ANALYSES SO FAR
Fig. 1 (A-G). Phenotypic Transition in Populations
This map represents the dynamic floral color transitions in red, mixed, and yellow populations between 1996 and 2016. Mixed plant populations were formerly assigned to the hybrid origin, while the present study states that the transitional stages are the consequences of a dynamic progression of single plants that begin with the yellow-flowered morph and consistently ends with the red floral shape. Thus, the coastal red morph D. puniceus has also become dominant in the East and North, whereas the former hillside and mountain morph of yellow-flowered D. australis is nowadays confined to the central Southern and Northern edges of San Diego County. The border extends at a speed between 300 and 800 m per year (arrows). The map is based on the monitoring of 70 populations over 20 years. Three inlets show the corresponding flower shape types.
Red circles: pure red-flowered D. puniceus populations of morph type depicted in Fig.1G; yellow circles: pure yellow populations D. australis of morph type shown in Fig; 1B; orange circles: populations in a transition stage with morph types depicted in Fig.1 C to E; dotted line: yellow boundary 1962; bold dotted line: yellow boundary 1996; dashed line: yellow boundary 2016; B1 … B6: sites along a 60 km long transect.
(B-G) Show six flower shapes representing mean values of sites B6 to B1 along a transect (more see movie S1).
Monitoring Plants in Transition
(A) The twenty plants were monitored once each year in San Diego and Riverside Counties, where color development was recorded between 1996 and 2008. A yellow individual from Campo (Y) and a red plant from Point Loma (R) show stable color readings during the whole period of observation. Both were plants growing well outside any transition zones on the extreme ends of their distribution zones.
Twenty plants (1-20) within the transition zones show the characteristic up-and-down development in the color transition of their color readings, more or less progressively tending towards orange and red tones.
(B. D) The transitioning plant, T27 & T30, have been monitored since 2002 in Ballena Valley, San Diego County, CA. In the field, the flower color steadily progressed from light yellow-orange to red (C & E).
Cutting from these plants monitored for 13 years under cultivar in Zurich. No phenotypic transition did occur in this plant under controlled laboratory conditions.
Fig. 3 (A-B). Heredity of the flower color of siblings of two different origins
The figure shows a comparison of plots of crossed red-flowered siblings (R3xR7 = Rc) from plants sampled at the coast against red-flowered siblings sampled 15 km inland (R14xR15 = Ri). The inland results show significantly more variation and demonstrate more extreme values compared to the coast values. The reversion rate from red to yellow state is between 1.2 and 4.8 % see* that lies well outside of the ±2SD range. In Fig.1, Rc and Ri are plotted on the map together with the transect.
Figure S1. Floral Color Monitoring in «Rainbow» & «Rancho» Populations
Monitors two transient populations by sampling every five to eight years. Sample selection was repeated in two populations with more than 250 individuals geographically spaced 15 km apart, both unmarked and in different plant individuals.
The color distribution in 2001 for Rainbow and 1997 for Rancho was broad and inclined to the yellow side. After that, the median color shifted permanently to 9.0. The variance (S 2) is lowest at the last census, missing yellow hues that were obvious 15-20 years ago. The result at both sites is a very similar floral color transition dynamics in populations within the transition zone. The censuses of both populations do not differ significantly (p=0.18) in 2016, and they converge to the same final state of more or less uniform red morphs.
Welch two-sample t-test. MCU = Methuen Color Unit.
Figure S2 (A-B). Age Structure at Rancho population
(A) The figure shows the age structure in a specific population of over 250 plants of the softwood shrub Diplacus aurantiacus situated at Rancho California Rd. The study used a sample survey along a 60 m long transect line with a maximum of 1 m lateral deviation.
The age structure is based on the number of tree-rings (N=32). It can be deduced that the average lifespan in Diplacus is 31 years, a possible lifetime of 10 to 70 years.
(B) Depicts the cross-section of a 23-year-old shoot; one can see more or less pronounced tree rings.
Figure S3 (A-F). Correlation of floral traits along a transect from coast to inland
The clinal variations of the leading floral features along with a transect line (see Figure 1) are shown. A nearly linear proportional correlation supported by the regression line is the corolla length (A) and the corolla width (B) with the distance to the coast. The positive correlation is appropriate (0.73 ... 0.76). Three more floral features like pedicel length (C), flower angle (D), style length (E) have a reasonable correlation (-0.954 ... 0.91). The Methuen flower color (F) is split into two sections. It rises sharply 40-45 km off the coastline but remains constant some 30 km off the coast at the MCU 9, red.
"H" refers to the "transition zone" with showy, differently colored flowers in populations. Differently colored flowers in populations have hitherto been used as a marker for assignment to a hybrid zone.
Figure S4. Inheritance at the individual level of two plants in transition
In a greenhouse experiment, the color morphs of two individuals in transition were tracked using inbred lines of cuttings harvested in the field. The cuttings of these individuals were taken twice from plants in Ballena Canyon (San Diego County) within 14 years.
The development of the individual cutting patterns was followed separately (see above the different shades). The overall broadly colored flower color variation of both progeny lines developed significantly towards redder shades during the observation period.
Figure S8. The floral Color variation on individual plant
The color deviation of the flowers of seven plants is applied. The standard deviation of plants in transition (T) is twice as high (0.16) as that of stable yellow or red plants (Y4, RbA, 0.08). This amount is relatively small compared to the annual average color increase of 0.3 to 0.4 MCU of plants in transition.
Figure S9 (A-B). Santa Catalina Island Herbarium specimen and field investigations
Shows a chronological list of Herbarium specimens of Diplacus species collected on the island of Santa Catalina, from which information about the flower color can be found on the labels. Around 1920, the form was also called Diplacus linearis, which corresponds to various yellow-flowered D. longiflorus. In the 40s and 50s of the last century, shapes with yellow-brownish tones predominated. End of the 90s, the flower color was consistently indicated as orange-red.
Today the plant occurs predominately on the less dry eastern slopes of the island. It is remarkably uniform and has a flower shape that distinguishes it on the one hand from the coastal morph D. puniceus and also from that on Santa Cruz Island (D. parviflorus or flemingii).
(Inlet). It is orange-red throughout the island, with two yellow exceptions with short pedicel, and can occur in dense populations in the open habitat. Hummingbirds pollinate it. This taxon should have its rank, e.g., «Diplacus cataliensis.» In summary, within just 70 years in the last century, the progression of this species has been unnoticed by science.