COMPLEX DISPERSAL STRATEGY IN PICRIS CUPULIGERA (ASTERACEAE, LACTUCEAE)

Daniel Petit UMR 1061, INRA/ Universite de Limoges, 123, av. A. Thomas, 87 Mcriscupuligera is an annual taxon of Lactuceae from North Africa 060 Limoges Cedex, Rlivin g mainly in arid, semi-arid and sub-humid areas and with mild France winters. We describe three kinds of dispersal: epizoochory of capitules linked to [email protected] their peduncles, anemochory of free central achenes of capitules, and barochory of peripheral achenes stuck to receptacle and involucral bracts. The proportion of each kind depends on the habitat, as zoochory is more important in open than in woodl and habitat. Anemochory is the earliest mode of dispersal as it happens as soon as the involucral bracts are dried out. Barochory occurs more progressively and last peripheral achenes are released independently unti 1 January. Key words: Zoochory, anemochory, barochory, bioclimatology, heterocarpy.

the annual species P. cupuligera, P. Picriscupuligera (Durieu) Walp.is an coronopifolia, P. willkommii, and P. annual plant belonging to the tribe asplenioides. Lactuceae (Asteraceae) [ 1 ]. This Heterocarpy is a feature consisting in species was first described from the presence of two kinds of seeds on Algeria by Durieu de Maisonneuve in the same plant, allowing a 1846 under the genus Spitzelia but diversification of dispersal modes. was later included in Picris subgenus It can concern morphology Spitzelia Schultz-Bip. in the flora of (dimorphism) and/or physiology Algeria [2]. It was also recorded in through differences in dormancy or Tunisia, Morocco and Spain [3]. germination[6], [7], [8].Heterocarpy Picris shares plumose pappus at the is an adaptation to unpredictable top of achenes with Leontodon, conditions which are often associated Hypochaeris, Helminthotheca, to arid and desert areas or disturbed Hedypnois, Scorzoneroides, and habitats[9], [10]. It is assumed that Vrospermum. All these genera are each morph is advantageous for a classified within the special type of environmental Hypochaeridinae, a subtriberecently conditions which vary as a function of revisited by molecular phylogenetic years and habitats. Heterocarpyin P. approaches [4], [5]. cupuligera has long been recognized. According to these authors, Indeed, in the Flora of Algeria [2], this Hedypnois is now included in the species was included in the sub-genus genus Leontodon whereas the former Sptizelia which gathers annual taxa subgenus Oporinia of Leontodon is with dimorph achenes: peripheral separated from Leontodon s.s. and ones without a beak surrounded by a renamed as genus Scorzoneroides. scarious cupule and central ones with Moreover, these authors extended the a long beak and a pappus of plumose content of Spitzelia and they now hairs. include the perennial species P. hispanica and P. scabra in addition to

Revue Agrobiologia 2013; N"04; 5-11 In a previous workrelated to media. calculated using the formula: Q2 = Asteraceae [ 11 ], we showed that in The collected specimens of P. 3.43 P/(M-m), with m = minima Morocco, heterocarpy was more cupuligera were dried and kept in means of the coldest month frequent in Lactuceaetribue than in herbarium sheets labeled with date (in°C),M = maxima means of the Cardueae, and in annual and and location and are now storedat hottest month (in °C) and P = annual hemicryptophyticspecies than in the University of Limoges rainfall (in mm)[ 15]. The perennial ones.In this paper, we collection. The plant architecture climagram was drawn using describe an original case where was described according to [12]. SYSTAT12[16]. heterocarpy is associated to another The inflorescence is divided in two The analyses of covariance dispersal type, epizoochory, by parts: the upper part (enrichment (ANCOVAs) were used to compare means of the strong hairs born by zone)with an increase in capitule the strength necessary for the outer bracts and by deciduous number borne by successive axes breaking of peduncle between two peduncles. The significance of this and the lower part (inhibition zone) stations, by considering their length complex strategy is discussed in the with a very low number of capitules. as covariable. All the statistical light of bioclimatic constraints. In order to quantify epizoochory analyses were performed using among other types of dispersal, we PAST2.17[17]. used a device imitating the coat of a Individuals of Picriscupuligera mammal. A small piece of wool was applied on each involucre so the were collected in two sites near Description of achene strong hairs of outer bracts grip on Casablanca (Morocco) in June 1985 heteromorphism it. The wool piece was hanged on a and May 1986. The first site is an The body of central achenes is ancient more or less abandoned suspension spring that measured the attenuate in a beak at the apex (about cemetery (open habitat), where strength necessary to detach the the third to the fourth of total length) grows a degraded matorral peduncles from the living plant. The and the wall bears longitudinal ribs with Asparagus s p p , length of the peduncle and the Chamaeropshumilis, strength necessary to split it from with around 36 wrinkles. The As phodelus micro carpus and the stem were recorded. pappus is composedof about 17 Urgineamaritima (altitude = 50 m, The distribution of the species was plumose bristles enlarged at the base 33°34'42" N and 7036'24" W). deduced from pur numerous field and 17 thin ones. The body of Herbaceous plants are dominant, trips in Morocco during 1985 and peripheral achenes is slightly including Poaceae, Brassicaceae 1986. A sample was collected in attenuate but never beaked, and is and Asteraceae. The second one is a each locality where the species was surrounded by a short scarious woodland with stands of Eucalyptus present and further kept in cupule (fig. 1). The central achenes spp in the area of Bouskoura, 10 km herbarium. These data were used to are independently released as the south of Casablanca (altitude =138 assess the bioclimagram of the involucre opens by desiccation, m, 33°26'56" N and 7°38'55" W). specieswhich wasdrawn according whereas the peripheral ones The spontaneous vegetation is the to the method of [13] and [14]. areclasped by the innermost same but less degraded with shrubs Briefly, the Emberger-Sauvage involucral bracts. of Pistadalentiscus and Phyllirea pluviometric coefficient was

