Selection during crop diversification involves correlated evolutionof the circadian clock and ecophysiological traits in Brassica rapa

Yulia Yarkhunova1, Christine E. Edwards2, Brent E. Ewers1, Robert L. Baker1, Timothy Llewellyn Aston1,

C. Robertson McClung3, Ping Lou3 and Cynthia Weinig1,4

1Department of Botany and Program in Ecology, University of Wyoming, Laramie, WY 82071, USA; 2Center for Conservation and Sustainable Development, Missouri Botanical Garden,

St Louis, MO 63166, USA; 3Department of Biological Sciences, Dartmouth College, Hanover, NH 03755, USA; 4Department of Molecular Biology, University of Wyoming, Laramie, WY

82071, USA

Author for correspondence:Cynthia Weinig

Tel: +1 307 7666378Email: cweinig@uwyo.edu

Received: 26 June 2015

Accepted: 14 October 2015

New Phytologist (2015)doi: 10.1111/nph.13758

Key words: Brassica rapa, circadian rhythms,crop diversification, ecophysiological traits,Farquhar model, leaf anatomy.

Summary

� Crop selection often leads to dramatic morphological diversification, in which allocation to

the harvestable component increases. Shifts in allocation are predicted to impact (as well as

rely on) physiological traits; yet, little is known about the evolution of gas exchange and

related anatomical features during crop diversification.� In Brassica rapa, we tested for physiological differentiation among three crop morphotypes

(leaf, turnip, and oilseed) and for correlated evolution of circadian, gas exchange, and pheno-

logical traits. We also examined internal and surficial leaf anatomical features and biochemical

limits to photosynthesis.� Crop types differed in gas exchange; oilseed varieties had higher net carbon assimilation

and stomatal conductance relative to vegetable types. Phylogenetically independent contrasts

indicated correlated evolution between circadian traits and both gas exchange and biomass

accumulation; shifts to shorter circadian period (closer to 24 h) between phylogenetic nodes

are associated with higher stomatal conductance, lower photosynthetic rate (when CO2 sup-

ply is factored out), and lower biomass accumulation. Crop type differences in gas exchange

are also associated with stomatal density, epidermal thickness, numbers of palisade layers,

and biochemical limits to photosynthesis.� Brassica crop diversification involves correlated evolution of circadian and physiological

traits, which is potentially relevant to understanding mechanistic targets for crop improve-

ment.

Introduction

In his early study, ‘The variation of plants and animals underdomestication’, Darwin recognized that studying domesticatedplants provided insights into fundamental evolutionary processes(Darwin, 1868). In the process of domestication, a crop speciesbecomes evolutionarily differentiated from a wild progenitor ashumans transition from harvesting and stewardship of plants inwild populations to imposing selection on plants in an agricul-tural ecosystem. This is commonly distinguished from cropdiversification, which occurs after domestication as humans selectfor secondary traits such as crop growth in diverse geographicregions, synchronized timing of life-history events, or enhancedflavor (Meyer & Purugganan, 2013; Abbo et al., 2014). Amongthe most visually apparent phenotypic changes during crop diver-sification, dramatic morphological differences have emergedamong related vegetable crops (Tsunoda et al., 1980; Puruggananet al., 2000) and between related vegetable and oilseed varieties(De Vries, 1997) as a consequence of selection to increase theharvestable component.

Selection during crop diversification may lead to correlatedevolution of multiple traits. Evolution of some traits (e.g. mor-phological ones) may be functionally dependent on other traits(e.g. physiological ones). During domestication or diversification,improvements in size or in relative allocation to the harvestablecomponent are probably enabled by physiological evolution. Forinstance, sink strength appears to be greater in oilseed vs veg-etable crops within a species, because oil per unit mass is moremetabolically expensive to produce than vegetative organs; car-bon assimilation is correspondingly higher in oilseed varieties(Evans, 1996; Ruuska et al., 2004; Goffman et al., 2005; Li et al.,2006). A small number of recent metabolomic studies also sug-gest correlated morphophysiological evolution (Del Carpio et al.,2014; Hennig et al., 2014). Nevertheless, despite the clear mech-anistic links in contemporary populations and across species(Wright et al., 2004; Sack et al., 2013) and the potential relevanceto crop improvement, examples of joint morphophysiologicalevolution during crop domestication or diversification are com-paratively rare (Cook & Evans, 1983; Smartt, 1988). Simultane-ously quantifying multiple morphological, physiological, and

� 2015 The Authors

New Phytologist� 2015 New Phytologist Trust

New Phytologist (2015) 1www.newphytologist.com

Research

developmental traits in domesticated crops that have undergonesubsequent diversification should offer insights as to suites ofmorphophysiological traits that evolve jointly to enhance yield(Evans, 1996).

Many developmental features could contribute to physiologi-cal diversification among crop types. One well-known regulatorycontrol on physiological traits is the circadian clock. Circadianrhythms are biological oscillations that are c. 24 h in durationand coordinate the diurnal timing of many physiological andmetabolic processes in plants. The circadian clock regulates eco-physiological traits (e.g. starch utilization and photosynthesis) inexperimental genetic lines such as mutant genotypes that expressdiscrete clock phenotypes that differ from the cognate wild-type(Dodd et al., 2005; Graf et al., 2010; Faure et al., 2012). Quanti-tative clock variation is also associated with net carbon assimila-tion and stomatal conductance in segregating progenies ofBrassica rapa (Edwards et al., 2009, 2011; Lou et al., 2011),potentially because clock cycles closer to 24 h enhance diurnaltiming of gas exchange. During crop diversification, the circadianclock and gas exchange traits may evolve in a correlated mannerif gas exchange responds to allele frequency changes at clock locior if photosynthetic evolution alters sugar status and therebyaffects clock function (Dalchau et al., 2011; Moghaddam & Vanden Ende, 2013). More proximately (and quite possibly indepen-dent of the circadian clock), leaf anatomical features or photosyn-thetic biochemistry could contribute to crop variation in gasexchange (Kebede et al., 1994).

Crucifers provide some of the most dramatic examples ofcrop morphological diversification. Reproductive organs arehighly altered in some members of Brassica oleracea, such ascauliflower and broccoli (Purugganan & Fuller, 2009). Vege-tative features are also highly modified among varieties ofboth B. oleracea and B. rapa. Distinct B. rapa accessions haveundergone selection for increased allocation to above-groundorgans (i.e. leaves in cabbage, pak choi, and turnip greencrops and seeds in oilseed types) as well as below-groundorgans (i.e. enlarged root-like hypocotyl in turnips). Accompa-nying morphological evolution, we hypothesize that crop typeswill exhibit physiological diversification as a consequence ofdifferences in overall size, sink strength, and proportionateallocation; the strength of morphophysiological correlationsremains as an empirical gap in our understanding of evolu-tion during domestication and diversification, and may informtargets for future crop improvement. Further, past studiesprovide evidence of genotypic variation in circadian clockperiod (Lou et al., 2011), but it remains unclear if circadiandifferences generalize to crop morphotype or if circadian phe-notypes correlate with physiological traits over evolutionarytime, which may again inform breeding for crop improve-ment.

