Lynda F. Delph

Professor of Biology

Department of Biology, Jordan Hall, 1001 E. Third Street,
Indiana University, Bloomington, IN 47405 USA
Office phone: 812-855-1831, Biology Fax: 812-855-6705, email: ldelph@indiana.edu

DEGREES
1979          B.S. (Honors), University of Arizona
1983          M.S., University of Arizona
1988          Ph.D., University of Canterbury
APPOINTMENTS
1989          Busch Postdoctoral Fellow at Rutgers University with Dr. T. Meagher
1989-96    Assistant Professor, Department of Biology, Indiana University
1996-02    Associate Professor, Department of Biology, Indiana University
1998-13     Associate Chair, Department of Biology, Indiana University
2001-        Senior Fellow, Indiana Molecular Biology Institute
2002-       Professor, Department of Biology, Indiana University
FELLOWSHIPS AND AWARDS
1983-84    Fulbright Fellowship
1994-95    Outstanding Junior Faculty Award, IU
1995          Senior Class Award for Teaching Excellence in Biology, IU
1997          Fulbright Fellowship
2000        Teaching Excellence Recognition Award, IU
2005         Trustees’ Teaching Award, IU
2005         Guggenheim Fellowship
2010         Fellow, American Association for the Advancement of Science

 

I am interested in evolutionarily based questions concerning adaptation, reproductive strategies, and speciation in plants. This involves looking at morphology, genetics, physiology, life histories, how they are interrelated and how they affect fitness. My work includes field, greenhouse, and lab work, and I use a combination of manipulative experiments, comparative approaches, artificial selection, and genetic approaches. Past research includes: (1) an investigation of the forces that select for gender dimorphism, and how this separation of gender affects other plant traits, (2) the evolution and quantitative genetics of sexual dimorphism, (3) how inbreeding affects inbreeding depression within populations, (4) how plant vigor affects male fitness, (5) mating-system evolution, (6) how traits such as flower size and number affect plant fitness, and (7) the evolution of sex chromosomes.

Most recently, my research has revolved around three main lines of investigation, and numerous smaller projects. Note that my graduate students often pursue their own research questions using their own systems (see the People page for the titles of past graduate student theses and the Organisms page for pictures of their study systems).

HALDANE'S RULE - We investigated whether Haldane's rule works in plants. Three closely related dioecious Silene species (S. latifolia, S. dioica, and S. diclinis) make a great system to test this in as all three have XY males, all three are crossable, and all flower relatively quickly. We completed a crossing experiment between all three species pairs and have found conformance with Haldane's rule for both hybrid male rarity and sterility (Brothers and Delph 2010 Evolution). Crosses between S. latifolia and S. diclinis show the most extreme example of the rule, so we peformed 15 types of crosses, including reciprocal crosses and later-generation crosses, to partition out which components of the genome are contributing to the incompatibilities (Demuth et al. 2014 Evolution). We found that male sterility was caused by X-autosome incompatibilities, which implicates recessive hemizygous alleles and is in accordance with the Dominance Theory for Haldane's rule. In contrast, the rarity of male hybrids was in accord with Faster-male evolution and the presence of the neo-sex chromosomes of S. diclinis.

Given the existence of Haldane's rule in crosses between S. latifolia and S. diclinis, we also investigated whether reinforcement was operating to prevent the production of unfit hybrids in sympatry. We performed intra- and inter-specific crosses using 7 allopatric and one sympatric popualtion of S. latifolia, followed by an experiment that varied the distance pollen had to travel down the style, and found that style length, rather than sympatry per se, confered reproductive isolation: whenever the S. latifolia mothers had long styles (i.e., large flowers), S. diclinis pollen was unable to achieve fertilization.

This work was in collaboration with my former PhD students Amanda Brothers and Phil Nista, my former postdoc Rebecca Flanagan, and Jeff Demuth (from the University of Texas at Arlington).

