Patterns of Rust Infection As a Function of Host Genetic Diversity and Host Density in Natural Populations of the Apomictic Crucifer, Arabis Holboellii Author(S): B
Total Page:16
File Type:pdf, Size:1020Kb
Patterns of Rust Infection as a Function of Host Genetic Diversity and Host Density in Natural Populations of the Apomictic Crucifer, Arabis holboellii Author(s): B. A. Roy Source: Evolution, Vol. 47, No. 1 (Feb., 1993), pp. 111-124 Published by: Society for the Study of Evolution Stable URL: http://www.jstor.org/stable/2410122 . Accessed: 29/01/2014 15:38 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Society for the Study of Evolution is collaborating with JSTOR to digitize, preserve and extend access to Evolution. http://www.jstor.org This content downloaded from 128.223.93.178 on Wed, 29 Jan 2014 15:38:41 PM All use subject to JSTOR Terms and Conditions Evolution,47(1), 1993, pp. 11 1-124 PATTERNS OF RUST INFECTION AS A FUNCTION OF HOST GENETIC DIVERSITY AND HOST DENSITY IN NATURAL POPULATIONS OF THE APOMICTIC CRUCIFER, ARABIS HOLBOELLII B. A. Roy' Rancho Santa Ana Botanic Garden, Claremont,CA 91711 USA and Rocky Mountain Biological Laboratory,Crested Butte, CO 81224 USA Abstract.-It is oftenassumed that geneticdiversity contributes to reduced disease incidence in natural plant populations. However, littleis known about the geneticstructure of natural popu- lationsaffected by disease. Here I presentdata fromthree apomictic (asexual) populationsof Arabis holboelliiinfectedby therusts Puccinia monoica and P. thlaspeos.An averageof 300 hostindividuals per populationwere genotyped (using seven variable allozyme loci) and scoredfor disease presence. Arabis holboelliipopulations are geneticallydiverse; the numberof clones detectedper population rangedfrom 6 to 27. There was substantialvariation in frequencyof host clones withinand among sites, and significantvariation among clones in susceptibilityto the differentrusts. Contraryto predictionsbased on frequency-dependentselection theorythere was not a consistentpositive relationshipbetween clone frequencyand disease incidence withinany of the populations (Spear- man's r = -0.096, P > 0.5). In addition, clonally diverse populations did not necessarilyhave decreased disease incidence. The population with the lowest overall (both pathogenscombined) disease incidence(7.5 ? 1.9%) had the smallestnumber of clones (6), the lowestspatial variability, and the highestArabis density.By comparison, another population had 22 clones, high spatial variability,low Arabis densityand significantlymore disease overall (16.8 ? 2.7%). Althoughthis studydoes not eliminate the possibilityof frequency-dependentpathogen attack in these popula- tions, the resultssuggest that it is likelyto be weak or intermittent. Key words.-Arabis, density,epidemiology, frequency-dependence, plant pathogens,Puccinia. Received December 6, 1991. Accepted June 12, 1992. Genetically variable clonal populations goingpathogen-mediated frequency-depen- are difficultto explain undertraditional eco- dent selection: logical theory,because the most productive Prediction 1: Within populations, geno- and competitive clone should drive all of types (clones) occurringat low frequency the othersto extinction(Bell, 1982; Sebens should have a fitnessadvantage ("rarityad- and Thorne, 1985; Begon et al., 1986). vantage") (Haldane, 1949; Jaenike et al., However, there are several ways genetic 1980; Hamilton, 1980; Antonovics and variabilityin clonal populations could be Ellstrand,1984; Parker,1989; Lively et al., maintained, including frequency-depen- 1990; Alexander, 1991; Parker,1992). That dent selectionby pathogens,heterogeneous is, genotypeswith rare resistance alleles will environmentsfavoring one clone then an- escape infectionuntil the resistantgenotype other, the continuous generation of new is common enough in the population that clones from sexual progenitors,rapid mu- selectionon the pathogenfavors strains that tation,and coexistenceof similarclones for can overcome the resistanthosts. short times due to neutral stability(Bell, Prediction2: Populations withmore vari- 1982; Sebens and Thorne, 1985). One ofthe ation in resistanceshould have less disease least studiedof these mechanismsis the po- thanmore uniform ones (Adams et al., 1971; tential for pathogens or parasites to exert Harlan, 1976; Bremermann, 1980, 1983; frequency-dependentselection. Barrett,1981; Alexander, 1988). Decreased Two predictionsare typicallymade con- disease incidence is generallyexpected in cerningpopulations thoughtto be under- variable versus uniform populations be- cause the absolute numberof hosts suscep- tible