
Leaf traits were measured at the time of flowering (May 2018 in Spain, 94 days after sowing; June 2018 152
in Switzerland, 65 days after sowing). Three individuals per crop species per plot were randomly marked 153
and their height measured with a ruler from the soil surface to the highest photosynthetically active 154
tissue. Of each marked individual, 1 to 10 healthy leaves were sampled and immediately wrapped in 155
moist cotton and stored overnight at room temperature in open plastic bags. For the subsequent leaf trait 156
measurements (specific leaf area [SLA] and leaf dry matter content [LDMC]) we followed standard 157
protocols (Cornelissen et al. 2003). The leaf samples were blotted dry to remove any surface water and 158
immediately weighed to produce a value for saturated weight. They were instantly scanned with a flatbed 159
scanner (CanoScan LiDE 120, Canon) and oven-dried in a paper envelope at 80 °C for at least 48 h. 160
After drying, the leaf samples were reweighed to obtain dry weight. Values for dry matter content were 161
calculated as the ratio of leaf dry mass (mg) to saturated leaf mass (g). The leaf scans were analyzed for 162
leaf area with the image processing software ImageJ 1 (Rasband 1997-2018). SLA was calculated as the 163
ratio of leaf area (cm2) to dry mass (g) and is a measure of the investment in dry mass per light-164
intercepting surface and used to assess light acquisition strategies; higher SLA values can be observed 165
under low light conditions as a larger leaf area per leaf mass (= thinner leaves) enables increased surface 166
for light capture at lower C costs (Weiher et al. 1999; Evans and Poorter 2001). LDMC is often inversely 167
related to SLA but also includes soil-water relations and hence indicates impacts of resource conditions, 168
including water (Pérez-Harguindeguy et al. 2013) 169
At the time of harvest (duration of crop growth from sowing to harvest given in table S2) all crops 170
in each plot were harvested. The three marked individuals used for the trait measurements were collected 171
separately, while all remaining plants per crop species per plot were pooled together. Plant shoots were 172
cut at the soil surface and biomass and seeds were separated. The total number of individuals per crop 173
species per plot was recorded. Fruits were air-dried. Afterwards, seeds were separated from chaff with 174
a threshing machine (in Switzerland: Allesdrescher K35, Baumann Saatzuchtbedarf, Germany; in Spain: 175
Hege 16, Wintersteiger, Austria). Vegetative biomass was oven-dried at 80 °C until constant weight. 176
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
6
Interception of PAR by the plant canopy was measured weekly with a LI-1500 (LI-COR 177
Biosciences GmbH, Germany). In each plot, three PAR measurements were taken around noon by 178
placing the sensor on the soil surface in the center of each of the three in-between rows. Light 179
measurements beneath the canopy were put into context through simultaneous PAR measurements of a 180
calibration sensor, which was mounted on a vertical post at 2 m above ground in the middle of the 181
experimental garden. FPAR (%) indicates the percentage of PAR that was intercepted by the crop 182
canopy. VWC in the soil was measured weekly with a ML3 ThetaProbe Soil Moisture Sensor (Delta-T, 183
Cambridge). The measurements were taken in the center of each of the three in-between rows per plot. 184
For further data analysis, we used FPAR and VWC values from the week of leaf trait measurements 185
(Spain: 92 days after sowing; Switzerland: 62 days after sowing). 186
2.5. Plant N analyses 187
For chemical analyses of the plant tissue we pooled the three dried leaf samples of the marked 188
individuals per plot and per crop species. Leaf samples were ground for 20 minutes in 1.2 ml tubes with 189
two stainless steel beads in a bead mill (TissueLyserII, Qiagen). Afterwards, either 100 mg (if available 190
in sufficient amount) or 4 mg (+/- 1 mg) (if the sample was too small) of ground leaf material were 191
weighed into tin foil cups or 5 × 9 mm tin capsules and analyzed for C and N contents. The 400 large 192
samples were analyzed on a LECO CHN628C elemental analyzer (Leco Co., St. Joseph, USA) and the 193
505 small samples on a PDZ Europa 20-20 isotope ratio mass spectrometer linked to a PDZ Europa 194
ANCA-GSL elemental analyzer (Sercon Ltd., Cheshire, UK), respectively. Eight samples were cross-195
referenced on both analytical devices (Fig. S1) and measured values from LECO were corrected to 196
account for the differences between the devices (correction factors are 1.0957 for N and 1.026 for C, 197
respectively). 198
2.6. Data analysis 199
Prior to data analysis the dataset was screened for missing data or incorrect data recording and the whole 200
plot was removed when any of these occurred. A total of 467 plots remained, 231 in Switzerland and 201
236 in Spain. 202
All statistical analyses were performed in R version 3.6.0. (R Core Team 2019). We used general 203
linear mixed-effects models using restricted maximum likelihood estimation to explain yield at the 204
community-level. We assessed the significance of the fixed effects using type-I ANOVA and the 205
Satterthwaite approximation of denominator degrees of freedom (lme4 (Bates et al. 2015) and lmerTest 206