Our overall aim was to test if the morphological variation aris-ing from selection for different crop types is associated with dif-ferences in circadian, leaf gas exchange, and anatomical traits indifferent B. rapa accessions. With this study we accomplished thefollowing specific aims: we first tested whether genotypic differ-ences in circadian period generalize to crop type; we then

explored if crop types differ in physiological (gas exchange) traitsand, if so, whether gas exchange covaries phylogenetically withcircadian period. Further, we tested the related gas exchangehypothesis in the crops: the rate of photosynthetic carbonassimilation (A) is limited by gas supply of CO2 (gs) and bio-chemical demand for CO2; demand is, in turn, regulated bylight-dependent and light-independent processes occurring in themesophyll, which are either limited by light harvesting (ribulose-1,5-bisphosphate (RuBP) regeneration) or by sugar metabolismthrough CO2 assimilation (Vcmax, maximum carboxylationcapacity of Rubisco) and triose phosphate utilization. Finally, weexamine leaf anatomical features (stomatal density and internalleaf features) that may contribute to physiological diversification.We explore how stomatal density and thickness of epidermismight contribute to CO2 supply and resistance differences, andhow palisade structure could affect light absorption, includingelectron transport and carbon fixation.

Materials and Methods

Study system

Brassica rapa L. (Brassicaceae) is an internationally importantcrop, and distinct subspecies are cultivated as oilseed crops (ssp.oleifera), leaf vegetable crops (ssp. chinenesis and pekinensis), androot vegetable crops (subsp. rapa). Accessions used here reflectthe global cultivation of the species; accession origin andgermplasm identifiers are presented in Supporting InformationTable S1.

Circadian measures

In all, 41 accessions were screened for circadian period(Table S1). Some accessions were screened previously by Louet al. (2011). Here, previously published data, data for additionalreplicates of accessions within the original panel, and data onadditional accessions are analyzed to test for circadian differencesamong accessions that generalize to crop type; within each acces-sion, between eight and 15 replicates were screened. For detailsof the circadian screens, which used leaf movement as a proxy toestimate circadian period, see Lou et al. (2011) as well as reviewsof this method (McClung, 2006; Greenham et al., 2015).

Leaf gas exchange measurements

In addition to differentiation of the circadian clock, we wereinterested in whether crop types differ in gas exchange, andwhether there is evidence of correlated evolution between circa-dian and gas exchange traits. A subset of crop accessions(Table S2) was sampled for physiological and leaf anatomicaltraits. For gas exchange experiments, nine to 10 replicates of 21representative genotypes were grown from seed in growth cham-bers (PGC-9/2 with Percival Advanced Intellus EnvironmentalController, Percival Scientific, Perry, IN, USA). Seeds wereplanted into 69 69 9 cm plastic pots filled with Redi-Earth pot-ting mix (Sunshine Redi-Earth Professional Growing Mix, Sun

New Phytologist (2015) � 2015 The Authors

New Phytologist� 2015 New Phytologist Trustwww.newphytologist.com

Research

NewPhytologist2

Gro Horticulture, Bellevue, WA, USA) and including a slow-release fertilizer (Scotts brand Osmacote Controlled Release Clas-sic, NPK; Scotts, Marysville, OH, USA). Three seeds wereplanted in each pot, and then covered with vermiculite and satu-rated with water. Plants were placed in the growth chamber withfollowing conditions: photoperiod of 15 : 9 h, 24 : 22°C,light : dark cycle, and photosynthetic photon flux density =500 lmol photons m�2 s�1. Plants were thinned to one seedlingper pot after cotyledons had expanded (c. after 5 d). At subjectivedawn, plants were watered daily to retain consistently high soilmoisture.

Gas exchange measurements were taken using a Li-Cor LI-6400 XT portable infrared gas analyzer with leaf chamber fluo-rometer (Open System Vers. 4.0, Li-Cor Inc., Lincoln, NE,USA). Measured traits were maximum photosynthetic rate (Amax)and stomata conductance (gs). Measurements were taken fromthe third fully developed true leaf at least 1 h after subjectivedawn under the following chamber conditions: photosyntheticphoton flux density (PPFD) = 1500 lmol m�2 s�1, ref [CO2] =400 lmol m�2 s�1, Tleaf = 24°C, and vapor pressure deficit basedon leaf temperature (VPDL, kPa) was kept between 1.3 and1.7 kPa (Long & Bernacchi, 2003). Measurements were taken3–5 min after all chamber conditions stabilized.

A–Ci response curves

A–Ci response curves were used to estimate biochemical limitsto photosynthesis, including the maximum rate of RuBP car-boxylation (Vcmax) and the maximum rate of electron transportdriving RuBP regeneration (Jmax). In further regard to Jmax,RuBP regeneration is controlled by the amount and activity ofCalvin cycle enzymes as well as NADP and ATP pool sizes,while NADP and ATP are supplied by the light-dependentreactions and are thus limited by electron transport betweenthe two photosystems.

Response curves were performed on the same leaf as the singletime point measurement again using six Li-Cor 6400 XTportable photosynthesis systems within a 5 d period for thegrowth chamber experiment. Because of the time-consumingnature of A–Ci phenotyping (in which measurement of one repli-cate can take 1–2 h and measurements can only be taken duringan approx. 2 h time window daily) and the fact that all replicatesshould be measured at a similar, preflowering stage, A–Ci curveswere collected only on three to four replicates within each of fourgenotypes for each of the three crop types (n = 44). Chamberconditions were maintained at PPFD = 1500 lmol m�2 s�1,Tleaf = 25°C and VPDL between 1.3 and 1.7 kPa. ReferenceCO2 concentration in the chamber was controlled with a CO2

mixer across the series of 400, 500, 600, 800, 1000, 1250, 1500,2000, 400, 300, 200, 100 and 50 lmol m�2 s�1 and mea-surements were recorded after equilibration to a steady state(5–7 min). The utility tool A/Ci Response Curve Fitting 10.0available at http://landflux.org/Tools.php, which is based on theequations of Farquhar et al. (1980) and an approach developedby Ethier & Livingston (2004), was used to estimate Vcmax andJmax for each A–Ci curve.

Leaf morphological measures

Stomatal density was measured in a subset of plants on which gasexchange data were collected. The fourth leaf was chosen forstomatal measurements; this leaf was fully expanded and wasdevelopmentally close to the third leaf, which was used for gasexchange measurements. Each leaf was first preserved in forma-lin-aceto-alcohol (FAA) solution (ratios of ethanol, ethanoic acid,and formalin were, by volume, 18 : 1 : 1), following severalethanol washes over the course of 1 wk until the leaf tissuebecame transparent. An aqueous solution of toluidine blue O(0.05%) was used to stain the tissue for 30–60 s depending onthe thickness of the leaves. Light micrographs were taken usingan Olympus BH2 compound light microscope at a magnificationof 9500. Within each leaf sample, stomata were counted visuallyin five to six areas chosen randomly. Five to six squares of dimen-sions 2009 200 lM were used to count stomata visually. Therange number of stomata per square was 10–35. The softwareprogram, Image J (Schneider et al., 2012), was used to calculatestomatal density, that is, the number of stomata counted withinan image area. In total, measurements were taken from five to siximages within each of six to eight replicates in each of 15 geno-types (equally representing the three crop types) (n = 639).

Leaf anatomical measures

Tissue samples were fixed in FAA and stored in 70% ethanol.Subsequently, leaves were dehydrated in a standard ethanol gradi-ent. Ethanol was replaced with Histoclear, and the tissue wasslowly infiltrated with paraffin at 60°C (Fowke & Rennie, 1995).Samples were embedded in paraffin, serially sectioned at 10 lM,and stained with toluidine blue O (Sakai, 1973). Light micro-graphs were captured using a Zeiss Primovert compound lightmicroscope and Zeiss camera and AxioVision software(AxioVision Rel 4.9.1; Carl Zeiss Imaging Solutions). The transi-tion from spongy mesophyll to palisade parenchyma was delin-eated by drawing a straight line through the center of multiplevascular bundles. All measurements were collected from a1000 lM long area of tissue in Photoshop CS4. Internal leafanatomy was likewise measured in a subset of samples includingtwo to five replicates in each of nine genotypes (equally represent-ing the three crop types) (n = 24), owing to the extensive timeneeded to clear, fix, embed, section, and analyze histological fea-tures.

