Minimal Loss of Genetic Diversity and No Inbreeding Depression

ORIGINAL ARTICLE

SUPPLEMENTARYMATERIAL (ONLINE)

Minimal loss of genetic diversity and no inbreeding depression

in blueflag iris (Iris versicolor) on islands in the Bay of Fundy

Nathaniel T. Wheelwright, Elise Begin, Claire Ellwanger,

Samuel Taylor, and Judy Stone

MATERIALS AND METHODS

Greenhouse protocols and physiological measurements - F1 plants were allocated positions within three blocks, each with 4 trays of up to 8 seedlings per tray. Within each tray every seedling represented a unique capsule x treatment x clone combination; seedlings from the South and West Beach study sites were laid out in a checkerboard pattern within which pollination treatments were randomized. Blocks were positioned on separate benches and rotated every 4-6 weeks.

Fluorescence was quantified using a LI-6400 photosynthesis system (LI-COR Biosciences, Inc., Lincoln, NE) equipped with a CO2 injector (LI-6400-01) and a combined light-source and fluorometer (LI-6400-40). Cuvette CO2 levels were maintained at 400 µmol mol−1 and block temperature was set at 24 °C. For measurements of Fv/Fm, which were made between 0800 and 0930 h, plants were placed in a closed cabinet overnight prior to measurements. Measurements of ΦPSII were made between 1200 and 1500 h using a photosynthetic photon flux density (PPFD) of 200 µmol m−2 s−1 (consistent with light levels commonly experienced by the leaves during daylight hours). Measurements were made on the youngest available leaves that had stopped expanding, usually the fifth or sixth youngest leaf.

Inbreeding depression: measurements and statistical analyses -Means and 95% intervals for statistics were obtained from 1000 simulations of fitted models (arm function sim), and the significance of the fixed effects treatment × site, site, and treatment were determined sequentially, using likelihood ratio tests (LRT) to compare models fit by maximum likelihood. We also incorporated random effects of treatment × clone (LRT P ≥ 0.30 [excluding the within-site treatment, all P ≥ 0.43 except proportion germinating, P = 0.045]) and clone (LRT P ≤ 0.001 [excluding the within-site treatment, P ≤ 0.01]) or, in greenhouse experiments, block. (In greenhouse experiments treatment × site was synonymous with treatment × clone because only one maternal clone was included from each site).

Relative performance (RP; Ågren and Schemske, 1993) = 1-ws/wo if wswo; RP = wo / ws -1 if wswo, where ws is the mean fitness of selfed individuals and wo the mean fitness of outbred individuals.

Tissue sampling and genotyping -DNA was extracted from dried leaf tissue using the QiagenDNeasy 96 Plant Kit (Valencia, California). PCR amplifications were performed in a volume of 25 µL using a GeneAmp PCR System 9700 (Applied Biosystems, Carlsbad, California). Reactions for tet4 and tet9 contained 1.5 mM MgSO4, 0.8 µM dNTPs, 0.5 µM each primer, 1.25 U Taq DNA Polymerase, and 10 ng genomic DNA. Reactions for IM200 differed in containing 2.0 mM MgSO4. Forward primers were labeled with a 6-FAM fluorescent tag. Cycling conditions for tet4 and tet9 consisted of an initial 2.5 min denaturation at 95°C followed by 35 cycles of 20 sec denaturation at 95°C, 30 sec annealing at 60-50°C, a 30 sec extension at 72°C, and a final 10 min hold at 72°C. Annealing temperatures started at 60°C and decreased by 0.5°C per cycle until the final annealing temperature of 50°C was reached. Cycling conditions for IM200 differed in that annealing temperatures during the touchdown phase decreased from 63°C to 57°C in 1°C increments, followed by 33 cycles at the final annealing temperature. PCR products were suspended with ROX 500 size standard in deionized formamide before loading onto an Applied Biosystems 3130 Genetic Analyzer.