SEXUAL DIMORPHISM - One ongoing project concerns the evolution of sexual dimorphism, the underlying genetics of sexual dimorphism, and sex-chromosome evolution in the dioecious species, Silene latifolia. The two sexes of this species have remarkably different life histories, which appear to be influenced by the sexual dimorphism in flower number - males make 17 times as many flowers as females over the same time period. Although females have a greater reproductive effort because of the high cost of producing fruit, males have higher costs of reproduction in the life-history trade-off sense (Delph and Meagher 1995 Ecology). By artificially selecting on flower number (to both increase and decrease the sexual dimorphism) we have uncovered correlated responses in morphological and physiological traits that have given us insights into why males pay a higher cost of reproduction and why males make smaller flowers than females (Delph et al. 2004 Evolution, Delph et al. 2005 Am Nat).

Silene latifolia - Pistillate flower from a female on the left and staminate flower from a male on the right.

A QTL-mapping project was performed, which involved crosses between the divergent selection lines to investigate the genetic architecture of sexually dimorphic traits.  We found that several QTL of major affect are only expressed when they are on the Y-chromosome, an unexpected and exciting result (Scotti and Delph 2006 Evolution). Sex-specific loci control more of the phenotypic variation in sexually dimorphic traits than loci that are expressed in both sexes, suggesting some resolution of sexual conflict. Moreover, results from the QTL study are congruent with our quantitative genetic (G-matrix studies), showing that traits within males are more genetically integrated than those in females (Delph et al. 2010 Evolution).

We planted artificial-selection lines out in the field to investigate how the divergent lines acquire fitness via two pathways - the pathway involved with mate acquisition (sexual selection) and the pathway involved with life history (fecundity and viability selection). This experiment revealed that sexual selection favored males that made lots of flowers early in the season, as this corresponded with relatively high fruit production by the females. However, making many flowers per se was not sexually selected in males, as it did not lead to a disproportionate siring success via increased attractiveness to pollinators - siring success was simply proportional to pollen production. Seed production by females was higher in plants that made relatively large flowers, and hence relatively few flowers. Both males and females that flowered extensively paid a cost in terms of longevity. Hence, the different forms of selection - sexual, fecundity, and viability - were in opposition in the males (sexual and fecudity selection favored males that made many small flowers, and viability selection favored males that made relatively few flowers) but not in the females. The forms of selection also operated differently between the sexes. These results go a long way toward explaining why the sexes are sexually dimorphic for flower size and number (Delph and Herlihy 2012 Evolution).

We performed a cross-classified breeding design to estimate the G matrix for both males and females. We found that the two sexes have different G matrices and that traits are highly intercorrelated (Steven et al. 2007 Evolution). This means that even if selection were the same in both sexes for certain traits sexual dimorphism would evolve and could include traits not directly under selection.


We also undertook two additional artificial selection experiments to investigate whether genetic correlations can be broken by correlational selection (Delph et al. 2011 Evolution). The results of one of these experiments, in which we tried to break the between-sex genetic correlation for calyx width, revealed that the correlation could be broken in only 4 generations. We followed this up with an experiment to test whether the correlation was actually broken by selecting only on one sex and seeing if the other sex responded. If we really did break the correlation, then little or no response should be seen in the sex not under selection, which is exactly what we found. We are currently evaluating whether breaking the between-sex correlation for calyx width resulted in other changes to the G matrix (Steven and Delph in prep.).


We investigated among-population differentiation in morphological and physiological traits that are sexually dimorphic via phenotypic selection analysis in natural populations in Europe that differ in floral and leaf traits (Delph et al. 2011 New Phytologist). We found that specific leaf area (a measure of leaf thickness that impacts leaf physiology and longevity) was under selection in hot, dry habitats, but only in males (selection against males with thin leaves). Combined with a G-matrix study, this provides evidence for intralocus sexual conflict for specific leaf area. This finding fits our results from an among-population crossing study, which revealed that males differ more among populations than do females as a result of selection (Yu et al. 2011 Journal of Evolutionary Biology).

We are now looking at the intersection of sexual dimorphism, sex-specific selection, sexually antagonistic selection, local adaptation, and whether these factors impact sex-chromosome evolution (currently funded by NSF). We have a reciprocal-transplant experiment going on in the field, to investigate local adaptation in Spain (hot and dry) and Croatia (cool and wet). We have planted pure-population-of-origin plants in field sites in both countries, as well as F2s. We plan to conduct phenotypic selection gradient analysis on these plants over the next two summers, and follow that with a QTL study of the F2s to determine whether sexually antagonistic traits conferring local adaptation map to the pseudoautosomal region of the sex chromosomes. If they do, this will implicate sexual antagonism in the evolution of sex chromosomes.