Plant harvest, leaf area, nitrogen content

After 5 wk of growth, plants were carefully removed from pots,and roots and shoots were washed to remove soil. One leaf perplant was scanned using the Epson Perfection V500 software at aresolution of 150 dpi (Epson, Nagano, Japan). The software pro-gram IMAGEJ was used to calculate leaf area (Schneider et al.,2012). Plants were cut at soil level, and the single collected leaf,the remaining above-ground portion, and the below-ground por-tion were dried at 65°C for 72 h. DW was then measured. Theoven-dried leaves were ground and analyzed for d15N, d13C, leaf

� 2015 The Authors

New Phytologist� 2015 New Phytologist TrustNew Phytologist (2015)

www.newphytologist.com

NewPhytologist Research 3

nitrogen concentration (wt% N), leaf carbon concentration (wt%C) and C : N ratio using an elemental analyzer (ECS 4010,Costech Analytical Technologies Inc., Valencia, CA, USA) cou-pled to a continuous-flow inlet isotope ratio mass spectrometer(CF-IRMS; Delta-plus XP, Thermo Scientific, Waltham, MA,USA). N data were collected on seven to eight replicates withineach of 12 genotypes equally representing the three crop types.

Data analysis

Statistical approach and data treatments We used ANOVAwith post hoc Tukey tests to identify significant crop type andgenotype within crop type effects (PROC GLM, SAS v.9.4., SASInstitute Inc., Cary, NC, USA). We observed a small number ofoutlier genotypes within crop types for circadian period. Notably,our assignment of ‘crop type’ and history of selection was basedon germplasm annotation, but our observation of morphologyand phenology suggested that some genotypic outliers might havebeen better assigned to a different crop type. For instance, in thegroup of oilseeds, the genotype WO-024 had an enlarged below-ground storage structure by midseason reminiscent of a turnip,and had a period phenotype that resembled other vegetable ratherthan oilseed types (Fig. S1). Likewise, several turnips had reducedvernalization requirements similar to oilseeds, and showed periodlengths closer to the average oilseed genotype. We thereforetested for genotype (crop type) and crop type effects either usingthe germplasm designation to assign genotypes to a crop type orusing phenotypic similarity to assign genotypes to crop type.Results were not affected by these differences in assignment, andwe therefore present the more conservative results using thegermplasm designation (Table S1).

We used linear regression to test for associations betweenquantitative clock variation and gas exchange traits, biomassaccumulation, and flowering time. We used both unadjusted val-ues of gas exchange in these analyses as well as Amax valuesadjusted for CO2 supply limitation and gs values adjusted for bio-chemical feedbacks on stomatal behavior. More specifically, weperformed the genotypic regression of Amax = gs to generate resid-uals and the variable rAmax; rAmax thus reflects variation in carbonassimilation after CO2 supply limitation is considered. The vari-able rgs reflects variation in stomatal behavior after biochemicaldemand is considered.

Phylogenetic analysis and phylogenetically independent con-trasts To reconstruct a phylogeny, we used amplified fragmentlength polymorphism data from a previously published study(Zhao et al., 2005), which consisted of 527 aligned charactersfrom 21 accessions representing three crop types (seven oilseeds,six leaf crops, and eight turnips; Table S1). Phylogeny recon-struction was carried out using a parsimony optimality criterionin PAUP* 4.0b10 (Swofford, 2002). In all parsimony analyses,heuristic searches were conducted using 1000 random additionreplicates and tree bisection and reconnection (TBR) branchswapping, saving 100 trees per replicate. All most parsimonioustrees were stored, saving branch lengths, which were used forphylogenetically independent contrasts (PICs). Bootstrap

analyses (1000 replicates) (Felsenstein, 1985) were used to assessbranch support employing a heuristic search with TBR branchswapping, with 10 random additions per replicate, saving nomore than 100 trees per replicate.

To calculate correlated changes in traits using a phylogeneticframework, we calculated PICs based on the most parsimonioustrees resulting from phylogeny reconstruction. The PIC approachis a comparative one that assumes a Brownian motion model ofevolution to test for evolutionary correlations of traits afterexplicitly accounting for phylogenetic relatedness. All PICs werecarried out using the PDAP module in MESQUITE (Midford et al.,2005). The two most parsimonious trees with stored branchlengths (see the Results section) were used to compute contrasts.Before carrying out analyses, we checked diagnostic plots of theabsolute values of the standardized PICs vs their standard devia-tions using the most parsimonious trees to determine whether thebranch lengths of the phylogenetic tree adequately fitted the tipdata (Garland et al., 1992; Diaz-Uriarte & Garland, 1996,1998); because branch lengths were adequately standardized forall characters on both trees, these analyses are not discussed fur-ther. We used PICs to carry out two different types of analyses.First, we calculated an independent contrast between circadianperiod and each gas exchange and allocation trait as well as flow-ering time, all of which are continuous traits, to understand pat-terns of correlated evolution between these traits. Second, weused PICs to assess whether shifts in crop type (which is a cate-gorical trait) were associated with significant shifts in gasexchange and allocation traits following Whittall & Hodges(2007). We calculated contrasts and then used a paired t-test tocalculate whether contrasts resulting from a shift in crop typewere significantly different from contrasts between nodes of thesame crop type. We also repeated these analyses with all most-parsimonious trees resulting from phylogeny reconstruction andusing trees that varied in rooting to assess how tree topology androoting affected the results of the PICs. Because all results wereconsistent regardless of variation in topology and rooting, theseanalyses are not discussed further.

Results

Using ANOVA, we found significant effects of genotype (croptype) on circadian period (Tables S1, S3a). We further found asignificant effect of crop type on circadian period duration(Fig. 1; Table S3a). In particular, we observed that oilseeds havecircadian cycles c. 1 h shorter on average and closer to 24 h thanthose of leaf crops or turnips (Fig. 1).

Analysis of gas exchange data showed that crop types also differsignificantly in both maximum net photosynthesis rate (Amax)and stomatal conductance (gs), such that oilseeds have higher val-ues of both traits relative to leaf crops and turnips (Fig. 2a,b;Table S3b,c). A strong positive association was observed betweenAmax and gs (Fig. 3a). No significant correlations were observedbetween unadjusted gas exchange values and circadian periodlengths (data not shown). However, rAmax (which accounts forgenotypic variation in stomatal conductance and CO2 supply)and circadian period were positively correlated (Fig. 3b). rgs

New Phytologist (2015) � 2015 The Authors

New Phytologist� 2015 New Phytologist Trustwww.newphytologist.com

Research

NewPhytologist4

(which accounts for biochemical feedbacks on stomatal behavior)and circadian period were negatively correlated (Fig. 3c). Thus,circadian period is related to leaf gas exchange once gas supply ordemand is statistically considered. Biomass was positively associ-ated with circadian period (Fig. 3d), suggesting that the highrAmax of the leaf crops and turnips (genotypes with long circadianperiod; Fig. 3b) is important to determining mass or simply thattheir preferential allocation of fixed carbon to vegetative organsrather than to metabolically expensive seed oil affects totalbiomass accumulation. No association was observed between cir-cadian period and the phenological trait, flowering time (datanot shown).