Power calculations -Given the expectation of little or no inbreeding depression, as in most predominantly selfing species and as shown in previous studies (Wheelwright et al., 2006, Zink and Wheelwright, 1997), we used power calculations to estimate how many samples would have been required to detect the (subtle) effect expected. Fig. S1 shows the relationship sample size and two-tailed confidence intervals based on the t-distribution and standard deviations back-calculated from 95% confidence intervals for RP across the South and West Beach sites (Table S3). The upper of the two panels represents traits measured in greenhouse experiments where samples were smaller, and the lower panel measurements of seed and reproductive traits that drew from a larger pool of maternal genotypes. A two tailed confidence interval is considered because we used Relative Performance (RP; Ågren and Schemske, 1993) to assess inbreeding (where ws is the mean fitness of selfed individuals and wo the mean fitness of outbred individuals: RP = 1-ws/woif wswo; RP = wo/ws-1 if wswo) explicitly allowing for the possibility that differences in performance could also be indicators of outbreeding depression. Dark gray shading indicates fitness effects from RP = 1 to RP = 0.95 (or 1.05) (0-0.05 on the y-axis). The lighter gray fill extends to RP 0.9 (or 1.1; 0.1 on the y-axis). Mean RP of a trait determined for a given sample size would need to fall beyond the confidence interval corresponding to the standard deviation of the trait. Vertical dashed lines indicate how sample sizes per treatment in experiments with I. versicolor (approximated as ¼ of the total sample sizes from Table 1) compare with the confidence intervals that could be obtained. It is clear that in greenhouse experiments (upper panel), sample standard deviations were sufficiently high that RP in the 1-0.9 range would rarely be detected as significant: to attribute statistical significance to RP in the 1-0.95 range would have required tripling of sample sizes. For the majority of seed and reproductive traits (lower panel) sample sizes were sufficient to detect some larger effects on RP as significant, but doubling sample sizes would still not have ensured that differences in the 1-0.97 range could be characterized as significant.

Fig. S1. Power calculations.

Table S1. Characteristics of primers and of fragments isolated by the three primer pairs on 320 individuals across 13 island populations and three mainland populations of Iris versicolor in the Bay of Fundy.

Primer / Sequence / Tm (C°) / Motif first isolated / Fragment size range / Number of
different
fragment lengths / Percent of individuals amplifying for a single fragment / Distinctcombinations of fragment lengths
Tet4F / CTAGCCTCGACGCTTTTCC / 64 / agc6 / 224-272 / 15 / 0.0 / 65
Tet4R / GCTGCAGCCTCAGCAGTAGT / 65
Tet9F / GTCCATCATGCCAAGTACGA / 64 / cata8 / 221-257 / 10 / 4.8 / 93
Tet9R / AACGCGAGTTTCCAAACATA / 62
IM200F / AACGGAAATGGCGAATAAACT / 57 / gaa23 / 350-374 / 9 / 1.6 / 68
IM200R / AGATCAGAAGTCCGAGCACCT / 63

______

Table S2. Influence of hand-pollination treatments (excluding within-site outbreeding treatment; cf. Table 1) on survivorship at key life-history stages (fitness traits) and measures of growth and physiology (performance traits) in Iris versicolor on Kent Island, New Brunswick, Canada. N refers to overall sample size, which was similar between the treatments (inbreeding, between-island outbreeding, open-pollinated [control]). Values are significance levels based on sequential fixed effects likelihood ratio tests comparing models fit using maximum likelihood and linear mixed effects or generalized linear mixed effects (Poisson distribution: number of leaves per plant; binomial distribution: proportion of flowers that set fruit, seeds matured, and seeds germinated). Random effects were treatment × clone and clone (or, in greenhouse experiments, block). 1 = traits measured in F1 plants grown in the greenhouse.

P-value, fixed effects likelihood ratio test
Trait / N / Treatment
(χ2, d.f. = 2) / Site
(χ2, d.f. = 1) / Treatment × Site
(χ2, d.f. = 2)
Proportion of flowers that set fruit / 609 flowers / 0.019 / 0.036 / 0.074
Number of seeds per capsule / 415 capsules / 0.065 / 0.572 / 0.302
Proportion of seeds matured / 415 capsules / 0.348 / 0.356 / 0.203
Proportion of seeds germinated / 109 capsules / 0.635 / 0.369 / 0.544
Capsule length / 417 capsules / 0.106 / 0.275 / 0.269
Seed size / 400 capsules / 0.863 / 0.982 / 0.228
Number of leaves1 / 49 plants / 0.324 / 0.100 / 0.590
Leaf length1 / 49 plants / 0.199 / 0.909 / 0.410
Leaf width1 / 49 plants / 0.159 / 0.564 / 0.079
Leaf thickness1 / 49 plants / 0.287 / 0.759 / 0.577
Effective quantum yield (ΦPSII) 1 / 50 plants / 0.946 / 0.184 / 0.749
Maximum quantum yield (Fv/Fm) 1 / 50 plants / 0.131 / 0.460 / 0.262

Table S3. Influence of hand-pollination treatments on Relative Performance (RP = 1 − ws/wo if ws≤ wo; RP = wo/ws− 1 if wswo) at key life-history stages, and measures of growth and physiology in Iris versicolor from Kent Island, New Brunswick, Canada. Means are shown across sites, and separately at the South and West Beach sites, along with 95% C.I. (obtained from 1000 simulations of the maximal models summarized in Table S2, i.e., excluding the within-site outbreeding treatment). Bold values of RP highlight two-tailed differences from 0 at the P < 0.05 level. 1 = traits measured in F1 plants grown in the greenhouse.

