Inbreeding of selfing and outcrossing plant

Inbreeding of selfing and outcrossing plant

We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

I am reading John H. Gillespie's Population Genetics A Concise Guide Section 4.3 Inbreeding. I do not understand these two paragraphs quoted below concerning selfing and outcrossing.

  1. The first paragraph states that an outcrosser individual leaves behind on average two gametes one in an ovule and the other in a pollen.

  2. The second paragraph states that a mutant selfer will leave behind three gametes for every two of the outcrossing plants, with two in its selfed offspring and one in its outcrossed offspring.

What are the effect of the ovules and pollen here? How are these two statistical conclusions drawn? I would like to have a very detailed explanation.

Some interesting evolutionary questions arise with species that are capable of both selfing and outcrossing. For example, in many plant species there is an intrinsic advantage to selfing, which leads to the evolutionary conundrum: Why don't all plant species self? The situation is illustrated in Figure 4.4. The outcrossing pedigree on the right represents a typical individual in an outcrossing population of constant size. This individual leaves behind, on average, two gametes, one carried in an ovule and the other in a pollen grain. These gametes appear as filled circles in the figure.

Figure 4.4: The gametes produced by a selfer and an outcrosser. The p to the right of an arrow indicates that the parent's contribution came from pollen; an o indicates it came from an ovule. The filled circles represent gametes from the illustrated parents; the open circles represent gametes chosen at random from the gamete pool.

Suppose a mutant appears that self-fertilizes all of its ovules, M illustrated on the left side of the figure. Suppose also that there is enough pollen in each individual of this species that the few grains needed for self-pollination by the mutant represent a small fraction of the total pollen. As a consequence, the selfing mutant has essentially the same quantity of pollen available for outcrossing as does a nonselfing individual. All else being equal, the selfing mutant will leave behind three gametes for every two of the outcrossing plants, as indicated by the three filled circles in the figure. Two of the gametes are in its selfed offspring; one is in its outcrossed offspring. Thus, the mutant should increase in frequency, perhaps leading to the establishment of selfing as the usual mode of reproduction.

I don't think it's a terribly rigorous statistical argument, so much as a toy model arguing that there can be a reproductive advantage to selfing in an outcrossing population. Here is the logic as I see it:

  1. Ovule and pollen represent the female and the male reproductive gametes in a monoecious plant (monoecious means that it has both male and female reproductive organs in the same individual).
  2. In sexual organisms that do standard sex (i.e. not a complicated variety of sex), both ovule and pollen (gametes) are necessary to generate a fertilized offspring (zygote).
  3. In outcrossing individuals, pollen is broadcast to fertilize other individuals (1 pollen gamete) and ovules receive pollen from other individuals (1 ovule gamete). Total: 2 gametes
  4. In selfing individuals in an outcrossing population, selfing individuals guarantee that they fertilize their own ovules. Thus, they replace the pollen of other individuals (1 pollen gamete + 1 ovule gamete).
  5. Selfing individuals still broadcast pollen to fertilize other outcrossing individuals (+1 pollen gamete). Total: 3 gametes

The key is in looking at the figure and understanding what it's arguing.

Again, these are not really rigorous ratios, details will be determined by exact biology of any case in real life. As far as I can tell from the text, the argument is not "these are the exact ratios in all organisms", but rather that "selfing gives the advantage that a selfer contributes all of the same gametes as the outcrosser, plus the male gametes to fertilize its own male gametes".

Secondary note: this model is only true when the proportion of selfers in the outcrossing population is very small. Note that as the proportion of selfers increases, everyone is just fertilizing their own ovules, and the net contribution goes back down to 2 gametes per generation. So this does not describe an equilibrium case for population composition. [As noted in comments, it is likely the case that it is equilibrium for population size, e.g. number of individuals.]

For more theory about this, you can read about the evolution of self-(in)compatibility. Here is a somewhat recent review. Evolution of self-compatibility in plants is extremely dynamic, as you might expect from the arms-race aspect.


Pulling up some notes from the comments on the assumptions of the model:

  • I think that this example works when population number is at equilibrium, but population composition (e.g. selfers vs. outcrossers) might not be at equilibrium.
  • I think that for the amount of pollen, we have to assume that all organisms contribute an equal, infinite amount of pollen. So selfers fertilize their own, and then also dump pollen into a pool into which outcrossers also dump infinite pollen.
  • However, there are only a few ovules; every individual contributes one. That is, outcrossers and selfers contribute one ovule per individual.
  • On average, everyone's pollen will be sampled roughly once (actually a little less) if every non-selfed ovule takes one pollen from an infinite pollen pool of uniform composition.
  • As number of selfers in the population increases, less and less of the common pool of pollen will be sampled.
  • If everyone in the population is a selfer, there will obviously be no outcrossing, because everyone just fertilizes their own ovule. This is the equilibrium state, if selfing is heritable and this stated advantage to selfing holds.

Inbreeding depression under mixed outcrossing, self-fertilization and sib-mating

Biparental inbreeding, mating between two relatives, occurs at a low frequency in many natural plant populations, which also often have substantial rates of self-fertilization. Although biparental inbreeding is likely to influence the dynamics of inbreeding depression and the evolution of selfing rates, it has received limited theoretical attention in comparison to selfing. The only previous model suggested that biparental inbreeding can favour the maintenance of stable intermediate selfing rates, but made unrealistic assumptions about the genetic basis of inbreeding depression. Here we extend a genetic model of inbreeding depression, describing nearly recessive lethal mutations at a very large number of loci, to incorporate sib-mating. We also include a constant component of inbreeding depression modelling the effects of mildly deleterious, nearly additive alleles. We analyze how observed rates of sib-mating influence the mean number of heterozygous lethals alleles and inbreeding depression in a population reproducing by a mixture of self-fertilization, sib-mating and outcrossing. We finally use the ensuing relationship between equilibrium inbreeding depression and population selfing rate to infer the evolutionarily stable selfing rates expected under such a mixed mating system.