The significant association between circadian period and bothgas exchange parameters and biomass (Fig. 3) can be explainedeither by inheritance of a combination of traits from a commonancestor (e.g. an oilseed ancestor that had both high gas exchangevalues and short 24 h circadian cycles) or by selection for corre-lated evolution of these two traits. Using PICs to assess correlatedevolution while taking phylogenetic relationships into account,we observed significant correlations between circadian period andrAmax (r = 0.53, P = 0.02) and total biomass (r = 0.47, P = 0.05)and a marginally significant correlation between period and rgs(r =�0.45, P = 0.06), as illustrated in Fig. 3. These results sug-gest that observed correlations are significant not because phylo-genetic groups have a common inherited relationship betweenthese traits, but because evolution in circadian rhythms is accom-panied by a corresponding response in the other traits regardlessof phylogenetic relationships. In the second analysis, we usedPICs to assess whether shifts in crop type (which is a categoricaltrait) between phylogenetic nodes were associated with shifts ingas exchange and allocation traits, but we did not find any signifi-cant association between crop diversification and evolution ofthese traits. This may reflect an absence of trait evolutionattributable to independent crop diversification events, or simply

that there is a limited number of discrete transitions betweennodes (crops) on the phylogeny and therefore low power to detecttraits associated with these transitions.

We were interested to further explore the source of genotypicdifferences in gas exchange parameters, Amax and gs. In order toexplain limitations on photosynthetic rates, we measured theresponse of photosynthesis to changes in intercellular mole frac-tions of CO2 (A–Ci curves) in a subset of genotypes using theFarquhar–von Caemmerer–Berry biochemical model of photo-synthesis. Fig. 4(a) shows the A–Ci curve in three crop types ofB. rapa and two parameters important to photosynthetic perfor-mance, including the initial slope (which denotes the carboxyla-tion-limited region where CO2 supply limits the activity ofRubisco while RuBP is not limiting) as well as the plateau (whichis the RuBP-limited region where CO2 no longer limits the car-boxylation reaction, but availability of RuBP limits the activity ofRubisco). Carboxylation-limited and RuBP-limited regionsappear qualitatively similar in turnips and leaf crops, while

Fig. 1 Circadian period length among different crop types estimated bycotyledon movement. Error bars indicate� 1SE. Different letters indicatestatistically significant differences among Brassica rapa crop types(P < 0.05).

(a)

(b)

Fig. 2 Differences in gas exchange traits among crop types. (a, b) Stomatalconductance (a) and maximum rate of CO2 assimilation (b) in Brassicarapa crop accessions measured at 1500 lmol quantam�2 s�1 and a leaftemperature of 24°C. Error bars indicate� 1SE. Different letters indicatestatistically significant differences among crop types (P < 0.05).

� 2015 The Authors

New Phytologist� 2015 New Phytologist TrustNew Phytologist (2015)

www.newphytologist.com

NewPhytologist Research 5

oilseed replicates appear to differ from the vegetable crop geno-types in both features. Analysis of the curves indicates that Vcmax,the maximum Rubisco carboxylation rate, is in fact significantlyhigher in oilseeds relative to leaf crops and tends to be higherin oilseeds relative to turnips (Fig. 4b; Table S3d). The valuesof Jmax, the maximum electron transport rate driving RuBPregeneration, and Vcmax are significantly correlated (R2 = 0.55,P < 0.001). Most of the points fall close to a straight line, and theslope of the fitted regression line equals 1.33, indicating that anincrease in intracellular CO2 more favorably affects RuBP regen-eration than carboxylation rate. Notably, the ratio of Jmax : Vcmax

is significantly higher in leaf crops than oilseed types or turnips(Fig. 4c; Table S3e). The high Jmax : Vcmax ratio indicates moreallocation to light harvesting (increased electron transport andRuBP regeneration) in leaf crops, with commensurately highercommitment to photosynthetic carbon metabolism (carboxyla-tion) in oilseeds. A strong negative association is observedbetween leaf area and percentage of nitrogen (Fig. 4d), possiblyindicating higher Rubisco concentrations in the smaller oilseedleaves.

Differences in CO2 supply could provide an additional expla-nation for higher carbon assimilation in oilseeds, that is, in addi-tion to more efficient photosynthetic biochemistry (as indicatedby Vcmax and Jmax earlier), oilseeds may have higher CO2 supplyas a result of leaf morphological features. The oilseed genotypeswe sampled in fact have 40% more stomata than vegetable geno-types (Fig. 5a–c; Table S3f), and stomatal density is positivelycorrelated with stomatal conductance, gs (r = 0.48, P < 0.001).

Further, internal leaf anatomical features, such as a thinneradaxial epidermis (Fig. 6a; Table S3 g), probably improve gassupply in oilseeds. Oilseeds also tend to have thicker mesophylland more palisade layers (Fig. 6b,c; Table S3 h). Thicker palisademesophyll is one mechanism to achieve a high Amax, becauseadditional palisade cells provide more space for chloroplasts, light

harvesting, and photosynthethic biochemistry. No crop type dif-ferences in abaxial epidermis were observed.

Discussion

In the present study, we examined hypotheses related to corre-lated morphophysiological evolution. To test for correlated traitevolution, we measured circadian rhythms, gas exchange, photo-synthetic biochemistry proxies, and leaf morphological andanatomical features among three morphologically diverse croptypes of B. rapa with past selection for different harvestable com-ponents (seed oil, tubers, and leaves). Differences in crop mor-phology that relate to size, sink strength, and/or allocation wereassociated with carbon fixation and stomatal conductance andwith leaf histological features that affect these gas exchange traits,suggesting physiological evolution during crop diversificationand thus filling an important empirical knowledge gap. As anovel component to crop diversification, the circadian clock alsoexhibited correlated evolution with gas exchange traits. This cor-related evolution may reflect a response to selection during diver-sification, in which crop varieties (or possibly just individualgenotypes) with greater sink strength, such as oilseeds, areselected for higher carbon fixation and a circadian period closerto 24 h enables higher photosynthesis. We further identified leaf-level biochemical features that contribute to genotypic differencesin gas exchange. Later, we discuss the results in the light of previ-ous studies examining phenotypic outputs of the clock and diver-sification of other crop species.

Circadian clock associations with physiological processes incrop morphotypes

Translational research has examined clock function and theArabidopsis thaliana clock model in crop plants. In several cases,

(a) (b)

(c) (d)

Fig. 3 Association among circadian period,gas exchange traits, and biomass. Eachsymbol represents a genotype; crop types arecoded by color. (a) Association betweenmaximum net photosynthesis rate (Amax) andstomatal conductance (gs) for 21 genotypesof three Brassica rapa crop types. The linerepresents the following relationship: r = 0.79(R2 = 0.61), P < 0.0001. (b) Associationbetween maximum net photosynthesis rate(Amax) and circadian period after statisticallyfactoring out stomatal conductance (gs). Theline represents the following relationship:r = 0.63 (R2 = 0.36), P = 0.0038. (c)Association between stomatal conductance(gs) and circadian period after statisticallyfactoring out Amax. The line represents thefollowing relationship: r =�0.58 (R2 = 0.29),P = 0.0091. (d) Association betweencircadian period and total biomassaccumulation. The line represents thefollowing relationship: r = 0.56 (R2 = 0.27),P = 0.0117.

New Phytologist (2015) � 2015 The Authors

New Phytologist� 2015 New Phytologist Trustwww.newphytologist.com

Research

NewPhytologist6

orthologs of A. thaliana circadian genes with conserved functionare found in crop species, including Oryza (Murakami et al.,2003), Poplar (Takata et al., 2009), and B. rapa (Murakami et al.,2003; Takata et al., 2009; Kim et al., 2012; Lou et al., 2012; Xieet al., 2015). Clock outputs documented in A. thaliana are alsoobserved in crop species. For instance, in a single genotype of alegume, Phaseolus vulgaris, the circadian clock regulates photo-synthesis and stomatal conductance (Fredeen et al., 1991; Hen-nessey & Field, 1991; Hennessey et al., 1993). In experimentalsegregating progenies of B. rapa, quantitative clock variation (andclock quantitative trait loci) is associated with gas exchange(Edwards et al., 2011), and in the wild near-relative, Boecherastricta, quantitative clock variation correlates with root : shootallocation and growth rates (Salmela et al., 2015).