(Kuznetsova et al. 2017) packages). Yield was log transformed. The fixed effects of the model were 207
country (Switzerland versus Spain), fertilizer (fertilized versus unfertilized), species number (2 versus 208
4) nested in diversity (monocultures vs mixtures), and interactions among the fixed effects (except 209
between the nested terms). Random terms were species composition, garden bed and the interactions of 210
garden bed with diversity and garden bed with species number. To test for effects of diversity within 211
each country, we performed pairwise multiple comparison Tukey tests. 212
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
7
We used the same linear mixed-effects models as mentioned above to analyze treatment effects 213
on environmental factors (FPAR, VWC), plant traits (SLA, LDMC, height, leaf N, C:N ratio) and 214
biodiversity effects (selection and complementarity effects) with country, fertilizer, species number 215
nested in diversity and interactions among these as fixed effects. Random terms were species 216
composition, garden bed and their interaction. Response variables were log transformed, except for 217
biodiversity effects. 218
Using structural equation modeling (SEM) we analyzed the sign and intensity of treatment effects 219
and abiotic factors on plant traits, biodiversity effects and community-level yield, based on our complex 220
a priori model (Fig. S2). To explain differences in community-level yield between mixtures and 221
monocultures, we calculated Δyield as the difference between the summed community-level yields of 222
all species in a mixture plot and the average of the mean community-level yields of all monocultures 223
corresponding to the species in the mixture plot. Thus, Δyield compares the observed yield in the mixture 224
with the expected yields in a mixture based on their yields in a monoculture. Δyield was square root 225
transformed to meet the assumptions of normal distribution. We quantified the biodiversity effect 226
(Δyield) and its two components, the complementarity and selection effects according to Loreau and 227
Hector (Loreau and Hector 2001). 228
∆𝑦𝑖𝑒𝑙𝑑 = 𝑁 ∙ ∆𝑅𝑌̅̅ ̅̅ ̅̅ ∙ �̅� + 𝑁 ∙ 𝑐𝑜𝑣(∆𝑅𝑌, 𝑀) (1) 229
Where N is the number of species in the plot. ΔRY is the deviation from expected relative yield 230
of the species in mixture in the respective plot, which is calculated as the ratio of observed relative yield 231
of the species in mixture to the yield of the species in monoculture. The first component of the 232
biodiversity effect equation (𝑁 ∙ ∆𝑅𝑌̅̅ ̅̅ ̅̅ ∙ �̅�) is the complementarity effect, while the second component 233
(𝑁 ∙ 𝑐𝑜𝑣(∆𝑅𝑌, 𝑀)) is the selection effect. We calculated ΔVWC and ΔFPAR according to equation 2. 234
∆𝑉𝑊𝐶 = 𝑉𝑊𝐶̅̅ ̅̅ ̅̅ ̅𝑚𝑖𝑥 − 𝑉𝑊𝐶̅̅ ̅̅ ̅̅ ̅
𝑚𝑜𝑛𝑜 (2) 235
Where 𝑉𝑊𝐶̅̅ ̅̅ ̅̅ ̅𝑚𝑖𝑥 is the average of all three measurements of VWC per mixture plot and 236
𝑉𝑊𝐶̅̅ ̅̅ ̅̅ ̅𝑚𝑜𝑛𝑜 the average of all three measurements of VWC of the respective monoculture plots. The same 237
was calculated for ΔFPAR. To scale up plant trait measurements to the community level, community-238
weighted means in mixtures and monocultures were used to calculate ΔSLA, ΔLDMC, ΔC:N ratio and 239
Δheight: 240
∆𝑆𝐿𝐴 = 𝐶𝑊𝑀. 𝑆𝐿𝐴𝑚𝑖𝑥 − 𝐶𝑊𝑀. 𝑆𝐿𝐴𝑚𝑜𝑛𝑜 (3) 241
Where CWM.SLAmix is the community-weighted mean of SLA of all species in a mixture plot 242
and CWM.SLAmono the community-weighted mean of SLA in monoculture of all the species in the 243
respective plot. Aboveground biomass of each species was used as weights. ΔLDMC, ΔC:N ratio and 244
Δheight were calculated likewise according to formula 3. The SEM was constructed using LAVAAN 245
(Rosseel 2012) in R. For each variable, we report the conditional coefficient of determination (R2c), 246
which represents the variation explained by fixed and random effects and was calculated with the 247
MUMIN package (Bartoń 2019). 248
249
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
8
3. Results 250
3.1. Response of community-level yield to treatments 251
Community-level yield was significantly affected by country, with 88% higher yields in Switzerland 252
compared with Spain (Fig. 1a, Table 1). The addition of fertilizer did not show any effect on yield in 253
either country nor on either diversity level. Mixtures produced significantly more yield than 254
monocultures –and 4-species mixtures yielded more than 2-species mixtures (Fig. 1a, Table 1). Tukey 255
post hoc tests revealed that productivity of mixtures was enhanced, particularly in Switzerland (Table 256
S3), where 4-species mixtures yielded 30% more than 2-species mixtures and 93% more than 257
monocultures. Also, 2-species mixtures in Switzerland yielded 48% more than monocultures. In Spain, 258
yield did not respond significantly to different diversity levels. The presence of a legume in mixtures 259
increased yields in Spain by 72% but decreased it by 23% in Switzerland (Fig. 1b, Table S3). The three-260
way interaction of country × fertilizer × legume indicates that mixtures without legumes yielded more 261
in Switzerland in both fertilizer treatments, while in Spain mixtures with legumes had higher yields, but 262
only in unfertilized soils (Fig. 1b, Table S3). 263
264
Figure 1: Community-level yields in kg dry weight per m2 visualizing the significant results from table 265
1. Differences in community-level yield between countries (a), diversity levels (a), mixture 266