We show that for a given rate of inbreeding, sib-mating is more efficient at purging inbreeding depression than selfing, because homozygosity of lethals increases more gradually through sib-mating than through selfing. Because sib-mating promotes the purging of inbreeding depression and the evolution of selfing, our genetic model of inbreeding depression also predicts that sib-mating is unlikely to maintain stable intermediate selfing rates.


Our results imply that even low rates of sib-mating affect plant mating system evolution, by facilitating the evolution of selfing via more efficient purging of inbreeding depression. Alternative mechanisms, such as pollination ecology, are necessary to explain stable mixed selfing and outcrossing.

Experimental and genetic analyses reveal that inbreeding depression declines with increased self-fertilization among populations of a coastal dune plant

Theory predicts that inbreeding depression (ID) should decline via purging in self-fertilizing populations. Yet, intraspecific comparisons between selfing and outcrossing populations are few and provide only mixed support for this key evolutionary process. We estimated ID for large-flowered (LF), predominantly outcrossing vs. small-flowered (SF), predominantly selfing populations of the dune endemic Camissoniopsis cheiranthifolia by comparing selfed and crossed progeny in glasshouse environments differing in soil moisture, and by comparing allozyme-based estimates of the proportion of seeds selfed and inbreeding coefficient of mature plants. Based on lifetime measures of dry mass and flower production, ID was stronger in nine LF populations [mean δ = 1-(fitness of selfed seed/fitness of outcrossed seed) = 0.39] than 16 SF populations (mean δ = 0.03). However, predispersal ID during seed maturation was not stronger for LF populations, and ID was not more pronounced under simulated drought, a pervasive stress in sand dune habitat. Genetic estimates of δ were also higher for four LF (δ = 1.23) than five SF (δ = 0.66) populations however, broad confidence intervals around these estimates overlapped. These results are consistent with purging, but selective interference among loci may be required to maintain strong ID in partially selfing LF populations, and trade-offs between selfed and outcrossed fitness are likely required to maintain outcrossing in SF populations.

© 2013 The Authors. Journal of Evolutionary Biology © 2013 European Society For Evolutionary Biology.

Materials and Methods

Seeds were collected from three populations, which differed in population size and degree of fragmentation (Fig. 1, Table 1). In each locality, cones were collected by climbing trees separated usually at least 10–20 m from each other and distributed throughout the sampling area. The population of Chiquihuitlán consisted of approx. 4000 reproductive individuals distributed in one 10 ha forest fragment and tiny forest patches, surrounded by a matrix of pasturelands and agricultural fields. Recruitment was detected. The Barillas population was larger: c. 40 000 adult individuals. Pinus chiapensis (Mart.) Andresen dominated the forest, distributed in 10–100 ha forest fragments and small patches, surrounded by agricultural lands. Recruitment was rare. El Rincon was the largest population (> 50 000 adult individuals) distributed in < 1500 ha of secondary forests dominated by P. chiapensis, which resulted from abandoning of corn-cropping areas. Tropical montane cloud forest was the original vegetation ( Cordova & del Castillo, 2001 Bautista-Cruz & del Castillo, 2005 ). Regeneration was abundant in recently abandoned open areas.

Map of Mexico and Guatemala showing the location of the Pinus chiapensis populations investigated in this study. (Map by Raúl Rivera García.)

Population State/ Department* Altitude (m) Latitude (N) Longitude (W) Mean d.b.h. (cm) Mean per-progeny germination rate (% ± SE)
Chiquihuitlán Oaxaca 1162 18°00′ 96°46′ 48.0 42.1 (5.9)
El Rincón Oaxaca 1737 17°21′ 96°18′ 28.1 36.1 (2.2)
Barillas Huehuetenango* 1683 15°47′ 91°19′ 41.6 26.9 (3.2)

Allozyme analyses

Allozyme studies were conducted on seedlings from fresh seeds, that were less than 4-months old germinated on moist filter paper at 24°°C. Seed radicles were ground in 0.2 m phosphate buffer pH 7.5 ( Conkle et al., 1982 ). Extracts were adsorbed onto Whatman 4 mm chromatographic paper wicks. Samples were loaded on 12% starch gels. Aspartate amino transferase (AAT, E.C. and glucose-6-phosphate isomerase (GPI, E.C. were resolved on tris-citrate/LiB (pH 8.3) gels that run for 6–8 h at 75 mA ( Conkle et al., 1982 ). Phosphoglucomutase (PGM, E.C. was resolved on histidine-citrate, pH 7.0 that run for 5–6 h at 50 mA ( Cheliak & Pitel, 1984 ). Staining protocols for GPI, PGM and AAT followed Conkle et al. (1982 ) Cheliak & Pitel (1984 ) and Levy (1989 ), respectively. We used a variable number of seeds and maternal plants reflecting between-population variation in number of maternal trees available for sampling. Where several zones of activity were observed for a single enzyme, numerals following the enzyme abbreviation were used for identification. Band interpretation was based on segregation patterns observed in megagametophytes ( Ramírez Toro, 2005 ), a haploid seed reserve tissue in conifers, and on the overall conservation of isozyme subunits and isozyme numbers in plants ( Gottlieb, 1981 Wendel & Weeden, 1989 ).