Here, we show that crop types of B. rapa differ significantly incircadian period. The mean period for the oilseed crop type was

c. 24 h, while the vegetable types have longer circadian cycles of25 h, on average, for leaf crops and nearly 26 h for turnips(Fig. 1). In combination with circadian clock differences, weobserved a significant crop type effect on both Amax and gs(Fig. 2), such that oilseeds have the highest photosynthetic rateswhile the leaf crops have the lowest. While clock period was notcorrelated with unadjusted values of gas exchange, we found thatcircadian period and stomatal conductance were negatively corre-lated after statistically factoring out differences in biochemicaldemand, as oilseeds with shorter circadian periods near 24 h hadhigher rgs (Fig. 3c). After statistically accounting for genotypicvariation in CO2 supply, circadian period and rAmax were posi-tively correlated; this suggests that long-period vegetable typeshave higher rAmax than oilseeds when gas supply is factored out(Fig. 3b). Although any mechanistic connection to the clockremains to be explained, the higher adjusted values of rAmax in

(a) (b)

(c) (d)

Fig. 4 Response of photosynthesis to internal CO2 concentrations at 25°C and light saturation. (a) Representative CO2 response curves of photosynthesisat 25°C for three crop types of Brassica rapa. (b) Differences in Rubisco carboxylation rate (Vcmax) (lmol CO2m

�2 s�1) between different crop types at25°C. (c) Differences in the ratio of the maximum electron transport rate driving ribulose-1,5-bisphosphate (RuBP) regeneration (Jmax) to the maximumRuBP carboxylation rate (Vcmax) at 25°C. Error bars indicate � 1SE. Different letters indicate statistically significant differences among crop types (P < 0.05).(d) Association between percentage nitrogen per unit leaf mass as a function of leaf area. Each symbol represents a genotype; crop types are coded bycolor. The line represents the following relationship: r =�0.44 (R2 = 0.19), P = 0.0007.

� 2015 The Authors

New Phytologist� 2015 New Phytologist TrustNew Phytologist (2015)

www.newphytologist.com

NewPhytologist Research 7

leaf crops relative to oilseeds may be explained, in part, by theirhigher Jmax : Vcmax; leaf crops may be more biochemically effi-cient in light harvesting and electron transport, while oilseedsmay allocate photosynthetic proteins preferentially to carbohy-drate metabolism to support metabolically expensive oil produc-tion (see later). Differences in carbon allocation and associatedfeedbacks on photosynthesis (Farrar & Jones, 2000; Paul &Pellny, 2003) could also contribute to higher adjusted rAmax invegetable types. Long-period vegetable crops also showed higher

amounts of total biomass accumulation than oilseeds in our study(Fig. 3d), suggesting that photosynthetic biochemistry of theformer genotypes, allocation feedbacks on photosynthesis, andpossibly allocation differences (e.g. to vegetative organs ratherthan metabolically expensive, low-mass seed oil) may outweightheir reduced CO2 supply in determining vegetative growth.

The observed association between period and adjusted rgs isconsistent with functional hypotheses regarding clock function,in which period lengths close to 24 h are predicted to increase gas

(a)

(b)

(c)

Fig. 5 Differences in stomatal density between Brassica rapa cropaccessions. (a) Stomatal density (number of stomata mm�2). Error barsindicate � 1 SE. Different letters indicate statistically significant differencesamong B. rapa crop types (P < 0.05). (b) Stomatal density of arepresentative oilseed; (c) stomatal density of a representative cabbage.

(a)

(b)

(c)

Fig. 6 Differences in anatomical parameters within different Brassica rapaaccessions from representative cross-sections. (a) Maximum depth ofadaxial epidermis; (b, c) oilseed (b) and turnip (c) cross-sections, in whichdifferences in adaxial epidermal depth and number of palisade layers arevisible. Bars, 100 lm. Error bars indicate � 1 SE. Different letters indicatestatistically significant differences among crop types (P < 0.05). Lightmicrographs have had the background removed and false color applied tohighlight differences in palisade parenchyma cell layers. Red lines delineatethe transition from palisade parenchyma to spongy mesophyll.

New Phytologist (2015) � 2015 The Authors

New Phytologist� 2015 New Phytologist Trustwww.newphytologist.com

Research

NewPhytologist8

exchange by providing a match to the duration of daily cycles.The association between period and adjusted rAmax as well as veg-etative biomass, in which cycles > 24 h correlate with increasedvalues of these two traits, requires further mechanistic investiga-tion; interestingly, vegetable crops of B. oleracea likewise exhibitlonger period lengths than floral crops (A. Millar, pers. comm.).Regardless of the exact mechanistic relationships, the resultsprovide a novel demonstration of correlated evolution betweenthe clock and gas exchange, which is potentially relevant tounderstanding the developmental origin of crop differences inphysiological traits as well as developmental targets for cropimprovement.

Gas exchange, leaf biochemistry, and yield

Selection for oil production is associated with higher gasexchange rates in multiple oil crops. For instance, photosynthesisaffects oil production in rapeseed, B. napus, by participating inlipid metabolism and serving as a source of NADPH and ATP(Ruuska et al., 2004; Goffman et al., 2005; Li et al., 2006). Culti-vars of short-season soybean (Glycine max) show improvement ofoil yield coupled with an increase in photosynthetic rates andstomatal conductance (Morrison et al., 1999; Cober et al., 2005).Oil production in Helianthus annuus is likewise positively corre-lated with leaf area, which is a proxy for photosynthetic rates insome species (Herv�e et al., 2001). Here, higher gas exchange isobserved in oilseed relative to vegetable crops, and, specifically,high values of gs in oilseeds support high values of Amax (Fig. 3a).In all of these species, high photosynthetic rates in oilseeds mightbe attributable to increased sink strength, because metabolicallyexpensive oil production drives high carbohydrate demand.

The main biochemical processes that contribute to photosyn-thetic performance in C3 plants are RuBP carboxylation rates(Vcmax) and how rapidly sugars are thus produced, as well asRuBP regeneration (expressed as Jmax) (Sage, 1994). The changein Jmax : Vcmax is correspondingly explained as a change in theallocation of photosynthetic proteins to light harvesting or tosugar production. Among B. rapa crop types, we observe thatVcmax values are highest among oilseeds while Jmax : Vcmax ratiosare higher among leaf crops than among either oilseeds orturnips. Combined with their high CO2 supply, oilseeds ofB. rapa seem to have more favorable allocation of these proteins,which leads to higher unadjusted values of Amax (Fig 2b). WhenCO2 supply is factored out, high Jmax : Vcmax values (in leaf crops)and possible feedback from allocation (in both leaf and rootcrops) may enhance adjusted values of rAmax in vegetable relativeto oilseed crops (Fig. 3b).

Patterns of carbohydrate production may be relevant tocrop diversification for different targets of harvest. The product ofcarboxylation of RuBP consists of two molecules of 3-phosphoglycerate, which is the precursor to a range of carbohy-drates that are stored and transported by other pathways. Asdescribed earlier, oilseeds probably require carbohydrates as asource of energy for oil production, which corresponds with ourresult of higher Vcmax values in this crop type (Fig. 4b). Variationin Vcmax can result from changes in Rubisco protein abundance

or its activation status, or a combination of both (Quick et al.,1991; Baker et al., 1997). Lower Vcmax rates are often explainedby lower amounts of Rubisco (Jacob et al., 1995; Nakano et al.,1997; Tissue et al., 1999) and sometimes by a low activation state(Sage et al., 1989; Cook et al., 1998). Usually, there is a positivecorrelation between light-saturating photosynthetic rate and Nconcentration, because N may reflect Rubisco abundance (Evans,1989). Consistent with this pattern, we observed that the amountof N was significantly correlated with the leaf area; oilseeds havesmaller leaves, more nitrogen per unit leaf mass, and potentiallymore Rubisco. Other processes that could explain crop differ-ences in RuBP regeneration and electron transport and that war-rant further investigation include regulation of enzymatic activityof photosynthetic pigment–protein complexes as well as enzymekinetics, activation states, and activation rates that affect photo-synthesis (Evans, 1987; Evans & Terashima, 1987; Bernacchiet al., 2002).