diversification (a), between the two-way interactions country × legume (b), country × diversity (a) and 267
country × mixture diversification (a) and between the three-way interaction country × fertilizer × 268
legume (b). 269
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
9
Table 1: Results of mixed effects ANOVA testing effects of the different treatments (country, fertilizer, 270
legume, diversity (monocultures vs. mixtures) and mixture diversification (2- vs. 4-species mixtures) on 271
community-level yield. SS: Sum of squares, MS: mean of squares, numDF: degrees of freedom of term, 272
denDF: degrees of freedom of error term, F-value: variance ratio, P: error probability. P-values in bold 273
are significant at α = 0.05 (* P < 0.05, ** P < 0.01, *** P < 0.001). n = 315 274
SS MS numDF denDF F-value P
country 4.691 4.691 1 24.98 12.64 0.002**
fertilizer 1.192 1.192 1 23.75 3.212 0.086
legume 0.201 0.201 1 43.24 0.542 0.466
diversity 1.524 1.524 1 42.67 4.107 0.049*
mixture diversification 1.933 1.933 1 43.78 5.209 0.027*
country × fertilizer 0.384 0.384 1 23.65 1.035 0.319
country × legume 16.99 16.99 1 257.1 45.77 8.91E-1***
country × diversity 15.01 15.01 1 79.23 40.45 1.20E-08***
country × mixture diversification 24.95 24.95 1 246.2 67.22 1.34E-14***
fertilizer × legume 0.356 0.356 1 254.9 0.958 0.329
fertilizer × diversity 0.059 0.059 1 76.7 0.158 0.692
fertilizer × mixture diversification 0.014 0.014 1 244.7 0.037 0.849
country × fertilizer × legume 1.518 1.518 1 253.9 4.089 0.044*
country × fertilizer × diversity 0.488 0.488 1 77.59 1.315 0.255
country × fertilizer × mixture
diversification 0.997 0.997 1 245.5 2.685 0.103
275
3.2. Environmental factors, plant traits and biodiversity effects 276
Neither environmental factors nor plant traits showed significant variation in response to the diversity 277
treatments (Fig. 2, Table S4, S5). Reduced water input resulted in a significantly lower volumetric soil 278
water content (VWC) in Spain. However, VWC did not vary significantly in response to fertilizer or 279
diversity treatments (Fig. 2a, Table S5). Both plant height and FPAR were significantly higher in 280
Switzerland than in Spain, indicating that canopy closure was more complete and that vegetative growth 281
was generally stronger in Switzerland than in Spain (Fig. 2b,e) and in fertilized compared with 282
unfertilized treatments (Table S5). SLA of crops was significantly higher in Switzerland compared with 283
Spain and LDMC showed an opposite behavior, with higher values in Spain than in Switzerland (Fig. 284
2c,d). Neither LDMC nor SLA responded significantly to fertilizer or diversity treatments. Leaf N and 285
C:N ratio both indicate significantly higher leaf N levels in Switzerland than in Spain (Fig. 2f,g, Table 286
S4, S5). Complementarity effects were stronger in Switzerland than in Spain and stronger in 4-species 287
than in 2-species mixtures in Switzerland (Fig. 2h). Selection effects showed no response to any 288
treatment factor (Fig. 2i). 289
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
10
290
Figure 2: Community-level means for the environmental factors volumetric soil water content (VWC) 291
(a) and absorption of photosynthetically active radiation (FPAR) (b), the plant traits specific leaf area 292
(SLA) (c), leaf dry matter content (LDMC) (d), plant height (e), leaf N (f) and C:N ratio (g) and the 293
biodiversity effects, divided into complementarity (h) and selection effects (i). Data are shown for both 294
countries and separated by levels of diversity. Complementarity and selection effects are only available 295
for mixtures. Brackets indicate significant differences between treatments and labels above brackets 296
indicate which treatment (mix. div. = mixture diversification) was significant at α = 0.05 (* P < 0.05, 297
** P < 0.01, *** P < 0.001). 298
299
300
301
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
11
3.3. Structural equation model to explain community-level yields 302
The SEM showed a good fit to the data (d.f = 5, p = 0.572, CFI = 1) and explored the links between 303
experimental treatment factors (country, fertilizer, mixture diversification), environmental factors 304
(ΔFPAR, ΔVWC) and the effect of these on the relevant plant traits (ΔSLA; ΔLDMC; ΔC:N ratio, 305
Δplant height) and finally linked these through selection and complementarity effects to yield 306
differences between crop monocultures and mixtures (Δyield) (Fig. 3). 307
The effect of complementarity on Δyield was 1.2 fold stronger than the selection effect (Fig. 3). 308
Both biodiversity effects were negatively correlated with each other, thus when one effect increased, the 309
other decreased. The selection effect was only related to Δheight, indicating that increasing plant height 310
in mixtures compared to monocultures increased the selection effect (Fig. 3). We suspect that most of 311
the selection effect is due to the tall-growing and high-yielding species Chenopodium quinoa dominating 312
the communities and their yield, particularly in Switzerland (Fig. S3). The complementarity effect was 313
positively related only to ΔC:N ratio. Increases in ΔC:N ratio indicate higher productivity per unit N in 314
mixtures compared with monocultures, which increased the complementarity effect. Overall, the model 315
explained 92% of yield differences between mixtures and monocultures (Fig. 3). 316