Figure l.Central and peripheral achenes of Picriscupuligera. Invoiucres open when they get dried (A) with outer achenes still attached to the receptacle and innermost bracts (B). Note the strong hairs on the outer bracts. The central achenes (C) have a deciduous pappus. D: central achene; E: peripheral achene.

Revue Agrobiologia 2012; N°3; 5-11 The mean number of central mean number of 12.75 ± 0.45(N = individuals still erect. In the tw achenes per capitule is 34.63 ± 4.98 13).Their release occurs when the populations studied, the dispersal t (standard deviation with N= 16 involucre bracts get broken several more than 90% of peripheri different plants). The number of months later. For example in mid- achenes was already done and Chi peripheral achenes is more January at the cemetery station, we test supported the homogeneity < homogenous and we recorded a considered the P. cupuligera results (p=0.28).

Table 1 .Remaining peripheral achenes in mid-January at station 1. It is assumed that each receptacle retains a mean of 13 peripheral achenes.

Theorical Number of number of Percentages of Plant attached peripheral Observed number of remaining peripheral numbers capitules achenes peripheral achenes achenes

Population 1 18 159 2067 149 7.21

Population 2 12 192 2496 203 8.13

An original feature of P. cupuligera variability of attachment of increases with the length of th is thatsome capitules are attached to peduncles to the stem.To measure peduncles (p=4.5 10"5) (fig. 2). Ii their peduncle and thus possibly the breaking strength, we took into other words, the shorter tin dropped off as a whole. As a result, account 28 measurements made on peduncle, the higher the probabilit; all the peripheral achenes are still station 1 and 12 onstation2. Using of breaking and transport b; linked to the head at the moment of this dataset, the strength necessary mammal coat. dispersal. We investigated the to split the peduncles from the stem

y = 26.63x-4.48 R* = 0.506 p=433 10-5 g ) (i n forc e Breakin g

Length of peduncle (in cm)

Figure 2. Breaking strength and length of peduncles

Revue Agrobiologia 2013; N°04; 5-11 We then compared the values p=0.175) as it is required for relationships. Tt is interesting to note recorded on the plants collected ANCOVAs, and no difference was that the deciduous peduncles are incemetery and woodland sites. recorded (Fl3 =0.004, p=0.95). We mostly oriented toward the outside Given the strength-length can conclude that the plants from of the whole plant (fig. 3). These relationship, we performed an both sites do not differ by intrinsic peduncles are characterized by a ANCOVA with the same data. We constitution of stems or peduncles. thickening at the base of the head. checked the homogeneity of slopes Both kinds of plants share the same between the two groups (F=1.917, strength-length of peduncle

Figure 3. Deciduous peduncles (thickened at the top) and persistent peduncles (others).

The plant is seen from the top. The principal axis corresponds to the wide circle and the order of successive secondary axes are represented by numbers ... ,<-• , - .-¬ We then countered the numbers of deciduous peduncles from 1 Oplants in both stations (fig. 4).