Leaf anatomy and gas exchange

In several studies, leaf anatomy has been shown to affect gasexchange performance (Reich et al., 1998; Niinemets, 1999;Wright et al., 2004; Franks & Farquhar, 2007). The overallresponse of stomatal conductance to stomatal density is straight-forward in the absence of other morphological changes, namelyan increase in stomatal density directly increases stomatal con-ductance for a given set of environmental conditions (Woodward& Bazzaz, 1988; Franks & Beerling, 2009). Several cultivatedspecies have increased stomatal density compared with the wildrelatives (e.g. in the genera Arachis, Beta, Gossypium, andPhaseolus), although some domesticated species show a decreasein stomatal density (e.g. the genera Avena, Cichorium, and Vicia)(Milla et al., 2013). We found that the increase in stomatal den-sity in oilseeds relative to vegetable crops (Fig. 5) was positivelyassociated with stomatal conductance. Leaf anatomical featurescan also affect several aspects of gas exchange. Specific mesophyllphenotypes, for instance, can affect light absorption and carbonfixation. Higher numbers of palisade mesophyll cell layers areassociated with higher photosynthetic rates (Vogelmann &Evans, 2002; Terashima et al., 2006). Fig. 6 shows an oilseedgenotype (159 – WO-181) that has five palisade layers, while thevegetable turnip (189 – VT-091) has three layers, a phenotypicpattern consistent with our observation of higher Amax in oilseedsthan in turnips. The uppermost 20% of the leaf absorbs light onthe leaf surface (Terashima & Saeki, 1983; Vogelmann et al.,1989); thus, a thinner upper epidermis may be beneficial forincreasing light absorption rates and CO2 fixation. We observethinner adaxial epidermis in oilseeds of B. rapa, which might con-tribute to their higher photosynthetic and stomatal conductancerates.

While past studies and the current study indicate an associa-tion between the circadian clock and gas exchange, there is noevidence in the literature that leaf morphological and anatomi-cal traits are under circadian clock regulation. There are manyannotated A. thaliana genes that control leaf anatomy anddevelopment (e.g. SDD1, YABBY genes, PHB, PHV, and

� 2015 The Authors

New Phytologist� 2015 New Phytologist TrustNew Phytologist (2015)

www.newphytologist.com

NewPhytologist Research 9

GTL). Using published A. thaliana transcriptome data (Coving-ton et al., 2008), we examined whether A. thaliana genes thatcontribute to anatomical or morphological features are undercircadian control, and found that no obvious candidate genesinvolved in anatomical patterning are regulated by the circa-dian clock.

Conclusions

Human selection during B. rapa diversification favored increasesin specific harvestable components. Our results, including PICs,indicate that circadian period is correlated over evolutionary timewith gas exchange and biomass accumulation in B. rapa. Theseresults are consistent with the long-standing view that targetedmodification of the circadian clock could enable further cropimprovement. More generally, the results suggest the hypothesisthat evolution at clock loci might partly contribute to increases inphotosynthetic rates and conductance commonly observed inoilseeds of other crops (Herv�e et al., 2001; Ruuska et al., 2004;Cober et al., 2005). The reversal in correlation between periodand rAmax (or biomass) (Fig. 3b,d) compared with period and rgs(Fig. 3c) could contribute to the maintenance of clock variationwithin the clade of B. rapa crop genotypes. We further observedthat enhanced photosynthesis is associated with leaf anatomicaland morphological traits in a manner consistent with other crops,and suggest that further functional studies, including clockmutants in crop morphotypes, are needed to dissect the specificgenetic mechanisms at work.

Acknowledgements

This work was supported by the National Science Foundationgrants (IOS-0923752 and IOS-1025965) to C.W., C.R.M. andB.E.E., and IOS-1306574 to R.L.B.

Author contributions

B.E.E., C.R.M., C.W., P.L. and Y.Y. planned and designed theresearch. C.E.E., P.L., R.L.B., T.L.A. and Y.Y. performed experi-ments, conducted growth-chamber experiments, performed leafhistological measures, and analyzed data. B.E.E., C.E.E., C.W.and Y.Y. wrote the manuscript.

References

Abbo S, van-Oss RP, Gopher A, Saranga Y, Ofner I, Peleg Z. 2014. Plant

domestication versus crop evolution: a conceptual framework for cereals and

grain legumes. Trends in Plant Science 19: 351–360.Baker N, Nogu�es S, Allen D. 1997. Photosynthesis and photoinhibition. In:

Lumsden P, ed. Plants and UV-B: responses to environmental change.Cambridge, UK: Cambridge University Press, 95–112.

Bernacchi CJ, Portis AR, Nakano H, von Caemmerer S, Long SP. 2002.

Temperature response of mesophyll conductance. Implications for the

determination of Rubisco enzyme kinetics and for limitations to photosynthesis

in vivo. Plant Physiology 130: 1992–1998.Cober ER, Morrison MJ, Ma B, Butler G. 2005. Genetic improvement rates of

short-season soybean increase with plant population. Crop Science 45: 1029–1034.

Cook A, Tissue D, Roberts S, Oechel W. 1998. Effects of long-term elevated

[CO2] from natural CO2 springs on Nardus stricta: photosynthesis,biochemistry, growth and phenology. Plant, Cell & Environment 21: 417–425.

Cook M, Evans L. 1983. Some physiological aspects of the domestication and

improvement of rice. Field Crops Research 6: 219–238.Covington MF, Maloof JN, Straume M, Kay SA, Harmer SL. 2008. Global

transcriptome analysis reveals circadian regulation of key pathways in plant

growth and development. Genome Biology 9: R130.Dalchau N, Baek SJ, Briggs HM, Robertson FC, Dodd AN, Gardner MJ,

Stancombe MA, Haydon MJ, Stan G-B, Gonc�alves JM. 2011. The circadian

oscillator gene GIGANTEAmediates a long-term response of the Arabidopsisthaliana circadian clock to sucrose. Proceedings of the National Academy ofSciences, USA 108: 5104–5109.

Darwin C. 1868. The variation of animals and plants under domestication.London, UK: John Murray.

De Vries I. 1997.Origin and domestication of Lactuca sativa L. Genetic Resourcesand Crop Evolution 44: 165–174.

Del Carpio DP, Basnet RK, Arends D, Lin K, De Vos RC, Muth D, Kodde J,

Boutilier K, Bucher J, Wang X. 2014. Regulatory network of secondary

metabolism in Brassica rapa: insight into the glucosinolate pathway. PLoS ONE9: e107123.

Diaz-Uriarte R, Garland T. 1996. Testing hypotheses of correlated evolution

using phylogenetically independent contrasts: sensitivity to deviations from

Brownian motion. Systematic Biology 45: 27–47.Diaz-Uriarte R, Garland T. 1998. Effects of branch length errors on the

performance of phylogenetically independent contrasts. Systematic Biology 47:654–672.