317
318
Figure 3: Mechanistic model showing the experimental treatment effects on the effects of environmental 319
factors on plant traits and subsequently on biodiversity effects and differences in yield between crop 320
mixtures and crop monocultures. Δ indicates differences between the respective measurements in 321
mixtures compared with monocultures, thus positive Δ values indicate higher values in mixtures 322
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
12
compared with monocultures and vice versa. Mean values for Δ values per country, fertilizer treatment 323
and species number are given in table S6. Displayed black arrows show significant positive (solid) or 324
negative (dashed) relationships (α = 0.05), grey arrows indicate direct effects of treatment factors on 325
traits and yield. Arrow thickness indicates effect size based on standardized path coefficients. Numbers 326
next to the variables indicate their explained variance (R2). Double-headed grey dashed arrows indicate 327
significant correlations. Non-significant tested relationships are not shown. 328
329
Plant traits, excluding plant height, were correlated among each other. ΔLDMC was strongly 330
negatively correlated with ΔSLA and positively with ΔC:N ratio. This suggests that a higher water 331
content in leaves of crop mixtures than crop monocultures (low ΔLDMC) is correlated to a higher leaf 332
area per unit mass in crop mixtures than in crop monocultures (high ΔSLA) and to less leaf N in crop 333
monocultures than in crop mixtures (low ΔC:N ratio) (Fig. 3). 334
ΔLDMC was negatively related to both environmental factors, ΔFPAR and ΔVWC, indicating 335
that higher light interception by the crop canopy and higher soil water contents decreased LDMC, thus 336
increasing leaf water contents in mixtures. The negative correlation between ΔC:N ratio and ΔFPAR 337
implies that leaf N content in mixtures increased with increasing light interception. ΔSLA was positively 338
related to both environmental factors, suggesting that increasing light interception and soil water content 339
increased SLA, thus promoting higher leaf area per leaf mass. Based on standardized effect sizes, the 340
effect of ΔFPAR on leaf traits was stronger than the effect of ΔVWC; specifically, the effect of ΔFPAR 341
on ΔLDMC and ΔSLA was 1.4 and 1.5 fold stronger than the effect of ΔVWC, respectively. ΔVWC 342
was positively correlated to the selection effect, with the effect of Δheight being 1.2 fold stronger than 343
the effect of ΔVWC on the selection effect. Thus, higher soil water contents in mixtures compared with 344
monocultures increased the selection effect in mixtures compared with monocultures (Fig. 3). 345
ΔFPAR varied in response to fertilizer treatment, with 10% higher values in unfertilized 346
treatments, indicating that crop mixtures intercepted more light than crop monocultures and that this 347
effect was stronger in unfertilized treatments (Table S6). ΔVWC varied significantly only among 348
countries, with more negative values in Switzerland compared with Spain. This indicates that soils in 349
crop monocultures had a higher water content than in crop mixtures and that this effect was more 350
pronounced in Switzerland than in Spain. All plant traits, except ΔSLA varied significantly among 351
countries; ΔLDMC was 163% higher and Δheight 207% higher in Spain than in Switzerland, while 352
ΔC:N was 104% higher in Switzerland than in Spain (Fig. 3, Table S6). ΔLDMC also varied 353
significantly among fertilization treatments, with 170% higher results in fertilized compared with 354
unfertilized treatments. Both, selection and complementarity effects responded significantly to country. 355
Complementarity effects were 540% higher in Switzerland than in Spain and selection effects were 356
350% higher in Switzerland than in Spain. Complementarity effect was the only variable responding to 357
mixture diversification and complementarity effects were 110% higher in 4- than in 2-species mixtures. 358
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
13
The effect of mixture diversification on complementarity effects was as strong as the effect from ΔC:N 359
ratio on complementarity effects (Fig. 3, Table S6). 360
361
4. Discussion 362
Our study found increasing yields from crop monocultures to 2- to 4-species mixtures, in particular at 363
the temperate site in Switzerland. Community-level yield did not respond to fertilizer treatments, but 364
varied strongly between the two countries. Selection and complementarity effects were linked to plant 365
height and C:N ratio, respectively, and explained 92% of the variation in community-level crop yield 366
differences between mixtures and monocultures. While C:N ratio was linked to light use, plant height 367
showed no link to environmental factors and only differed between countries. The effect of mixture 368
diversification on crop yield was not mediated through any of the abiotic factors or plant traits measured 369
in this study, but acted directly upon complementarity effects. 370
371
Positive biodiversity–productivity relationships are context–dependent 372
In Switzerland community-level yield increased from monoculture to mixture and from 2- to 4-species 373
mixtures, while the diversity effects on yield in Spain remained weak. The differences between the two 374