To date, several of my postdocs (Steve Carroll, Jan Gehring, Maureen Levri, Michele Arntz, Janet Steven, Ivan Scotti, Chris Herlihy, Daniela Bell, Ingrid Anderson, and Maia Bailey), two of my graduate students (Frank Frey and Laura Weingartner), and around 21 undergraduates have worked with me on various aspects of this system. The correlational selection work was in collaboration with my former colleague Butch Brodie (now at UVA) and the among-population crossing study was in collaboration with my colleague Mike Wade.


Two flowers from males of Silene latifolia are pictured here - the one on the left is from the selection line
for small flowers and the one on the right is from the selection line for larger flowers.

 

GYNODIOECY - Another avenue of investigation involves the study of gynodioecy, in which both females and hermaphrodites coexist in populations. We have focused on costs of restoration, using both simulation models and experiments. We have also investigated the evolutionary dynamics of nuclear-cytoplasmic gynodioecy by investigating mitochondrial gene-sequence polymorphism.
    In the past, our main natural study species for our work on gynodioecy was the cushion plant Silene acaulis, which is circumpolar in the northern hemisphere in arctic and alpine habitats. Our field work on this species took place in the mountains of Colorado and in northern Finland (furthermore, we are growing plants from all over the world, including Alaska, Colorado, Iceland, Greenland, Norway, and Finland in the greenhouses at IU).
    We started out using S. acaulis to evaluate mechanisms that might be responsible for why outcrossed seeds from female seed-parents outperformed those from hermaphrodite seed-parents. Once we had shown that seeds from females are not better provisioned than those from hermaphrodites (Delph et al. 1999 Am J Bot), we turned to alternative hypotheses. These hypotheses (which aren't mutually exclusive) deal with pollen-tube competition, pleiotropic affects of genes, and biparental inbreeding. Results indirectly support the idea that nuclear restorer alleles have negative pleiotropic affects (or costs of restoration) (Delph and Mutikainen 2003 Evolution).
    Past work with this species also includes an investigation of population structure, which revealed positive spatial autocorrelation within 1-3 m, and panmixia past that (Gehring and Delph 1999 Heredity).
    We found remarkable sequence polymorphism in a mitochondrial gene of this species (22 substitutions over 3 geographic regions), and these are combined within a small number of haplotypes per population, indicating that either selection or demographic forces have perpetuated a small number of very ancient haplotypes (Städler and Delph 2002 PNAS). Pascal Touzet (University of Lille) and I followed up on this line of investigation in other Silene species, as this level of polymorphism is unprecendented. We found that sequence polymorphism in mitochondrial genes is abundant in gynodioecious species, but absent or nearly absent in hermaphroditic or dioecious species (Touzet and Delph 2009 Genetics). We have followed this up by looking at similar genes in various Lobelia species that vary in the frequency of females (from 0 to 60%). Remarkably, mitochrondrial haplotype diversity is significantly correlated with the frequency of females (Delph and Montgomery 2014 International Journal of Plant Sciences.). This finding strengthens the hypothesis that diversity is maintained by balancing selection.
    We have also worked on Brassica napus with regard to the cost of restoration. We made use of the fact that individuals can be characterized as having a male-fertile cytoplasm or one of two male-sterility genes together with one of two restorer genes. In other words, we used individuals whose genotype is known for the mitochondrial genes causing male sterility and the nuclear genes restoring male sterility and could therefore directly measure the cost of the restorers. One of the restorers, Rfp, should act in a silent manner, while the other, Rfn, should act more constitutively. We have compared the cost of these two types of restorers and found that Rfn carries a higher cost of restoration (Montgomery, Bailey, Brown, and Delph in press).
    To date, several of my postdocs (Maia Bailey, Steve Carroll, Jan Gehring, Ben Montgomery, Pia Mutikainen, Molly Nepokroeff, and Thomas Städler), three of my graduate students (Debbie Marr, Maia Bailey, and Dana Dudle), and several undergraduates have worked with me on various aspects of gynodioecy.

  
Silene acaulis from Pennsylvania Mtn. in Colorado - pistillate flowers from a female plant on the left and perfect flowers from a hermaphrodite plant on the right.

Back to top