Fig. 4. Proportions of deciduous peduncles in plants of both stations (N=10) peduncle s deciduou s

o f There is a significant greater proportion of deciduous capitules in the station 1 than in station 2 (ANCOVA: F=l 1.45, P=0.011). Proportio n

Stations

Revue Agrobiologia 2012; N°3; 5-11 We then tested whether the lengths lengths observed in Bouskoura corresponding to the inhibition zone of peduncles differed between both station are longer than in cemetery (ANOVA: N=17, F=12.85 and p = stations.Indeed, there is a station. Moreover, there is a 0.002, fig. 5). marginally significant difference significant difference in the (Kruskal-Wallis, p=0.063). The proportion of principal axis

Figure 5.Proportion of inhibition zone in the plants of both stations (N= 17) zon e

inhibitio n We also tested whether there was a significant difference

o f in the divergence angle of the successive ramifications from top to bottom between the plants of both stations. The means calculated with 5 plants from each site are

Proportio n given in table 2. They show that the angle, rather open in the first top ramifications, gets more and more acute downward. However, the angles recorded in station 2 are regularly more acute in station 2 than in station 1 Stations (Wilcoxon,p=0.043).

Table 2. Means of divergence angles in the successive ramifications of the principal axis

Station 1 Station 2

Order of ramifications Angle mean Standarddev Angle mean Standard dev

1 66.4 4.98 57 12.55

2 58 9.08 . 38.6 2.19

3 44 4.18 35.6 3.78

4 s 56.75. .9.43 37.5 3.54

5 56.75 15.67 32 9.90

Wilcoxon p -0.043

In conclusion, the architectures of the plants from both stations are clearly different, and summarized in fig. 6. The deciduous peduncles are Figure 6.Architectures of plants in locatedlower above the ground and both stations.h:inhibition zone; H: more distantly from the principal total length of principal axis; 1,2,3, axis in the case of plants growing in 4, 5: order of the successive open habitat (stationl) than those of Open habitat ramifications. woodland habitat (station 2). Cemetery

Woodland Bouskoura

Revue Agrobiologia 2013; N"04; 5-11 Bioclimatology According to fig. 7, Picriscupuligera is mainly found in arid, semi-arid and sub-humid stages, with mild winters (m>4°C).

Figure 7. Bioclimagram of Picriscupuligera. The lighter the blue lines, the higher the probability of Q 2 presence of E cupuligera. S: saharian stage, A: arid stage, SA: semi-arid stage, SH: sub-humid stage, and H: humid stage.

m°C Discussion since there are deciduous peduncles Barochory should be the system of Picriscupuligera is an annual plant that are dropped as a whole, bearing security as the mother plant found showing an original mode of all the peripheral achenes of a same favorable conditions to develop and dispersal. Indeed, the relatively capitule at the same time. We produce seeds. However this common dimorphism of achenes, recognize here the synaptospermy strategy is not appropriate in the described by Zohary[18] in the case of arid environment. Indeed, associated to anemo- and barochory, Near-East (several seeds stuck the water supply is low from the end is enriched by the possibility of together). However, this feature is of spring (flowering time)until the epizoochory. It is possible to facultative and varies according to first autumn rainfallsandit would quantify the theorical proportion of the habitat. The shade due to lead to a strong competition types of dispersal in open habitats. woodlands has an. effect on the between the seedlings. At this time, During the first days after capitule shape of the plants by comparison to it is unlikely that the released desiccation, the central achenes are individuals grown in sunny habitats: achenes can germinate due toa the first to be released and brought (i) the secondary axes are closer to strong dormancy. As several away by anemochory. They the principal axis, (ii) the axes Hypochaeridinae living in represent about the 75%of total bearing the deciduous peduncles are Mediterranean climates, we seeds per capitule. The facultative more elevated above the ground, hypothesize that the germination epizoochory allows diversifying the and (Hi) the peduncles are generally occurs during early spring rainfalls fete of peripheral achenes. The longer. As a result, the deciduous [8], [19]. Of note, we showed that proportion of capitules brought by peduncles are about twofold less most peripheral achenes were the fur of mammal is random but the frequent and less accessible to the already dropped off in mid-January. maximum is around 60% in open fur of mammals than in open It means that their release is habitats. So the percentage of habitat. So the epizoochory is progressive, limiting the attraction barochory is something between predicted to be as at least twofold of harvesterants suchas 10% (numerous mammals walking more effective for dispersal in open Messorspry'm autumn [20]. The around the plant) to 25 % (no than in woodland habitat, assuming possibility to escape from mammal present). an equal abundance of mammals in competition is enhanced by We demonstrated that epizoochory both environments. epizoochory. depends on the plant architecture