countries were diverse and included differences in precipitation and irrigation, hours of sunshine, soil 375
nutrients and soil texture. Light availability in Spain was presumably not limiting, but could have been 376
an inhibiting factor. Lower SLA values in Spain indicate that the plants had less leaf area per dry leaf 377
mass, which could be the plants’ effort to reduce leaf area exposed to high irradiance. 378
From the three growth-limiting resources we focused on in this study, soil water and N availability 379
were the most promising to explain the missing positive diversity-productivity relationship in Spain. 380
Soil water content in Spain was kept low by reducing irrigation to the amount needed for plant survival 381
(see chapter 2.1) (mean VWC in Spain: 5.7 ± 0.4 %). Combined with a generally hotter climate in Spain, 382
the crops were more prone to water stress. Crop yields in intercropping under drought conditions are 383
expected to decrease (Coll et al. 2012). Also, positive diversity effects on crop water availability in 384
intercropping remain contested and Brooker et al. (2015) suggest that these effects are limited to 385
intercropping systems where at least one species has a low water demand. The crop species planted in 386
this experiment were not adapted to the dry conditions in Spain, since they were Swiss cultivars bred 387
for use under temperate climatic conditions. A further explanation for the absence of a positive diversity-388
productivity relationship in Spain can be the increased allocation of C to belowground productivity in 389
response to dry conditions. In our study we were interested in positive productivity effects on crop yield, 390
hence the focus on above-ground biomass. However, increased belowground investment can lead to a 391
decrease in aboveground productivity while maintaining overall community productivity (Kahmen et 392
al. 2005). 393
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
14
Besides arid conditions in Spain, low soil fertility could also have reduced biodiversity effects on 394
productivity, as observed before in an agricultural experiment (Fridley 2003). The increase in 395
community-level yields in response to legume presence in Spain indicates that N sparing by the legumes 396
could increase leaf N content for neighboring crops. Since the N sparing effect was only observed in 397
unfertilized communities in Spain, we suggest that fertilizer addition reduced the formation of nodules 398
during early legume growth, inhibiting N-fixation during later growth stages (Naudin et al. 2011). Also, 399
available soil water content is an important parameter controlling N2-fixation of legumes, either directly 400
by influencing nodulation or indirectly by reducing plant growth and thus N2-fixation (Sprent and 401
Minchin 1983). Therefore, we suggest that drier conditions in Spain might have limited plant growth 402
directly and also inhibited N2-fixation, further reducing N availability for crops. 403
404
Complementarity effects increased yields in mixtures compared with monocultures 405
In our study, complementarity effects were shown to drive positive biodiversity–productivity 406
relationships in Switzerland. The effect of complementarity was mainly linked to changes in C:N ratio 407
between mixtures and monocultures, implying that crops grown in mixtures were producing more 408
biomass per unit leaf N than crops in monoculture. This increase in efficiency could partially explain 409
the higher productivity often observed at higher diversity (Fargione et al. 2007). The C:N ratio was 410
dependent on the fraction of intercepted light but not on soil water contents. The negative link between 411
C:N ratio and FPAR indicated that higher light interception was linked to higher leaf N content per unit 412
biomass. Thus, crops responded to a more complete canopy cover and thus lower light access by 413
increasing their SLA (Fig. 3) and leaf N content to have a larger photosynthetically active leaf area. 414
High leaf N and high SLA are a common plant response to lower light conditions (Lambers et al. 1998; 415
Reich et al. 1998; Evans and Poorter 2001; Funk et al. 2017). However in Switzerland, at higher mixture 416
diversification (i.e. in 4-species mixtures) the relation between FPAR and C:N ratio became positive 417
(table S6), indicating that with increasing diversification a shift in N use efficiency occurred. Thus, 4-418
species mixtures in Switzerland produced more yield per unit N than 2-species mixtures or 419
monocultures. Increasing N efficiency in more diverse communities has been observed before in semi-420
natural (van Ruijven and Berendse 2005; Fargione et al. 2007) but, to the best of our knowledge, not in 421
intercropping systems. 422
Complementarity effects were the only variable responding to mixture diversification, with an 423
increase of complementarity effects in 4- over 2-species mixtures. Strikingly, mixture diversification 424
did not affect any of the environmental factors or plant traits measured in this study. This suggests that 425
the plant traits measured in this study were not able to describe how mixture diversification affects 426
complementarity. We suspect belowground processes driving yield increases from 2- to 4-species 427
mixtures or that dynamics of nutrients other than N, or reduced impacts of pests, were possibly playing 428