Dodd AN, Salathia N, Hall A, Kevei E, Toth R, Nagy F, Hibberd JM, Millar

AJ, Webb AA. 2005. Plant circadian clocks increase photosynthesis, growth,

survival, and competitive advantage. Science 309: 630–633.Edwards CE, Ewers BE, Williams DG, Xie Q, Lou P, Xu X, McClung CR,

Weinig C. 2011. The genetic architecture of ecophysiological and circadian

traits in Brassica rapa. Genetics 189: 375–390.Edwards CE, Haselhorst MS, McKnite AM, Ewers BE, Williams DG, Weinig

C. 2009. Genotypes of Brassica rapa respond differently to plant-inducedvariation in air CO2 concentration in growth chambers with standard and

enhanced venting. Theoretical and Applied Genetics 119: 991–1004.Ethier G, Livingston N. 2004.On the need to incorporate sensitivity to CO2

transfer conductance into the Farquhar–von Caemmerer-Berry leaf

photosynthesis model. Plant, Cell & Environment 27: 137–153.Evans J. 1987. The relationship between electron transport components and

photosynthetic capacity in pea leaves grown at different irradiances. AustralianJournal of Plant Physiology 14: 157–170.

Evans JR. 1989. Photosynthesis and nitrogen relationships in leaves of C3 plants.

Oecologia 78: 9–19.Evans J, Terashima I. 1987. Effects of nitrogen nutrition on electron transport

components and photosynthesis in spinach. Australian Journal of PlantPhysiology 14: 59–68.

Evans LT. 1996. Crop evolution, adaptation and yield. Cambridge, UK:

Cambridge University Press.

Farquhar G, von Caemmerer SV, Berry J. 1980. A biochemical model of

photosynthetic CO2 assimilation in leaves of C3 species. Planta 149: 78–90.

Farrar J, Jones D. 2000. The control of carbon acquisition by roots. NewPhytologist 147: 43–53.

Faure S, Turner AS, Gruszka D, Christodoulou V, Davis SJ, von Korff M,

Laurie DA. 2012.Mutation at the circadian clock gene EARLY MATURITY 8adapts domesticated barley (Hordeum vulgare) to short growing seasons.Proceedings of the National Academy of Sciences, USA 109: 8328–8333.

Felsenstein J. 1985. Confidence limits on phylogenies: an approach using the

bootstrap. Evolution 39: 783–791.Fowke LC, Rennie PJ. 1995. Botanical microtechnique for plant cultures. In:

Gambourg OL, Philips GC, eds. Plant cell, tissue and organ culture. Heidelberg,

Germany: Springer, 217–228.Franks PJ, Beerling DJ. 2009.Maximum leaf conductance driven by CO2 effects

on stomatal size and density over geologic time. Proceedings of the NationalAcademy of Sciences, USA 106: 10343–10347.

New Phytologist (2015) � 2015 The Authors

New Phytologist� 2015 New Phytologist Trustwww.newphytologist.com

Research

NewPhytologist10

Franks PJ, Farquhar GD. 2007. The mechanical diversity of stomata and its

significance in gas-exchange control. Plant Physiology 143: 78–87.Fredeen AL, Gamon JA, Field CB. 1991. Responses of photosynthesis and

carbohydrate-partitioning to limitations in nitrogen and water availability in

field-grown sunflower. Plant, Cell & Environment 14: 963–970.Garland T, Harvey PH, Ives AR. 1992. Procedures for the analysis of

comparative data using phylogenetically independent contrasts. SystematicBiology 41: 18–32.

Goffman FD, Alonso AP, Schwender J, Shachar-Hill Y, Ohlrogge JB. 2005.

Light enables a very high efficiency of carbon storage in developing embryos of

rapeseed. Plant Physiology 138: 2269–2279.Graf A, Schlereth A, Stitt M, Smith AM. 2010. Circadian control of

carbohydrate availability for growth in Arabidopsis plants at night. Proceedings ofthe National Academy of Sciences, USA 107: 9458–9463.

Greenham K, Lou P, Remsen SE, Farid H, McClung CR. 2015. TRiP: Tracking

Rhythms in Plants, an automated leaf movement analysis program for circadian

period estimation. Plant Methods 11: 33.Hennessey TL, Field CB. 1991. Circadian rhythms in photosynthesis:

oscillations in carbon assimilation and stomatal conductance under constant

conditions. Plant Physiology 96: 831–836.Hennessey TL, Freeden AL, Field CB. 1993. Environmental effects on circadian

rhythms in photosynthesis and stomatal opening. Planta 189: 369–376.Hennig K, de Vos RC, Maliepaard C, Dekker M, Verkerk R, Bonnema G.

2014. A metabolomics approach to identify factors influencing glucosinolate

thermal degradation rates in Brassica vegetables. Food Chemistry 155: 287–297.Herv�e D, Fabre F, Berrios EF, Leroux N, Al Chaarani G, Planchon C, Sarrafi A,

Gentzbittel L. 2001.QTL analysis of photosynthesis and water status traits in

sunflower (Helianthus annuus L.) under greenhouse conditions. Journal ofExperimental Botany 52: 1857–1864.

Jacob J, Greitner C, Drake B. 1995. Acclimation of photosynthesis in relation to

Rubisco and non-structural carbohydrate contents and in situ carboxylaseactivity in Scirpus olneyi grown at elevated CO2 in the field. Plant, Cell &Environment 18: 875–884.

Kebede H, Martin B, Nienhuis J, King G. 1994. Leaf anatomy of two

Lycopersicon species with contrasting gas exchange properties. Crop Science 34:108–113.

Kim JA, Kim JS, Hong JK, Lee Y-H, Choi B-S, Seol Y-J, Jeon CH. 2012.

Comparative mapping, genomic structure, and expression analysis of eight

pseudo-response regulator genes in Brassica rapa.Molecular Genetics andGenomics 287: 373–388.

Li R-J, Wang H-Z, Mao H, Lu Y-T, Hua W. 2006. Identification of

differentially expressed genes in seeds of two near-isogenic Brassica napus lineswith different oil content. Planta 224: 952–962.

Long S, Bernacchi C. 2003. Gas exchange measurements, what can they tell us

about the underlying limitations to photosynthesis? Procedures and sources of

error. Journal of Experimental Botany 54: 2393–2401.Lou P, Wu J, Cheng F, Cressman LG, Wang X, McClung CR. 2012.

Preferential retention of circadian clock genes during diploidization following

whole genome triplication in Brassica rapa. Plant Cell 24: 2415–2426.Lou P, Xie Q, Xu X, Edwards C, Brock M, Weinig C, McClung C. 2011.

Genetic architecture of the circadian clock and flowering time in Brassica rapa.Theoretical and Applied Genetics 123: 397–409.

McClung CR. 2006. Plant circadian rhythms. Plant Cell 18: 792–803.Meyer RS, Purugganan MD. 2013. Evolution of crop species: genetics of

domestication and diversification. Nature Reviews Genetics 14: 840–852.Midford P, Garland TJr, Maddison W. 2005. PDAP package of Mesquite.Version 1.15. [WWW document] URL www. mesquiteproject. org/

pdap_mesquite [accessed 9 June 2012].

Milla R, de Diego-Vico N, Mart�ın-Robles N. 2013. Shifts in stomatal traits

following the domestication of plant species. Journal of Experimental Botany 64:3137–3146.

MoghaddamMRB, Van den Ende W. 2013. Sweet immunity in the plant

circadian regulatory network. Journal of Experimental Botany 64: 1439–1449.Morrison MJ, Voldeng HD, Cober ER. 1999. Physiological changes from

58 years of genetic improvement of short-season soybean cultivars in Canada.

Agronomy Journal 91: 685–689.

