Pollination syndromes are a key component of flowering plant diversification, prompting questions about the architecture of single traits and genetic coordination among traits. Here, we investigate the genetics of extreme floral divergence between naturally hybridizing monkeyflowers Mimulus parishii (self-pollinated) and M. cardinalis (hummingbird-pollinated).

We mapped quantitative trait loci (QTLs) for 18 pigment, pollinator reward/handling, and dimensional traits in parallel sets of F₂ hybrids plus recombinant inbred lines and generated nearly isogenic lines (NILs) for two dimensional traits, pistil length, and corolla size.

Our multi-population approach revealed a highly polygenic basis (n = 190 QTLs total) for pollination syndrome divergence, capturing minor QTLs even for pigment traits with leading major loci. There was significant QTL overlap within pigment and dimensional categories. Nectar volume QTLs clustered with those for floral dimensions, suggesting a partially shared module. The NILs refined two pistil length QTLs, only one of which has tightly correlated effects on other dimensional traits.

An overall polygenic architecture of floral divergence is partially coordinated by genetic modules formed by linkage (pigments) and likely pleiotropy (dimensions plus nectar). This work illuminates pollinator syndrome diversification in a model radiation and generates a robust framework for molecular and ecological genomics.

All hybrids in this study were generated from two highly inbred parental lines: Sierran CE10 for Mimulus cardinalis (Yuan et al., 2013) and PAR for M. parishii (Fishman et al., 2015; Nelson et al., 2021; Liang et al., 2023). F₁ hybrids were generated with PAR as the seed parent and selfed to generate F₂ seeds, while recombinant inbred lines (RILs) were generated by single-seed-descent from F₂ individuals through 3-6 generations of self-fertilization (Sotola et al., 2023).

Plant growth conditions for QTL mapping - F₂ hybrids were grown in two separate greenhouse common gardens at the University of Connecticut (UC) and the University of Montana (UM). At UC, seeds (CE10, PAR, F₁ hybrids, F₂ hybrids, and NILs) were sown into 98 cell trays (1.25"sq x 2"H), and then after four weeks, all seedlings were transferred to 18 pocket trays (3.25" sq x 3.14"H) and grown to flowering.  At UM, F₂ seeds (and a few CE10 and PAR controls) were germinated on wet sand in Petri dishes, seedlings transplanted into 3” square pots filled with Sunshine #4 soil-less potting mix, and plant grown to first flower and beyond (life history and plant architecture traits presented elsewhere). The RIL mapping population was grown at the University of Georgia (UGA) with seeds sown into 2.5” pots filled with Metromix 830 soil. Seedlings were transplanted into 4” pots and grown to flowering. In all three greenhouses, plants were grown under natural light supplemented with sodium lamps to a 16‐h daylength and fertilized regularly (UC: 2-3 times per week, UM and UGA: weekly).

Phenotypic measurements for QTL mapping

In the UC_F₂ growout, we measured three pigment, two pollinator reward/handling and nine dimensional floral traits, plus flowering time, on F₂s (n = 253) plus parental lines and F₁ hybrids (n =8 each) (Fig. 1). We scanned the ventral petal of each flower to quantify petal lobe anthocyanin (PLA) and carotenoid (PLC) pigment intensity. The proportion of red (R), green (G), and blue (B) pixels in a square area of the same size of the adaxial surface of the petal were estimated from scanned images using Image J (http://rsbweb.nih.gov/ij/) and the relative petal lobe anthocyanin and carotenoid concentrations estimated by [(R + B)/2] − G and [(R + G)/2] – B, respectively. Nectar guide anthocyanin (NGA) and nectar guide trichome length (NGT), were visually scored in F₁ and F₂ hybrids on semi-quantitative scales defined by the parental extremes (CARD = 9 and 7, respectively, PAR = 1 for both). Nectar volume (NEV) was measured for two flowers per individual on their first day of opening, using a pipette accurate to 1.5μL. To reduce environmental effects, nectar was measured at 4:00 PM-7:00 PM after watering at 12:00-1:00PM each day. Floral dimensions (Fig. 1c) were measured on one of the second pair of open flowers using a digital caliper, and stigma-anther distance calculated as pistil length – stamen length. 

In addition to pollen number and viability (Sotola et al. 2023) and life history traits to be presented elsewhere, the UM_F₂ grow-out measured 9 traits shared with the UC_F₂ growout: flowering time, two floral pigment traits (PLA and PLC), and five floral dimension traits (corolla length and width, corolla tube width, and pistil and stamen length) and calculated stigma-anther separation. PLA and PLC were measured and calculated as described above, except that we scanned one of the two lateral petals. The standard corolla dimension traits (Fishman et al., 2002) were measured with an engineering ruler on one flower of the first pair on the day it opened.

In the RIL growout at UGA (n = 145 RILs, 18 each of CE10 and PAR parents and F₁ hybrids), we recorded day to first flower (flowering time) and measured two floral pigment traits (PLA and PLC), nectar volume, and five floral dimension traits (corolla length and width, corolla tube width, and pistil and stamen length) and calculated stigma-anther separation. We measured floral pigment traits on the ventral petal as described above and used an engineering ruler to measure the floral dimensional traits on one flower from the first pair. To estimate nectar volume, we used a capillary tube to collect nectar from two flowers on their first day of opening and took the average; nectar was measured between 1:00 PM - 3:00 PM after watering in the morning at 8:00 AM - 9:00 AM. Other than pistil length (PIL), which includes the stigma lobes in the UC_F₂ and UGA_RIL measurements but not in the UM_F₂ dataset, and thus stigma-anther separation (SAS), traits with a shared abbreviation represent the same floral dimension.

• Fishman L, Beardsley P, Stathos A, Williams CF, Hill JP. 2015. The genetic architecture of traits associated with the evolution of self-pollination in Mimulus. New Phytologist 205: 907–917.
• Fishman L, Kelly AJ, Willis JH. 2002. Minor quantitative trait loci underlie floral traits associated with mating system divergence in Mimulus. Evolution 56: 2138–2155.
• Liang M, Chen W, LaFountain AM, Liu Y, Peng F, Xia R, Bradshaw HD, Yuan Y-W. 2023. Taxon-specific, phased siRNAs underlie a speciation locus in monkeyflowers. Science 379: 576–582.
• Nelson TC, Stathos AM, Vanderpool DD, Finseth FR, Yuan Y-W, Fishman L. 2021. Ancient and recent introgression shape the evolutionary history of pollinator adaptation and speciation in a model monkeyflower radiation (Mimulus section Erythranthe). PLoS Genetics 17: e1009095.
• Sotola VA, Berg CS, Samuli M, Chen H, Mantel SJ, Beardsley PA, Yuan Y-W, Sweigart AL, Fishman L. 2023. Genomic mechanisms and consequences of diverse postzygotic barriers between monkeyflower species. Genetics 225: iyad156.
• Yuan Y-W, Sagawa JM, Young RC, Christensen BJ, Bradshaw HD. 2013. Genetic dissection of a major anthocyanin QTL contributing to pollinator-mediated reproductive isolation between sister species of Mimulus. Genetics 194: 255–263.