Relative Performance (RP)
Trait / Pooled / 95% C.I. / South / 95% C.I. / West / 95% C.I.
Proportion of flowers that set fruit / −0.112 / −0.228,0.006 / −0.209 / −0.355,−0.078 / −0.016 / −0.204,0.184
Number of seeds per capsule / −0.040 / −0.112,0.033 / −0.048 / −0.143,0.050 / −0.033 / −0.141,0.078
Proportion of seeds matured / 0.027 / −0.086,0.143 / 0.050 / −0.082,0.176 / 0.004 / −0.186,0.194
Proportion of seeds germinated / 0.184 / −0.302,0.616 / 0.240 / −0.423,0.771 / 0.129 / −0.536,0.699
Capsule length / −0.016 / −0.066,0.034 / 0.012 / −0.058,0.081 / −0.045 / −0.114,0.028
Seed size / −0.020 / −0.085,0.043 / 0.037 / −0.052,0.125 / −0.077 / −0.164,0.019
Number of leaves1 / 0.152 / −0.050,0.329 / 0.214 / −0.103,0.470 / 0.089 / −0.174,0.322
Leaf length1 / −3×10−4 / −0.101,0.102 / −0.029 / −0.181,0.131 / 0.029 / −0.101,0.154
Leaf width1 / 0.021 / −0.082,0.121 / 0.061 / −0.094,0.210 / −0.020 / −0.149,0.110
Leaf thickness1 / −0.039 / −0.158,0.087 / −0.024 / −0.205,0.167 / −0.054 / −0.206,0.107
Effective quantum yield (ΦPSII)1 / −0.028 / −0.131,0.078 / −0.039 / −0.188,0.128 / −0.018 / −0.157,0.127
Maximum quantum yield (Fv/Fm)1 / 0.001 / −0.008,0.010 / 0.007 / −0.007,0.021 / −0.004 / −0.015,0.007

Table S4. Influence of hand-pollination treatments (including within-site outbreeding treatment) on measures of size and physiology (performance traits) in additional Kent Island Iris versicolor clones grown in the greenhouse but not included in the main analyses because they were unbalanced across treatments. The main analysis (Tables 1, S2) only incorporated data from two clones (WB5, S2) that had plants represented in all treatments in a balanced design. In the present analysis, additional clones were only represented in the following treatments: S25 (control), S28 (inbreeding, within-island outbreeding, control), and WB28 (within-island outbreeding, control). N refers to overall sample size, which was similar between the treatments (inbreeding, between-island outbreeding, within-island outbreeding, open-pollinated [control]). P-values based on likelihood ratio tests of fixed effects in linear mixed effects models. Variables included treatment, site, and treatment × site interaction.

Trait / P-value, fixed effects likelihood ratio test
N / Treatment / Site / Treatment × Site
Number of leaves / 85 plants / 0.201 / 0.958 / 0.984
Leaf length / 85 plants / 0.094 / 0.320 / 0.681
Leaf width / 85 plants / 0.453 / 0.032 / 0.140
Leaf thickness / 85 plants / 0.555 / 0.825 / 0.413
Effective quantum yield (ΦPSII) / 85 plants / 0.233 / 0.001 / 0.875
Maximum quantum yield (Fv/Fm) / 86 plants / 0.316 / 0.028 / 0.025

Table S5: Relative Performance (RP = 1-ws/wo if ws≤ wo, wo/ws−1 if wswo ) for Kent Island Iris versicolor clones grown in the greenhouse (see Table S4). Means and 95% confidence intervals determined by 1000 simulations based on mixed effects models (Table S4). Results are presented pooled across sites, and separately for the South and West Beach sites. There are no cases in which RP is significantly different from 0.

Trait / All Sites / South / West Beach
RP / 95% C.I. / RP / 95% C.I. / RP / 95% C.I.
Number of leaves / 0.004 / −0.216, 0.222 / 0.210 / −0.127, 0.479 / −0.202 / −0.474, 0.145
Leaf length / −0.014 / −0.110, 0.086 / −0.008 / −0.152, 0.139 / −0.020 / −0.149, 0.116
Leaf width / −0.056 / −0.162, 0.052 / −0.010 / −0.172, 0.151 / −0.102 / −0.228, 0.040
Leaf thickness / 0.021 / −0.089, 0.125 / −0.003 / −0.164, 0.153 / 0.045 / −0.106, 0.187
Effective quantum yield (ΦPSII) / −0.002 / −0.100, 0.096 / −0.074 / −0.203, 0.062 / 0.071 / −0.076, 0.207
Maximum quantum yield (Fv/Fm) / 0.003 / −0.006, 0.012 / 0.003 / −0.010, 0.016 / 0.002 / −0.011, 0.015