a role. Concerning possible belowground processes, other studies observed that the presence of a legume 429
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
15
can increase N or P availability in the soil surrounding its roots (Temperton et al. 2007; Zhang et al. 430
2019). In our study, this observation could potentially be important, as all 4-species mixtures contained 431
one leguminous crop, while not all 2-species mixtures did so. As an alternative to not measuring the 432
appropriate traits, it could also be that not enough traits were measured to fully capture niche differences. 433
As suggested by earlier studies, when linking plant traits to biodiversity effects, a large range of traits is 434
required to explain niche differences (Kraft et al. 2015; Cadotte 2017). 435
Although we did see facilitative processes in Spain, especially the N sparing effect of the legumes 436
in unfertilized plots, we propose that facilitative processes were not strong enough to compensate for 437
the difficult growing conditions in Spain. 438
439
Selection effects due to Chenopodium quinoa 440
In this study, selection effects had a similarly strong effect on yield differences between mixtures and 441
monocultures as had complementarity effects. While it is often assumed that positive biodiversity–442
productivity relationships are driven mainly by niche differentiation (Cardinale 2013), we show here 443
that selection effects are nearly as important. The selection effect was linked to plant height, thus 444
selection effects increased with increasing plant height in mixtures compared to monocultures. A 445
relation between plant height and selection effects has been observed before (Cadotte 2017) and is 446
probably due to plant height being related to competition for light, where taller plants outcompete shorter 447
plants (Westoby 1998). 448
Selection effects were significantly higher in Switzerland than in Spain. Combined with the link 449
to plant height, we concluded that one highly productive and tall-growing species was causing this effect 450
Chenopodium quinoa (Fig. S3). It has been observed before that C. quinoa was highly competitive 451
(Buckland 2016) and that abundant irrigation could increase its yields (Walters et al. 2016). Thus, the 452
selection effect was considerably less important in Spain than in Switzerland. 453
454
5. Conclusion 455
Our study showed that crop productivity increased with diversity under temperate conditions but only 456
weakly when crops were grown under semiarid conditions with limited availability of water and strong 457
irradiance. Increases in productivity in mixtures compared to monocultures were caused to almost the 458
same extent by complementarity and selection effects. Selection and complementarity effects were 459
explained by different plant trait syndromes. The selection effect was maximized in plots with tall plants 460
and was probably caused by one single species, Chenopodium quinoa, which was highly productive and 461
tall-growing. Complementarity effects were linked to an increased C capture per unit N, thus indicating 462
that crops in mixtures were more efficient. However, complementarity effects were also stronger in 4- 463
compared with 2-species mixtures and this link was not mediated through any of the measured plant 464
traits, suggesting that other ecological processes must have been responsible for the positive diversity 465
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
16
effect on yield beyond two-species mixtures. This finding suggests that the drivers of diversity effects 466
from monocultures to mixtures are not the same as from 2- to 4-species mixtures and should therefore 467
be targeted specifically in future studies. Nevertheless, by identifying links between relevant plant traits 468
and positive biodiversity–productivity relationships through selection and complementarity effects, this 469
study helps pave the way to select crop species for mixtures that optimize sampling and complementarity 470
effects. 471
472
Acknowledgments 473
We are grateful to Elisa Pizarro Carbonell, Carlos Barriga Cabanillas and Anja Schmutz for their help 474
with the field experiment, and Johan Six for comments on the experimental design. We also thank the 475
Aprisco de Las Corchuelas and the University of Zurich for allowing us to use their experimental 476
gardens. The study was funded by the Swiss National Science Foundation (PP00P3_170645). 477
478
Author contribution 479
N.E. and C.S. conceived the study with input from L.S. and R.W.B. N.E., L.S. and C.S. collected the 480
data; N.E. assembled and analyzed the data with the help of C.S.; N.E. and C.S. wrote the paper. All 481
authors discussed data analyses and results and revised the manuscript. 482
483
References 484
Bartoń K. 2019. MuMIn: Multi-Model Inference. R package version 1.43.15. 485 Bates D, Mächler M, Bolker B, Walker S. 2015. Fitting Linear Mixed-Effects Models using lme4. 486
Journal of Statistical Software 67: 1-48. 487 Bedoussac L, Justes E. 2009. The efficiency of a durum wheat-winter pea intercrop to improve yield 488
and wheat grain protein concentration depends on N availability during early growth. Plant 489 and Soil 330: 19-35. 490
Brooker RW, Bennett AE, Cong W-F, Daniell TJ, George TS, Hallett PD, Hawes C, Iannetta PPM, Jones 491 HG, Karley AJ et al. 2015. Improving intercropping: a synthesis of research in agronomy, plant 492 physiology and ecology. New Phytologist 206: 107-117. 493