Data analysis

Mating systems were analysed using Ritland's (2002 ) multilocus approach and software, which uses progeny arrays to calculate multilocus (tm) and single locus (ts) outcrossing rates, the parental F estimate, the extent that siblings share the same father (correlation of paternity) and the extent of variation among arrays for selfing rate (correlation of selfing). Multilocus outcrossing rate estimation assumes that the loci studied are unlinked ( Ritland, 2002 ). In P. chiapensis estimates of linkage disequilibrium were not significant for any pair of the above loci (R. F. del Castillo unpublished). Outcrossing rates were estimated from progeny genotypes with the Newton–Raphson method. Standard errors were calculated by 1000 bootstraps using the progeny array as resampling unit. Biparental inbreeding was tested by analysing the correlations of ts among loci, rs. The 1 – rs values are indicative of the fraction of biparental inbreeding over the total inbreeding detected and tend to be less biased than the traditional tmts used for biparental inbreeding assessment ( Ritland, 2002 ).

Inbreeding depression from mating to seedling stage may bias our mating system estimates, which were based on analyses of seedlings. To correct for such a bias, we used Eqn 1, with the following estimate of inbreeding depression:

(Ws and Wo are the viability of the self-and outcross seeds, respectively). Seed viability was estimated from the proportion of germinated seeds out of the total number of seeds attempted to germinate per progeny, usually 100 seeds, as already described. Ungerminated seeds had an empty or necrotic embryo when dissected and no detectable enzymatic activity. Ws and Wo were obtained from linear regression of per-progeny seed viability on per-progeny F-values, estimated from the isozyme analyses described earlier. A total of 50 progenies, for which we have both germination and isozyme data, were used in this analysis. For prediction purposes, no transformation was applied since untransformed data gave better fit and better (more homogenous) residual behavior than transformed data (angular and log odds), recommended for proportions ( Draper & Smith, 1981 ) or log transformation used for inbreeding depression analysis ( Charlesworth & Charlesworth, 1987 ). A lack-of-fit analysis was performed to test the adequacy of the model ( Draper & Smith, 1981 ). Wo and Ws were estimated from the values predicted at F = 0 and F = 0.5, respectively. The latter value is the expected F of self progenies from parents that were themselves product of random outcrossing, and was used in our analysis since the estimated F of the maternal plants from all studied populations did not reveal inbreeding (see the Results section).

To obtain an upper bound of the corrected outcrossing rate, we replaced the sm and δ values from Maki's formula (Eqn 1) with the upper 95% bound estimates of the estimated multilocus selfing rate sm = 1 – tm, assuming a normal distribution, and an estimate of δ using the upper 95% predicted values of Ws and Wo. Similarly, a lower bound estimate of the corrected outcrossing was obtained with the 95% lower bound estimates of selfing, with a δ-value calculated from the lower 95% predicted values of Ws and Wo. The SAS system version 9.1.3 (proc GLM) was used for statistical analyses ( SAS Institute, 1986 ).


Ågren, J., Schemske, D.W.: Outcrossing rate and inbreeding depression in two annual monoecious herbs, Begonia hirsuta and B. semiovata.-Evolution 47: 125–135, 1993.

Baumann, H., Künkele, S.: Die Gattung Serapias L.-eine taxonomische Übersicht.-Mitt. Bl. Arbeitskr. Heim. Orch. Baden-Württ. 20: 10–651, 1989.

Carr, D.E., Dudash, M.P.: Inbreeding depression in two species of Mimulus (Scrophulariaceae) with constrasting mating systems.-Amer. J. Bot. 83: 586–593, 1996.

Charlesworth, D., Charlesworth, B.: Inbreeding depression and its evolutionary consequences.-Annu. Rev. Ecol. Evol. 18: 237–268, 1987.

Delforge, P. (ed.): Orchids of Europe, North Africa and the Middle East.-Timber Press, Portland 2005.

Elle, E., Carney, R.: Reproductive assurance varies with flower size in Collinsia parviflora (Scrophulariaceae).-Amer. J. Bot. 90: 888–896, 2003.

Falconer, D.S. (ed.): Introduction to Quantitative Genetics. Second Edition.-Longman Press, London 1981.

Hedrick, P.W., Kalinowski, S.T.: Inbreeding depression in conservation biology.-Annu. Rev. Ecol. Evol. 31: 139–162, 2000.

Husband, B.C., Schemske, D.W.: Magnitude and timing of inbreeding depression in a diploid population of Epilobium angustifolium (Onagraceae).-Heredity 75: 206–215, 1996.

Johnson, S.D., Nilsson, L.A.: Pollen carryover, geitonogamy, and the evolution of deceptive pollination systems in orchids.-Ecology 80: 2607–2619, 1999.

Johnson, S.D., Peter, C.I., Ågren, J.: The effects of nectar addition on pollen removal and geitonogamy in the non-rewarding orchid Anacamptis morio.-Proc. roy. Soc. London B Biol. Sci. 271: 803–809, 2004.

Karoly, K.: Inbreeding effects on mating system traits for two species of Lupinus nanus.-Amer. J. Bot. 79: 49–56, 1992.

Neiland, M.R.M., Wilcock, C.C.: Maximisation of reproductive success by European Orchidaceae under conditions of infrequent pollination.-Protoplasma 187: 39–48, 1995.

Ollerton, J.: Reconciling ecological processes with phylogenetic patterns: the apparent paradox of plant-pollinator systems.-J. Ecol. 84: 767–769, 1996.