Revue Agrobiologia 2012; N°3; 5-11 In conclusion, the climatic Lactuceae) to be diphyletic. climates. Agrobiologia 2:21-28. constraints prevailing in arid and American Journal of Botany [13] Benfekih, L., Foucart, A., Peti semi-arid stages havefavored the 93:1193-1205. D. 2011. Central Sahara development of complex dispersal [5] Enke, N., Gemeinholzer B., populations of Locustamigratc strategies. Epizoochory, an original Zidorn, C. 2012. Molecular and riacinerascens (Orthopterz feature of Picriscupuligera, phytochemical systematics of the Acrididae) in irrigated perimeter; contributes to limit the intraspecific subtribe Hypochaeridinae is it a recent colonisatio competition and the predation by (Asteraceae, Cichorieae). event? Annales de la Societ harvesterantswhich are very Organisms Diversity and Evolution Entomologique de France 47: 14' abundant in these areas. 12:1-16. 153.

[6] Imbert, E., Escarre, J., Lepart, J. [14] Benkenana, N., Harrat, A 1997. Seed heteromorphism in We thank Mrs. F. Vallet of the Petit, D. 2012. The Pamphagida Crepis sancta: performance of two University of Limoges for her (Orthoptera) from East Algeria an morphs in different environments. technical assistance, and Dr G. description of a new specie; Oikos 79:325-332. Saladin for pictures and Zootaxa3168:22-38. constructing remarks. This work [7] Fumanal, B., Chauvel, B., [15] Sauvage, C. 1963. Etage received financial support via the Sabatier, A., Bretagnolle, F. 2007. bioclimatiques. Atlas du Marot program Tassili n°08MDU726. Variability and cryptic notice explicative (Rabat). heteromorphism of Ambrosia [16] Systat Inc. 2007. SYSTA artemisiifolia seeds: What 12.Manual and software. [1] Lack, H.-W. 1974. Die Gattung consequences for its invasion in Picris L., s. L, imostrnediterran- France? Annals of Botany 100:305¬ [17] HammerD.A.T., Harper P.D. I westasiatischen Raum. 313. Ryan P.A.S. 2001: Paleontologies statistics software package fo Dissertationen der Universitat [8] Cruz-Mano, G., Narbona, E., education and data analysis Verband der wissenschaftlichen Buide, M. L. 2010. Germination Palaeont. Electr.4 (9 Gesellschaften Osterreich patterns of dimorphic achenes in (VWGO), Dissertationen der three related species of http://palaeoelectronica.org Universitat Wien, vol.116, Austria, Scorzoneroides (Asteraceae, 2001J/past/issueI_01.htm. 184 pp. Lactuceae) growing in different [18] Zohary, M. 1950. Evolutionar environments, Annales Botanici [2] Battandier, LA. Trabut, L.1888. trends in the fruiting head c Fennici47:337-345. Flore de TAlgerie et catalogue des Compositae. Evolution 4:103-109. plantes du Maroc. Tome 1. [9] Venable, D. L. 1985. The [19] El-Keblawy, A. 2003. Effect Dicotyledones, par Battandier. 872 evolutionary ecology of seed of achene dimorphism on dormanc p. Alger. heteromorphism. American and progeny traits in the tw [3] Greuter, W. 2006-2009. Naturalist 126:577-595. ephemerals Hedypnoiscretica an Crepisaspera (Asteraceae' Compositae (pro parte majore). -In: [10] Venable, D.L.,Lawlor, L. 1980. Canadian Journal of Botany 81(6 Greuter, W. & Raab-Straube, E. von Delayed germination and dispersal 550-559. ( e d , ) : Compositae. in desert annuals: escape in space Euro+MedPlantbase - the and in time. Oecologia46:272-282. [20] Azcarate, F.M., Peco, B. 200: information resource for Euro- Spatial patterns of seed predatio Mediterranean plant diversity. [11] Petit, D. 2011. Strategies de byharvester ants (MessorForeX) i reproduction et de dissemination Mediterranean grassland an [4] Samuel, R., Gutermann, W., des Asteraceae dans les zones arides scrubland.InsectesSociaux50:120 Stuessy, T.F., Ruas, C.F., Lack, H.- duMaroc. Agrobiologia 1:16-19. W., Tremetsberger, K., Talavera, S., Hermanowski, B., Ehrendorfer, F. [12] Petit, D. 2012. The 2006. Molecular phylogenetics inflorescences of Asteraceae as an reveals Leontodon (Asteraceae, adaptation to unpredictable

Revue Agrobiologia 2013; N"04; 5-11