Brooker RW, Maestre FT, Callaway RM, Lortie CL, Cavieres LA, Kunstler G, Liancourt P, Tielbörger K, 494 Travis JMJ, Anthelme F et al. 2008. Facilitation in plant communities: the past, the present, 495 and the future. Journal of Ecology 96: 18-34. 496
Buckland KR. 2016. Increasing the Sustainability of Utah Farms by Incorporating Quinoa as a Novel 497 Crop and Protecting Soil Health. in Plants, Soils and Climate, p. 145. Utah State University, 498 Utah. 499
Cadotte MW. 2017. Functional traits explain ecosystem function through opposing mechanisms. 500 Ecology Letters 20: 989-996. 501
Cardinale BJ. 2013. Towards a general theory of biodiversity for the Anthropocene. Elementa: Science 502 of the Anthropocene 1. 503
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
17
Cardinale BJ, Srivastava DS, Duffy JE, Wright JP, Downing AL, Sankaran M, Jouseau C. 2006. Effects of 504 biodiversity on the functioning of trophic groups and ecosystems. Nature 443: 989-992. 505
Coll L, Cerrudo A, Rizzalli R, Monzon JP, Andrade FH. 2012. Capture and use of water and radiation in 506 summer intercrops in the south-east Pampas of Argentina. Field Crops Research 134: 105-507 113. 508
Cooper PJM, Gregory PJ, Keatinge JDH, Brown SC. 1987. Effects of fertilizer, variety and location on 509 barley production under rainfed conditions in Northern Syria 2. Soil water dynamics and crop 510 water use. Field Crops Research 16: 67-84. 511
Cornelissen JHC, Lavorel S, Garnier E, Díaz S, Buchmann N, Gurvich DE, Reich PB, Ter Steege H, 512 Morgan HD, van der Heijden MGA et al. 2003. A handbook of protocols for standardised and 513 easy measurement of plant functional traits worldwide. Australian Journal of Botany 51: 335-514 380. 515
Duchene O, Vian J-F, Celette F. 2017. Intercropping with legume for agroecological cropping systems: 516 Complementarity and facilitation processes and the importance of soil microorganisms. A 517 review. Agriculture, Ecosystems & Environment 240: 148-161. 518
Evans JR, Poorter H. 2001. Photosynthetic acclimation of plants to growth irradiance: the relative 519 importance of specific leaf area and nitrogen partitioning in maximizing carbon gain. Plant, 520 Cell and Environment 24: 755-767. 521
Fargione J, Tilman D, Dybzinski R, Lambers JH, Clark C, Harpole WS, Knops JM, Reich PB, Loreau M. 522 2007. From selection to complementarity: shifts in the causes of biodiversity-productivity 523 relationships in a long-term biodiversity experiment. Proceedings of the Royal Society - 524 Biological Sciences 274: 871-876. 525
Fridley JD. 2003. Diversity effects on production in different light and fertility environments: an 526 experiment with communities of annual plants. Journal of Ecology 91: 396-406. 527
Funk JL, Larson JE, Ames GM, Butterfield BJ, Cavender-Bares J, Firn J, Laughlin DC, Sutton-Grier AE, 528 Williams L, Wright J. 2017. Revisiting the Holy Grail: using plant functional traits to 529 understand ecological processes. Biol Rev Camb Philos Soc 92: 1156-1173. 530
Hauggaard-Nielsen H, Ambus P, Jensen ES. 2001. Temporal and spatial distribution of roots and 531 competition for nitrogen in pea-barley intercrops - a field study employing 32P technique. 532 Plant and Soil 236: 63-74. 533
Huang Y, Chen Y, Castro-Izaguirre N, Baruffol M, Brezzi M, Lang A, Li Y, Härdtle W, von Oheimb G, 534 Yang X et al. 2018. Impacts of species richness on productivity in a large-scale subtropical 535 forest experiment. Science 362: 80-83. 536
Huston MA. 1997. Hidden treatments in ecological experiments: re-evaluating the ecosystem 537 function of biodiversity. Oecologia 110: 449-460. 538
Jensen ES. 1996. Grain yield, symbiotic N2 fixation and interspecific competition for inorganic N in 539 pea-barley intercrops. Plant Soil 182: 25-38. 540
Jesch A, Barry KE, Ravenek JM, Bachmann D, Strecker T, Weigelt A, Buchmann N, de Kroon H, Gessler 541 A, Mommer L et al. 2018. Below-ground resource partitioning alone cannot explain the 542 biodiversity-ecosystem function relationship: A field test using multiple tracers. Journal of 543 Ecology 106: 1-17. 544
Kahmen A, Perner J, Buchmann N. 2005. Diversity-dependent productivity in semi-natural grasslands 545 following climate perturbations. Functional Ecology 19: 594-601. 546
Kottek M, Grieser J, Beck C, Rudolf B, Rubel F. 2006. World Map of the Köppen-Geiger climate 547 classification updated. Meteorologische Zeitschrift 15: 259-263. 548
Kraft NJ, Godoy O, Levine JM. 2015. Plant functional traits and the multidimensional nature of 549 species coexistence. Proceedings of the National Academy of Sciences 112: 797-802. 550
Kuznetsova A, Brockhoff PB, Christensen RHB. 2017. lmerTest Package: Tests in Linear Mixed Effects 551 Models. Journal of Statistical Software 82: 1-26. 552
Lambers H, Chapin FS, Pons TL. 1998. Plant Physiological Ecology. Springer, New York. 553
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
18
Li C, Hoffland E, Kuyper TW, Yu Y, Li H, Zhang C, Zhang F, van der Werf W. 2020. Yield gain, 554 complementarity and competitive dominance in intercropping in China: A meta-analysis of 555 drivers of yield gain using additive partitioning. European Journal of Agronomy 113. 556
Li L, Tilman D, Lambers H, Zhang F-S. 2014. Plant diversity and overyielding: insights from 557 belowground facilitation of intercropping in agriculture. New Phytologist 203: 63-69. 558