Parker, I.M., Nakamura, R.P., Schemske, D.W.: Reproductive allocation and the fitness consequences of selfing in two sympatric species of Epilobium (Onagraceae) with contrasting mating systems.-Amer. J. Bot. 82: 1007–1016, 1995.

Pellegrino, G., Gargano, D., Noce, M.E., Musacchio, A.: Reproductive biology and pollinator limitation in a deceptive orchid, Serapias vomeracea (Orchidaceae).-Plant Species Biol. 20: 19–26, 2005a.

Pellegrino, G., Musacchio, A., Noce, M.E., Palermo, A.M., Widmer, A.: Reproductive versus floral isolation among morphologically similar Serapias L. species (Orchidaceae).-J. Hered. 96: 15–23, 2005b.

Pico, F.X., Mix, C., Ouborg, N.J., Van Groenendael, J.M.: Multigenerational inbreeding in Succisa pratensis: effects on fitness components.-Biol. Plant. 51: 185–188, 2007.

Proctor, M., Yeo, P. (ed.): The Pollination of Flowers.-Collins, London 1973.

Rathcke, B., Real, L.: Autogamy and inbreeding depression in mountain laurel, Kalmia latifolia (Ericaceae).-Amer. J. Bot. 80: 143–146, 1993.

Robertson, J.L., Wyatt, R.: Reproductive biology of the yellow-fringed orchid, Platanthera ciliaris.-Amer. J. Bot. 77: 388–389, 1990.

Smithson, A.: Pollinator limitation and inbreeding depression in orchid species with and without nectar rewards.-New Phytol. 169: 419–430, 2006.

Tremblay, R.L., Ackerman, J.D., Zimmerman, J.K., Calvo, R.N.: Variation in sexual reproduction in orchids and its evolutionary consequences: a spasmodic journey to diversification.-Biol. J. Linn. Soc. 84: 1–54, 2005.

Van der Cingel, N.A. (ed.): An Atlas of Orchid Pollination.-A.A. Balkema, Rotterdam 1995.

Van der Pijl, L., Dodson, C.H. (ed.): Orchid Flowers: their Pollination and Evolution.-University of Miami Press, Coral Gables 1966.

Van Waes, J.M., Derbergh, P.C.: In vitro germination of some Western European orchids.-Physiol. Plant. 67: 253–261, 1986.

Vöth, W.: Können Serapias blüten Nesttäuschblumen sein?-Die Orchidee 31: 159–162, 1980.


Sex is costly. You could die trying to find a mate. Your mate could kill you, or give you a disease. You could be unable to find a mate in the first place, in which case you’d be better off if you could reproduce asexually. Even without those risks, though, even in a simple genetic simulation, sexual reproduction means you only pass on half of your genes to your offspring.

So why do it? We know that it’s possible to reproduce without sex lots of things do. It’s not just bacteria and protists, either: asexual reproduction occurs in some plants, insects, snails, amphibians, and reptiles, among many others. The logic of natural selection suggests that sex must confer some benefit that outweighs all the costs, at least in some situations. Essentially all of the proposed benefits of sex have to do with outcrossing, or mixing your genes with those of another, genetically distinct, individual.

Nevertheless, a lot of things that reproduce sexually do so without outcrossing. This is especially common in plants, where it’s called “self-pollination” or just “selfing.” Selfing is thought to provide short-term advantages relative to outcrossing–basically by avoiding the costs I’ve listed above. However, selfing also doesn’t provide most of the benefits associated with sex, so it’s thought to be a bad strategy in the long term. This leads to selfing being thought of as a “dead-end” strategy: the short-term advantages make it unlikely that a selfing species will return to outcrossing, and the reduced genetic variation produced by selfing make diversification less likely.

Erik Hanschen and colleagues have tested these predictions in the volvocine algae (I’m among the “colleagues,” as are John Wiens, Hisayoshi Nozaki, and Rick Michod): do selfing species ever return to outcrossing, and do they have a lower rate of diversification than outcrossing species? Both mating systems exist within the volvocine algae, and so they make a good test case. Roughly speaking, the term heterothallic refers to outcrossing species and homothallic to selfing species:

Figure 1 from Hanschen et al. 2017. Diversity of mating systems in the volvocine green algae and their respective life cycles. (A) In outcrossing (heterothallic) species, distinct genotypes (male on left and female on right) sexually differentiate producing either eggs or sperm. A diploid zygospore (red) is produced after fertilization. Sexual offspring hatch and enter the haploid, asexual phase of the life cycle. (B) In selfing (homothallic) monoecious species, a single genotype is capable of producing both gamete types. Upon sexual differentiation, each sexual colony produces both sperm and eggs. (C) In selfing (homothallic) dioecious species, a single genotype sexually differentiates, producing either eggs or sperm, but not both within the same colony. Cartoons in panels (A–C) are shown with anisogamous, Volvox-like morphology for illustrative purposes only.

But this is a case where being happy with ‘rough’ definitions can get us into trouble, especially since they’re used a bit differently in other groups. Heterothallism, at least in the volvocine algae, is basically synonymous with genetic sex (or mating type) determination. In other words, a given genotype of a heterothallic species can produce either sperm or eggs, never both. This leads to obligate outcrossing, because a spheroid can never mate with another of the same genotype.