Murakami M, Ashikari M, Miura K, Yamashino T, Mizuno T. 2003. The

evolutionarily conserved OsPRR quintet: rice pseudo-response regulators

implicated in circadian rhythm. Plant and Cell Physiology 44: 1229–1236.

Nakano H, Makino A, Mae T. 1997. The effect of elevated partial pressures of

CO2 on the relationship between photosynthetic capacity and N content in rice

leaves. Plant Physiology 115: 191–198.Niinemets €U. 1999. Components of leaf dry mass per area – thickness anddensity – alter leaf photosynthetic capacity in reverse directions in woody

plants. New Phytologist 144: 35–47.Paul MJ, Pellny TK. 2003. Carbon metabolite feedback regulation of leaf

photosynthesis and development. Journal of Experimental Botany 54: 539–547.

Purugganan MD, Boyles AL, Suddith JI. 2000. Variation and selection at the

CAULIFLOWER floral homeotic gene accompanying the evolution of

domesticated Brassica oleracea. Genetics 155: 855–862.Purugganan MD, Fuller DQ. 2009. The nature of selection during plant

domestication. Nature 457: 843–848.Quick W, Schurr U, Fichtner K, Schulze ED, Rodermel S, Bogorad L, Stitt M.

1991. The impact of decreased Rubisco on photosynthesis, growth, allocation

and storage in tobacco plants which have been transformed with antisense rbcS.Plant Journal 1: 51–58.

Reich P, Ellsworth D, Walters M. 1998. Leaf structure (specific leaf area)

modulates photosynthesis–nitrogen relations: evidence from within and across

species and functional groups. Functional Ecology 12: 948–958.Ruuska SA, Schwender J, Ohlrogge JB. 2004. The capacity of green oilseeds to

utilize photosynthesis to drive biosynthetic processes. Plant Physiology 136:2700–2709.

Sack L, Scoffoni C, John GP, Poorter H, Mason CM, Mendez-Alonzo R,

Donovan LA. 2013.How do leaf veins influence the worldwide leaf economic

spectrum? Review and synthesis. Journal of Experimental Botany 64: 4053–4080.

Sage RF. 1994. Acclimation of photosynthesis to increasing atmospheric CO2:

the gas exchange perspective. Photosynthesis Research 39: 351–368.Sage RF, Sharkey TD, Seemann JR. 1989. Acclimation of photosynthesis to

elevated CO2 in five C3 species. Plant Physiology 89: 590–596.Sakai WS. 1973. Simple method for differential staining of paraffin embedded

plant material using toluidine blue O. Biotechnic and Histochemistry 48: 247–249.

Salmela MJ, Greenham K, Lou P, McClung CR, Ewers BE, Weinig C. 2015.

Variation in circadian rhythms is maintained among and within populations in

Boechera stricta. Plant, Cell & Environment. doi: 10.1111/pce.12670.Schneider CA, Rasband WS, Eliceiri KW, Schindelin J, Arganda-Carreras

I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S. 2012. 671

NIH image to imageJ: 25 years of image analysis. Nature Methods 9:671–675.

Smartt J. 1988. Morphological, physiological and biochemical changes in

Phaseolus beans under domestication. In: Gepts P, ed. Genetic resourcesof Phaseolus beans. Dordrecht, the Netherlands: Springer, 143–161.

Swofford DL. 2002. Phylogenetic analysis using parsimony (* and other methods).Version 4. Sunderland, MA, USA: Sinauer Associates.

Takata N, Saito S, Tanaka Saito C, Nanjo T, Shinohara K, Uemura M. 2009.

Molecular phylogeny and expression of poplar circadian clock genes, LHY1 andLHY2. New Phytologist 181: 808–819.

Terashima I, Hanba YT, Tazoe Y, Vyas P, Yano S. 2006. Irradiance and

phenotype: comparative eco-development of sun and shade leaves in

relation to photosynthetic CO2 diffusion. Journal of Experimental Botany57: 343–354.

Terashima I, Saeki T. 1983. Light environment within a leaf I. Optical properties

of paradermal sections of Camellia leaves with special reference to differences inthe optical properties of palisade and spongy tissues. Plant and Cell Physiology24: 1493–1501.

Tissue DT, Griffin KL, Ball JT. 1999. Photosynthetic adjustment in field-grown

ponderosa pine trees after six years of exposure to elevated CO2. Tree Physiology19: 221–228.

� 2015 The Authors

New Phytologist� 2015 New Phytologist TrustNew Phytologist (2015)

www.newphytologist.com

NewPhytologist Research 11

Tsunoda S, Hinata K, Gomez-Campo C. 1980. Brassica crops and wildallies: biology and breeding. Tokyo, Japan: Japan Scientific Societies

Press.

Vogelmann T, Bornman J, Josserand S. 1989. Photosynthetic light

gradients and spectral regime within leaves of Medicago sativa.Philosophical Transactions of the Royal Society of London Series B: BiologicalSciences 323: 411–421.

Vogelmann T, Evans J. 2002. Profiles of light absorption and chlorophyll within

spinach leaves from chlorophyll fluorescence. Plant, Cell & Environment 25:1313–1323.

Whittall JB, Hodges SA. 2007. Pollinator shifts drive increasingly long nectar

spurs in columbine flowers. Nature 447: 706–709.Woodward F, Bazzaz F. 1988. The responses of stomatal density to CO2 partial

pressure. Journal of Experimental Botany 39: 1771–1781.Wright IJ, Reich PB, Westoby M, Ackerly DD, Baruch Z, Bongers F,

Cavender-Bares J, Chapin T, Cornelissen JH, Diemer M. 2004. The

worldwide leaf economics spectrum. Nature 428: 821–827.Xie Q, Lou P, Hermand V, Aman R, Park HJ, Yun D-J, Kim WY,

Salmela MJ, Ewers BE, Weinig C. 2015. Allelic polymorphism of

GIGANTEA is responsible for naturally occurring variation in circadian

period in Brassica rapa. Proceedings of the National Academy of Sciences,USA 112: 3829–3834.

Zhao J, Wang X, Deng B, Lou P, Wu J, Sun R, Xu Z, Vromans J,

Koornneef M, Bonnema G. 2005. Genetic relationships within Brassicarapa as inferred from AFLP fingerprints. Theoretical and Applied Genetics110: 1301–1314.

Supporting Information

Additional supporting information may be found in the onlineversion of this article.

Fig. S1 Root morphology of different Brassica oilseed plant geno-types.

Table S1 List of experimental Brassica rapa genotypes

Table S2 Genotypic means for the gas exchange parameters indifferent genotypes of Brassica rapa

Table S3 ANOVA tables

Please note: Wiley Blackwell are not responsible for the contentor functionality of any supporting information supplied by theauthors. Any queries (other than missing material) should bedirected to the New Phytologist Central Office.

New Phytologist is an electronic (online-only) journal owned by the New Phytologist Trust, a not-for-profit organization dedicatedto the promotion of plant science, facilitating projects from symposia to free access for our Tansley reviews.

Regular papers, Letters, Research reviews, Rapid reports and both Modelling/Theory and Methods papers are encouraged. We are committed to rapid processing, from online submission through to publication ‘as ready’ via Early View – our average timeto decision is <27 days. There are no page or colour charges and a PDF version will be provided for each article.

The journal is available online at Wiley Online Library. Visit www.newphytologist.com to search the articles and register for tableof contents email alerts.

If you have any questions, do get in touch with Central Office (np-centraloffice@lancaster.ac.uk) or, if it is more convenient,our USA Office (np-usaoffice@lancaster.ac.uk)

For submission instructions, subscription and all the latest information visit www.newphytologist.com

New Phytologist (2015) � 2015 The Authors

New Phytologist� 2015 New Phytologist Trustwww.newphytologist.com

Research

NewPhytologist12

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.