Loreau M, Hector A. 2001. Partitioning selection and complementarity in biodiversity experiments. 559 Nature 412: 72-76. 560
Miyazawa K, Murakami T, Takeda M, Murayama T. 2009. Intercropping green manure crops—effects 561 on rooting patterns. Plant and Soil 331: 231-239. 562
Mouquet N, Lagadeuc Y, Devictor V, Doyen L, Duputié A, Eveillard D, Faure D, Garnier E, Gimenez O, 563 Huneman P et al. 2015. Review: Predictive ecology in a changing world. Journal of Applied 564 Ecology 52: 1293-1310. 565
Naudin C, Corre-Hellou G, Voisin A-S, Oury V, Salon C, Crozat Y, Jeuffroy M-H. 2011. Inhibition and 566 recovery of symbiotic N2 fixation by peas (Pisum sativum L.) in response to short-term 567 nitrate exposure. Plant and Soil 346: 275-287. 568
Pérez-Harguindeguy N, Díaz S, Garnier E, Lavorel S, Poorter H, Jaureguiberry P, Bret-Harte MS, 569 Cornwell WK, Craine JM, Gurvich DE et al. 2013. New handbook for standardised 570 measurement of plant functional traits worldwide. Australian Journal of Botany 61. 571
R Core Team. 2019. R: A Language and Environment for Statistical computing. R Foundation for 572 Statistical Computing, Vienna, Austria. 573
Rasband WS. 1997-2018. ImageJ. U. S. National Institutes of Health, Bethesda, Maryland, USA. 574 Reich PB, Ellsworth DS, Walters MB. 1998. Leaf structure (specific leaf area) modulates 575
photosynthesis-nitrogen relations: evidence from within and across species and functional 576 groups. Functional Ecology 12: 948-958. 577
Rosseel Y. 2012. lavaan: An R Package for Structural Equation Modeling. Journal of Statistical 578 Software 48: 1-36. 579
Sekiya N, Yano K. 2004. Do pigeon pea and sesbania supply groundwater to intercropped maize 580 through hydraulic lift?—Hydrogen stable isotope investigation of xylem waters. Field Crops 581 Research 86: 167-173. 582
Spehn EM, Scherer-Lorenzen M, Schmid B, Hector A, Caldeira MC, Dimitrakopoulos PG, Finn JA, 583 Jumpponen A, O'Donovan G, Pereira JS et al. 2002. The role of legumes as a component of 584 biodiversity in a cross-European study of grassland biomass nitrogen. Oikos 98: 205-218. 585
Sprent JL, Minchin FR. 1983. Environmental effects on the physiology of nodulation and nitrogen 586 fixation. in Temperate legumes, physiology, genetics and nodulation (eds. DG Jones, DR 587 Davies), pp. 269-317. Pitman Books, London, UK. 588
Stomph TJ, Dordas C, Baranger A, de Rijk J, Dong B, Evers J, Gu C, Li L, Simon J, Jensen ES et al. 2020. 589 Designing intercrops for high yield, yield stability and efficient use of resources: Are there 590 principles? Advances in Agronomy 160: 1-41. 591
Sun S-J, Meng P, Zhang J-S, Wan X. 2013. Hydraulic lift by Juglans regia relates to nutrient status in 592 the intercropped shallow-root crop plant. Plant and Soil 374: 629-641. 593
Temperton VM, Mwangi PN, Scherer-Lorenzen M, Schmid B, Buchmann N. 2007. Positive interactions 594 between nitrogen-fixing legumes and four different neighbouring species in a biodiversity 595 experiment. Oecologia 151: 190-205. 596
Tilman D, Lehman CL, Thomson KT. 1997. Plant diversity and ecosystem productivity: theoretical 597 considerations. Proceedings of the National Academy of Sciences 94: 1857-1861. 598
Tilman D, Reich PB, Knops J, Wedin DA, Mielke T. 2001. Diversity and Productivity in a Long-Term 599 Grassland Experiment. Science 294. 600
van Ruijven J, Berendse F. 2005. Diversity-productivity relationships: Initial effects, long-term 601 patterns, and underlying mechanisms. Proceedings of the National Academy of Sciences 102: 602 695-700. 603
Vandermeer J. 1989. The ecology of intercropping. Cambridge University Press, Cambridge. 604
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint
19
Von Felten S, Niklaus PA, Scherer-Lorenzen M, Hector A, Buchmann N. 2012. Do grassland plant 605 communities profit from N partitioning by soil depth? Ecology 93: 2386-2396. 606
Walters H, Carpenter-Boggs L, Desta K, Yan L, Matanguihan J, Murphy K. 2016. Effect of irrigation, 607 intercrop, and cultivar on agronomic and nutritional characteristics of quinoa. Agroecology 608 and Sustainable Food Systems 40: 783-803. 609
Weiher E, Van der Werf W, Thompson K, Roderick M, Garnier E, Eriksson O. 1999. Challenging 610 Theophrastus: A common core list of plant traits for functional ecology. Journal of Vegetation 611 Science 10: 609-620. 612
Weisser WW, Roscher C, Meyer ST, Ebeling A, Luo G, Allan E, Beßler H, Barnard RL, Buchmann N, 613 Buscot F et al. 2017. Biodiversity effects on ecosystem functioning in a 15-year grassland 614 experiment: Patterns, mechanisms, and open questions. Basic and Applied Ecology 23: 1-73. 615
Westoby M. 1998. A leaf-height-seed (LHS) plant ecology strategy scheme. Plant and Soil 199: 213-616 227. 617
Wright AJ, Wardle DA, Callaway R, Gaxiola A. 2017. The Overlooked Role of Facilitation in Biodiversity 618 Experiments. Trends in Ecology & Evolution 32: 383-390. 619
Zhang D, Lyu Y, Li H, Tang X, Hu R, Rengel Z, Zhang F, Whalley WR, Davies WJ, Cahill JF, Jr. et al. 2019. 620 Neighbouring plants modify maize root foraging for phosphorus: coupling nutrients and 621 neighbours for improved nutrient-use efficiency. New Phytologist. 622
623
.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted June 18, 2020. ; https://doi.org/10.1101/2020.06.17.157008doi: bioRxiv preprint