In homothallic species, a given genotype can produce both sperm and eggs, so it is possible for two spheroids of identical genotype to mate. This situation comes in two flavors, monoecious and dioecious (B & C, respectively, in the figure above). Sexual spheroids of monoecious species are hermaphrodites: a single spheroid produces both sperm and eggs. In dioecious species, a single genotype will produce separate male and female sexual spheroids, the males producing sperm and the females producing eggs. A third flavor of monoecy, not represented in the figure, is androdioecy, in which a single genotype produces both hermaphrodite and male spheroids (for more on this, see “Another step toward understanding sex determination in Volvox“).

So heterothallic species are obligate outcrossers, i.e. sexual reproduction always involves two genotypes, and homothallic species are facultative selfers, i.e. sexual reproduction can involve one genotype (selfing) or two (outcrossing). Does outcrossing ever evolve from selfing? This is predicted to be rare, because the short-term benefits of selfing prevent invasion by an outcrossing genotype. Hanschen and colleagues used ancestral character state reconstructions to test this:

Figure 2 from Hanschen et al. 2017. The evolution of outcrossing (black) and selfing (green). Branch color refers to the most likely state inferred by maximum likelihood (ML) reconstruction. Pie charts at nodes represent scaled marginal likelihoods from ML reconstruction. Numbers at select nodes indicate Bayes factors (support for that character state against the next most likely state), which explicitly take phylogenetic uncertainty into account, colored by which state is most supported. Interpretation of Bayes factors (Kass and Raftery 1995): 0–2 barely worth mentioning, 2–6 positive, 6–10 strong, >10 very strong. Chlamy., Chlamydomonas Astre., Astrephomene Col., Colemanosphaera N., NIES U., UTEX.

Let’s zoom in on one particularly interesting part of that reconstruction. Here’s the group known as “section Volvox” (or sometimes Euvolvox) and its closest relatives:

Within section Volvox, only two species are heterothallic: Volvox perglobator and Volvox rousseletii. Thus it seems likely that the most recent common ancestor of this group was also heterothallic. In this case, Volvox perglobator and Volvox rousseletii must have evolved homothallism, in other words lost genetic sex determination. If, on the other hand, the most recent common ancestor of section Volvox were homothallic, this would imply that several lineages evolved genetic sex determination (heterothallism) independently:

In the first reconstruction, three changes are required: from heterothallism to homothallism near the base of the section Volvox part of the tree, then two changes from homothallism to heterothallism in Volvox perglobator and Volvox rousseletii. In the second, four changes are required, all from the ancestral heterothallic state to homothallism: in Volvox kirkiorum, in Volvox ferrisii, in Volvox barberi, and in the most recent common ancestor of Volvox globator and Volvox capensis.

In this kind of reconstruction, the path that requires the fewest changes is preferred. This is called the most parsimonious reconstruction. It’s preferred because species inherit most of their traits from their ancestors. Some things change most remain the same (a concept I attempted to explain to Cornelius Hunter). Humans, for example, inherited most of our traits from our common ancestor with chimpanzees. Yes, we are bipedal and have less hair and bigger brains, among many other derived traits. But we have many more traits that are inherited: forward-facing eyes, absence of a tail, nails instead of claws, five digits, seven cervical vertebrae, placenta, mammary glands, endothermy, mitochondria, and on and on. All the traits we share with other apes, primates, mammals, vertebrates, animals, and eukaryotes. Change is the exception rather than the rule, so all else equal a reconstruction that requires fewer changes is more likely to be true (Hanschen et al. actually used likelihood and Bayesian methods rather than parsimony, but the basic principle of ‘fewer changes are better’ remains the same).

So at least in this case, the most parsimonious reconstruction has two outcrossing species, Volvox perglobator and Volvox rousseletii, descending from selfing (homothallic) ancestors. If that’s right, it means that the idea that the short-term benefits of selfing prevent invasion by an outcrossing genotype is wrong, or at least not always right. Furthermore, selfing does not seem to prevent speciation along with section Volvox, the clade that includes Pleodorina californica, Pleodorina japonica, and Volvox aureus seem to have diversified from a selfing ancestor.

The long-term persistence of selfing in volvocine algae may be due, at least in part, to their life cycle. Volvocine algae are ‘haploid dominant’, that is, the multicellular phase of their life cycle has only one set of chromosomes. Most of the multicellular organisms we’re used to thinking about are diploid dominant, i.e. they have two sets of chromosomes in the multicellular stage. This is true for animals, for example, where the haploid stage of the life cycle is unicellular and ephemeral: the gametes. In diploid-dominant organisms, selfing can lead to inbreeding depression, a reduction in fitness due to increased homozygosity. A lot of detrimental traits are only expressed when both gene copies are the same, and inbreeding makes this more likely. Self-fertilization is the most extreme version of inbreeding just one generation of self-fertilization means that every gene is homozygous in the offspring. Since homozygosity is not an issue for haploid organisms, it may be that the costs of selfing are lower than in diploids.

There’s another possible reason that the balance of costs and benefits of selfing may be different in volvocine algae, one that we didn’t mention in the paper. We classified homothallic species as selfing, but they are really only facultatively selfing they are perfectly capable of outcrossing if genetically distinct individuals are around when they start mating. In a genetically diverse population, only a tiny minority of matings will be between genetically identical individuals. In this case, the costs of selfing may be low simply because it doesn’t happen that much.

When selfing is likely to happen is when diversity is low. This could well be the case in some situations, for example when a single individual has dispersed to a new pond. Such an individual would likely reproduce asexually for a while, and its progeny would eventually enter the sexual phase. If no other individuals had shown up by then, all of the matings would be between genetically identical (or nearly so) individuals. Nevertheless, a homothallic individual that colonizes a new pond is probably better off than a heterothallic one. When volvocine algae have sex, usually in late summer or fall, they produce a dormant spore that is resistant to cold, desiccation, and other environmental insults (red circles in Figure 1 above). The progeny of a heterothallic colonist would be unable to mate, unable to produce spores, and therefore unlikely to survive the winter.