to a given pathogen genotypewill be ' Presentaddress: BotanyDepartment, University of lower, the distance between susceptible California,Davis, CA 95616 USA. plants will be greater, making pathogen 111 ?0 1993 The Society forthe Study of Evolution. All rightsreserved. This content downloaded from 128.223.93.178 on Wed, 29 Jan 2014 15:38:41 PM All use subject to JSTOR Terms and Conditions 112 B. A. ROY transmissiondifficult, and pathogenpropa- geneticdifferences in susceptibility(Falcon- gules landing on resistanthosts will be lost er, 1981). Second, the clonal nature of the to the system (Burdon, 1987; Alexander, plants may make it easier to detect fre- 1988). quency-dependenteffects. This is because a The distance between susceptible plants population composed of clones originating can be largebecause eithera) hostresistance fromdifferent parental stocks may be more alleles are diverseand thusthe frequencyof stronglydifferentiated than a sexual popu- individuals in a susceptiblegenotype is low, lation in whichthere may be more genomic or b) because the actual physicalproximity sharing(Tibayrenc and Ayala, 1988). (density)of hosts is low. Thus, host density, like host frequency,is typicallypredicted to MATERIALS AND METHODS be positivelyassociated with disease inci- dence because more and closerhost 'targets' The Hosts are assumed to improve the probabilityof The hosts, Arabis holboellii (Brassica- pathogentransmission (Anderson and May, ceae), are biennial to short-livedperennial 1979; Burdon and Chilvers, 1982; Alex- plants occurring on well-drained soils ander, 1988; Burdon et al., 1989). Never- throughoutNorth America and extending theless,host densityis not always positively north into Greenland (Rollins, 1941). Re- associated with disease incidence (Burdon productionin A. holboelliimay be sexual or and Chilvers, 1982; Burdon, 1987). apomictic (Bocher, 1947, 1951, 1969; Ny- The presentstudy is partof a largereffort gren, 1954). I have verified,by emascula- to determinewhether frequency-dependent tion and crossingexperiments (Roy, 1992; selection by pathogenicrust fungiis struc- unpubl. data) and by electrophoresis of turingnatural populations of the apomictic progenyarrays (Roy and Rieseberg, 1989), (asexual) crucifer,Arabis holboellii.Arabis that the plants in the studypopulations re- holboelliidisplays heritable variation in re- produce by pseudogamous apomixis. In sistanceto rustinfection, and host fitnessis pseudogamous apomicts, pollen is neces- reduced by infection(Roy, 1992). For ex- sary for successfulseed production (prob- ample, in a singleseason, uninfectedplants ably requiredfor endosperm formation) but floweredabout 31% of the time versus0.4% does not fertilizethe ovules (Richards,1986; for those infectedby Puccinia monoica or Asker and Jerling,1992). The geneticcon- neverfor those infected by P. thlaspeos.Sur- sequence of thisform of apomixis is thatall vival of uninfectedplants and of those in- progenyfrom a single motherare identical fectedby P. thlaspeos was relativelyhigh, to each otherand to theirmother (Richards, 65% and 64% respectively,but declined to 1986). Arabis holboelliiplants do not spread 40% or less forthose infectedby the more vegetatively,but theyare neverthelessclon- common rust,P. monoica. The goals of the al (sensu Buss, 1985; Ellstrandand Roose, presentstudy were to determine,in natural 1987), because all the membersof a family populations of Arabis holboellii,(1) the ex- are geneticallyidentical. I have chosen to tentof clonal variability,(2) whetherclones call theseindependent plants individuals in- varied in frequencyand in disease incidence stead of ramets to emphasize that there is withinand among populations,and (3) the no physiologicalconnection between plants. effectsof host densityon disease incidence. I referto groups of individuals with iden- The fact that Arabis holboellii is clonal tical allozyme phenotypesas clones. has importantimplications for this study. First,it should be possible to assess disease The Pathogens incidenceat the clonal level,in situ,because The pathogens, Puccinia monoica (Pk.) each genotypein a given population is rep- Arth. and P. thlaspeos C. Schub. [listed as resented by many replicate, genetically P. holboelliiin Anonymous (1960) and Ar- identical individuals. I am assuming that thur(1962)], cause systemicrust disease on the environmentalinfluences on infection Arabis species (Anonymous, 1960; Farr et probabilityare spread among these repli- al., 1989). Althoughthese rustsare closely cated genotypesand that differencesin in- related to each other (Savile, 1974; Cum- fectionat the clonal level thereforereflect