So the life cycle and ecology of volvocine algae might lead to different costs and benefits of selfing than we’re used to thinking about. A haploid dominant life cycle means that inbreeding is much less of a problem, and the ecological utility of sexually-produced spores might make a homothallic colonist more likely to survive the first year in a new environment than a heterothallic one. Facultative homothallism might provide volvocine algae with the benefits of outcrossing (when other genotypes are around) without the cost of potentially being unable to find a mate. From this perspective, we might want to ask not why homothallism persists in volvocine algae, but rather why it is so rare!

Hanschen ER, Herron MD, Wiens JJ, Nozaki H, and Michod RE. 2017 Repeated evolution and reversibility of self-fertilization in the volvocine green algae. Evolution (doi: 10.1111/evo.13394)

General Overviews

The relative advantages of inbred and outbred mating systems in plants have been discussed since at least the 19th century. Barrett 2010 provides a detailed account of Darwin’s thoughts on mating system. In the 20th century, a series of canonical models for the evolution of self-fertilization arose from the field of population genetics. Charlesworth and Charlesworth 1979 clearly outlines the sequence of important models until the late 1970s, and Jarne and Charlesworth 1993 includes all current major models. Parallel to the development of theory, descriptive science has expanded the corpus of knowledge on the distribution of mating systems among taxa. Fryxell 1957 reviews mating system diversity in depth. Because both selfing and asexual reproduction require only a single parent, it is easy to conflate the two. Holsinger 2000 explains the key differences between selfing and asexual reproduction.

Barrett, Spencer C. H. 2010. Darwin’s legacy: The forms, function and sexual diversity of flowers. Philosophical Transactions of the Royal Society of London B: Biological Sciences 365:351–368.

This high-level review evaluates adaptive explanations of the morphological diversity of flowers. It extensively catalogs different types of floral adaptations, including those that are less well known. The author traces modern discussion of the topic back to Darwin, who explained most floral adaptations as a means of maximizing outcrossing.

Charlesworth, Deborah, and Brian Charlesworth. 1979. The evolutionary genetics of sexual systems in flowering plants. Proceedings of the Royal Society of London B: Biological Sciences 205:513–530.

The authors review the development of concepts in a chronological sequence of canonical models and relate how each built on previous work. This article provides the clearest view available of the historical context of models published previously.

Fryxell, Paul A. 1957. Mode of reproduction of higher plants. The Botanical Review 23:135–233.

This taxonomically broad review classifies angiosperm mating systems based on frequency of cross-fertilization, self-fertilization, and asexual reproduction through seed (see Outcrossing Rate). It gives a sense of the proportions of species practicing these forms of reproduction.

Holsinger, Kent E. 2000. Reproductive systems and evolution in vascular plants. Proceedings of the National Academy of Sciences 97:7037–7042.

This review provides an introduction to the population genetic consequences of mating system in land plants. In particular, the comparison of the effects of self-fertilization and asexual reproduction reveals important symmetries while pointing out potential sources of confusion between them.

Jarne, Philippe, and Deborah Charlesworth. 1993. The evolution of the selfing rate in functionally hermaphrodite plants and animals. Annual Review of Ecology and Systematics 24:441–466.

The authors of this review recommend a reorientation of mating system research toward an empirical focus. They argue that the possible theoretical advantages of selfing and outcrossing have been exhaustively described, but that basic data on selfing rate, inbreeding depression, and inbreeding coefficients are insufficient to decide their importance. The authors also explain the uses and limitations of some of the methods of estimating these parameters.

Users without a subscription are not able to see the full content on this page. Please subscribe or login.


Selfing rates in S. squalidus populations

A frequency histogram of self-fruit-set values for the 191 wild-sampled S. squalidus plants revealed a highly right-skewed distribution of selfing rates (Fig. 2a). The majority of individuals were strongly SI, with mean self-fruit-set per capitulum values of ≤ 2, whilst only six individuals with mean self-fruit-set values per capitulum > 2 but < 25 could be classified as PSC (Fig. 2a). When these PSC individuals were omitted from the frequency histogram of self-fruit-set, the distribution remained highly right-skewed, with the majority of individuals (61.8%) exhibiting zero self-fruit-set (Fig. 2b). Whole population samples were all highly SI, with mean self-fruit-set per capitulum values all < 2. The greater population mean self-fruit-set per capitulum values for Cl, Cw, Kr, Ox and Tt, relative to the other populations, were primarily a result of the presence of PSC individuals in these population samples (Fig. 3). No sample population exhibited significantly higher selfing rates compared with the others after Bonferroni correction for multiple tests of significance (range of P-values from 0.05 to 0.95 α = 0.005 Kr and Bl, respectively Fig. 3).

(a) Self-fruit-set per capitulum values for the entire sample of 191 Senecio squalidus individuals from 11 British sample populations. Pseudo-self-compatible individuals with mean self-fruit-set per capitulum values between 2 and 25 are labelled individually. (b) Self-fruit-set per capitulum values for a subset of 185 confirmed self-incompatible S. squalidus individuals from 11 British sample populations.

Mean self-fruit per capitulum values including and excluding self-compatible individuals for 11 Senecio squalidus sample populations. Diamonds, all plants squares, all SI plants. Bl, Bristol Cl, Carlisle Cw, Crewe Gw, Glasgow Hd, Holyhead Is, Inverness Kr, Kirriemuir Ox, Oxford SB, St. Blazey ST, Stoke-on-Trent Tt, Tamworth.

Inheritance of selfing rates

The strength of SI in the three classes of parental plants (forced-self PSC plants, naturally occurring PSC plants and outcrossed SI plants) followed the same patterns in their progeny. However, self-fruit-set values were considerably reduced in the progeny (Fig. 4). There was a nonsignificant positive correlation between the selfing rates of parental plants and of those their progeny (P = 0.09, Fig. 5). Narrow-sense heritability was low (h 2 = 0.06 Fig. 5) compared with the coefficient of additive genetic variance (CVa = 30.67 Fig. 5).

Mean self-fruit-set per capitulum values for Senecio squalidus plants and their progeny derived from (1) pseudo-self-compatible (PSC) plants generated through forced selfing, (2) naturally occurring PSC plants and (3) outcrossing self-incompatible (SI) plants. Numbers in brackets after class descriptions are sample sizes for parents and progeny, respectively. Standard error bars refer to interindividual variation in self-fruit-set.

Inheritance of selfing rates for seven pseudo-self-compatible (PSC) and five self-incompatible (SI) parental Senecio squalidus plants and their F1 progeny. Mean parental values are presented where the cross involves two different parents. Standard error bars indicate variation between progeny. h 2 , narrow sense heritability calculated according to Equation 1. CVa, percentage additive genetic coefficient of variation calculated according to Equation 2.

Mating system and inbreeding depression

A comparison of nine fitness trait values measured for selfed and outcrossed progeny arrays demonstrated that all except one (days to flower Table 1) were of the expected relative magnitude if the assumption that outcrossed progeny were fitter than selfed progeny holds. A statistical comparison between selfed and outcrossed progeny confirmed that these differences were significant at the 5% confidence level for six of the eight traits (Table 1). These differences in mean measured fitness traits between selfed and outcrossed progeny correspond to inbreeding depression values ranging from −0.03 to 0.38 (days to flower and leaf length, respectively Table 1). Mean inbreeding depression at each life history stage ranged from 0.18 to 0.25 (reproduction and germination, respectively Table 2). A variety of characteristic ‘unfit’ phenotypes were identified within the selfed progeny arrays studied, including ‘dwarfed’, ‘leafy’, ‘twisted’ and ‘floppy’ phenotypes (Fig. 6).

Fitness character Mean inbred value (SE, sample size) Mean outcrossed value (SE, sample size) Two-sample t-test (df) Mann–Whitney U-test Inbreeding depression value (δ)
% seed germination a 96.8 (0.01, 5) 87.6 (0.07, 7) n/a d 0.3845 d 0.09
Days to germinate a 6.45 (0.26, 281) 3.84 (0.12, 213) < 0.001 (385) < 0.0001 0.40
Leaf length (cm) b 10.98 (0.22, 154) 17.57 (0.39, 57) < 0.001 (92) < 0.0001 0.38
Lobes/leaf b 7.78 (0.14, 154) 10.97 (0.45, 57) < 0.001 (66) < 0.0001 0.29
Height (cm) b 44.03 (1.50, 154) 51.56 (2.95, 57) 0.025 (86) 0.0357 0.15
Days to flower b 104.15 (1.86, 62) 107.22 (3.21, 46) 0.410 (74) 0.6343 −0.03
Capitula/primary inflorescence c 7.71 (0.29, 65) 8.92 (0.35, 48) 0.009 (98) 0.0295 0.14
Maximum flowering capitula/day c 22.01 (1.31, 119) 28.00 (3.64, 52) 0.126 (64) 0.9318 0.21
% viable pollen c 38.18 (1.95, 119) 47.82 (3.23, 38) 0.013 (66) 0.0099 0.20
  • Superscripts after fitness traits refer to a germination, b vegetative growth and c reproduction life history stages.
  • d Refers to the significance of differences in germination counts tested using the χ 2 test (1 degree of freedom) only, because sample size was limited by number of progeny arrays.
  • SE, standard error df, degrees of freedom.
Life history stage Mean inbreeding depression value (δ)
Germination 0.25
Vegetative growth 0.20
Reproduction 0.18

A selected range of representative phenotype classes observed in forced-selfed progeny arrays of Senecio squalidus. Plants were photographed at approx. 100 d old when wild-type phenotype progeny were beginning to flower. (a) All crossed progeny and most force-selfed progeny were indistinguishable from the wild-type phenotype presented here. (b) ‘Leafy’ phenotypes tended to developed leaves in place of captula (19% force-selfed A34 progeny). (c) ‘Twisted’ phenotypes had twisted stems and leaf petioles (16% force-selfed Bill progeny). (d) ‘Dwarf’ phenotypes were much smaller than wild-type phenotypes and were often chlorotic and failed to flower (17% force-selfed A34 progeny). (e) ‘Floppy’ phenotypes were an extreme form of the ‘twisted’ phenotypes where the stems could no longer support the leaves in an upright position (26% force-selfed Bill progeny). (f) Range of morphologies observed for mid-cauline leaves among the force-selfed and cross progeny arrays. The wild-type is sixth counting from the left.


Most flowering plants that are predominantly self-pollinating have an annual life history [1–3]. Interpretations of this association usually involve one of two main hypotheses. (i) Compared with perennials, annuals may generally accrue greater fitness benefits from selfing through 'reproductive assurance', i.e., because ovules may be generally more outcross-pollen-limited and/or pollen grains may be more outcross-ovule-limited [2, 4–8]. (ii) Perennials may incur a higher fitness cost of selfing through seed discounting and inbreeding depression hence, possibly most selfers are annuals simply because relatively few perennials can be selfers [9, 10].

A recent third hypothesis, the 'time-limitation' hypothesis, predicts that both selfing and the annual life cycle are concurrent products of strong 'r-selection' associated with high density-independent mortality risk in ephemeral habitats with a severely limited period of time available to complete the life cycle [11]. Both the traditional reproductive assurance hypothesis and the time-limitation hypothesis involve a fitness advantage for selfing through ensuring that at least some reproduction occurs, but they involve very different selection mechanisms – pollinator/mate-limitation (where outcross pollen is not available at all due to a lack of pollinators or mates), versus time-limitation (where outcross pollen is available but arrives too late to allow sufficient time for development of viable seeds). Accordingly, these two hypotheses for selfing involve very different assumptions and predictions.

The time-limitation hypothesis has direct and indirect components. The indirect component predicts higher selfing rates in annuals as a trade-off of selection for earlier reproductive maturity in annuals [12, 13] (Figure 1a). More rapid floral maturation is expected to result in smaller flowers with increased overlap of anther dehiscence and stigma receptivity in both space (reduced herkogamy) and time (reduced dichogamy) thus, increasing the frequency of selfing as an incidental consequence [12] (Figure 1a). If selfing also shortens the time between flower maturation and ovule fertilization, then higher selfing rates for annuals in time-limited habitats may also be predicted as a direct fitness benefit abbreviating the time between anthesis and ovule fertilization may ensure that there is enough remaining time in the growing season (after ovule fertilization) to allow complete seed and fruit maturation [11] (Fig 1b). Selection favors selfing here by favoring increased overlap in anther dehiscence and stigma receptivity in both space and time, which are in turn facilitated by smaller flower size and shorter flower development time, respectively (Figure 1b).

Two components of the 'time-limitation' hypothesis for the evolution selfing in annuals. In (a), selfing is a trade-off of selection favoring a shorter time to reproductive maturity (fully developed flowers) under strong r-selection. As a tradeoff (dashed arrows), flowers become smaller with greater overlap in location and timing of anther dehiscence and stigma receptivity, thus increasing the rate of selfing as an incidental consequence. In (b) strong r-selection favors a shorter pollination time directly i.e., selfing is selected for directly because it shortens the amount of time between flower maturation and ovule fertilization, thus leaving sufficient remaining time for seed and fruit maturation before the inevitable early mortality of the maternal plant under strong r-selection. In this case, smaller flower size and shorter flower development time are favored by selection because they facilitate selfing (see text).

However, the two components of time-limitation cannot be separated clearly, as they operate simultaneously i.e., earlier onset of flowering, shorter flower development time, smaller flowers and selfing can all be interpreted to have direct fitness benefits because they may all contribute directly to accelerating the life cycle [11]. Indeed, time-limitation associated with strong r-selection would be expected also to favor an acceleration of the final stage in the life cycle – seed/fruit development time (Figure 1) – resulting, as a trade-off, in smaller seeds and/or fruits [11].

The time-limitation hypothesis remains untested. Some recent studies have explored the rapid growth and maturation time of annuals in terms of bud development rates and ontogeny [13–15]. However, these studies have compared growth and development rates between selfing and outcrossing populations of only a single species. Since their effective sample size is only one, this makes it difficult to extrapolate the predominant selection pressures that may have promoted the general association of selfing with the annual life cycle.

The objective of the present study was to compare, for annuals exclusively, life history traits associated with selfing versus outcrossing using several species from a wide range of plant families. Phylogenetically-independent contrasts (PIC) were used to control for confounding effects due to common ancestry among species [16]. Using a database of 118 species involving 14 families, plant size, flower size, and seed size were compared between selfing and outcrossing annuals. The time-limitation hypothesis predicts that all of these traits should be smaller in selfing annuals because the severely time-limited growing season that promotes selfing also imposes an upper limit on the maximum sizes that can be attained for plant traits [11] (Figure 1). The trend for outcrossers to be taller, and have larger flowers and larger seeds has often been noted [1, 17–19]. We used a multi-species, across-family comparison, however, to investigate whether this trend also holds true within annuals exclusively.

Data on the timing of life history stages (i.e. age at first flower, bud development time, and flower longevity) were also obtained from a greenhouse study of 25 annual species involving 5 families. The time-limitation hypothesis predicts that selfers should produce mature flowers more quickly and should have shorter flowering times.


I thank my advisor, L Delph, for advice throughout the planning and completion of this project. In addition, I thank my committee members J Bever, C Lively, and M Wade for their comments and suggestions. I Anderson offered valuable help and advice during the location of populations in northern Alabama. I thank C Fuzzell, E Speck, T and M Winchell, and the Willis family for access to their private property. I thank R Davis for work during the experiment. Comments from J Koslow, T Linksvayer, J Moorad, M Neiman, D Waller, and two anonymous reviewers considerably improved the manuscript. This work was supported by Indiana University, the National Science Foundation, Sigma Xi, and the Indiana Academy of Science.


  1. Chayson

    I think you are wrong. I'm sure. I can defend my position. Email me at PM.

  2. Sayad

    Quite right! It is excellent idea. It is ready to support you.

  3. Cuetzpalli

    I can recommend that you visit the site, with a huge number of articles on the topic that interests you.

  4. Sumner

    This message, amazing))), very interesting to me :)

  5. Killian

    Here it is yes!

  6. Arale

    There was a mistake

Write a message