Ventajas del precedente vinculante

I I I . l - INTRODUCTION -

Edinburgh populations of ^ vulgaris th at are polymorphic for capitulum type exhibit two peak periods of seed production and hence seedling recruitment per year. These occur in la te spring/early summer and in autumn (Abbott, pers. comm.). As a consequence, i t is feasible th a t the re la tiv e fitn esses of the rad iate and non-radiate morphs may d iffe r for plants which grow in populations th at develop and reproduce at d ifferen t times of the year. Abbott (1986) using material of both morphs derived from a large polymorphic Edinburgh population, investigated differences between the rad iate and non-radiate morphs of S^ vulgaris for a range of lif e history t r a i t s , in experiments in itia te d under glass in June and November, He found th at although the rad iate morph produced greater dry weight than the non-radiate morph in both experiments, only in the study in itia te d in June was the difference between morphs sig n ifican t,

A difference in re la tiv e fitn e s s between morphs might also vary with density, In studies on the re la tiv e fitnesses of chasmogamous (open pollinated) and cleistogamous (se lf pollinated) offspring of Impatiens capensis, Waller (1984) found th at although seedlings of chasmogamous offspring always exhibited the greater fitn ess (based on dry weight), differences in fitn e s s varied sig n ific a n tly at d iffe re n t d en sities.

In the investigations reported in th is chapter, an analysis has been conducted on the possible e ffects of planting date and density on the re la tiv e fitn esse s of the rad iate and non-radiate morphs grown under conditions which approximate to those in the wild. The effect of planting date on the re la tiv e fitn esse s of the two morphs was examined by in itia tin g experiments in Spring and Autumn, 1984, at the Botanic Gardens St. Andrews, while the effects of density were investigated by raisin g plants of each

morph in high and low density pure stands and in mixtures. Relative fitn esse s were measured in terms of survivorship and fecundity,

I I I . 2 - MATERIALS and METHODS ~

Plant m aterial was derived from seeds collected in May 1984 for the Spring 1984 experiment, and September 1984 for the Autumn 1984 experiment. Seedlings were produced from seeds collected from plants raised under conditions previously described in Chapter II section 2.

With the experiment in itia te d in Spring 1984 (planted 20.5,84), comparisons were made between three treatments previously described in Chapter XI, i . e . pure stands of the radiate and non-radiate morphs and a mixed stand in which morphs were present at a 1:1 ra tio with individuals spaced at 3cm, in terv als (high density). This represented a tra d itio n a l de Wit replacement series design (de Wit, 1960). In addition, two further treatments were included; these were a pure stand of radiate plants and a pure stand of non-radiate plants in which spacing between individuals was increased to 6cm (low density). These additional plots allowed an assessment of density effects in pure stands.

In Autumn 1984 at the same s it e at St.Andrews, a second experiment was in itia te d (planted 7-10.10.84) in which the design of the Spring 1984 experiment was repeated but with the addition of a low density mixed stand in which plants were spaced at 6cm. in terv als. Thus, the design of th is experiment represented two de Wit replacement series, one series in which plants were raised a t high density (3cm. spacing) and the other in which plants were raised at low density (6cm, spacing).

In each experiment, rep lic atio n was eight-fold and plots were arranged, one metre apart, in a randomised block design, with rep lic ates arranged in blocks. Within each p lo t, seedlings were planted a t points of a 14 x 12 grid. Plots were monitored on a weekly basis to assess m ortality throughout the seedling, vegetative and flowering stages. Due to a large "edge" e ffect

noted in plots raised within the 1983 experiments (see Chapter II) the two ; outer rows of each plot were designated as guard rows (Guard Row P lants). ‘ Data were collected separately from guard rows and inner rows (Plot Plants) I before analysis.

At harvest a record was taken of the number of plants to set seed, and *•; the number of capitula produced per plant ( i.e a fte r a l l plants in a plot : had finished fru itin g . Plants which had survived in the Spring 1984 experiment were harvested on 30,8.84 while those from the Autumn 1984 experiment were collected on 30,6.85. In the Spring 1984 experiment, an i

additional record was taken of ovule and seed number per capitulum plus ovule f e r t i l i t y . This entailed sampling a single capitulum from each of 20 plants in the outer two rows and 30 plants from the inner rows of each plot over a five day period when most plants were fru itin g . Capitula were teased open using fine forceps and the number of f ill e d seed recorded. The number of ovules per capitulum was calculated as the to ta l number of f ille d and aborted seed produced in a capitulum, while ovule f e r t i l i t y was computed as the proportion of ovules to produce f ille d seed.

111,3 - ANALYSIS OF DATA - I I I . 3.1 - Spring 1984 Experiment -

Density e ffe c ts in pure stand

Differences between morphs raised in high and low density pure stands for survivorship, number of plants to f r u it and to ta l seed produced per p lo t, were examined by a three factor analysis of variance. In the ANOVA i

(see Table I I I . l ) , a mixed model was assumed with morphs (treatment term) and density considered as fixed e ffe c ts, and rep lic atio n as a random effe c t. The treatment x density x re p lic a te interaction was used as an estimate of the error term to te s t the significance of the treatment x density in teraction; while the treatment and density main effe c ts were tested for significance against the treatment x replicate and density x

Table I I I . l Three factor ANOVA model to examine differences between radiate (R) and non-radiate (N) plants in survivorship and fecundity when raised in high and low density pure

stands. Source of Variation d .f Treatment (R v N) 1 Density 1 Replication 7 Treatment x Density 1 Treatment x Replication 7 Density x Replication 7

Treatment x Density x Replication 7

Total 31

— 3 2 —

rep lic a te interactions respectively.

For seed and capitulum number per plant, and ovule number, seed number and ovule f e r t i l i t y per capitulum, the mean square of the treatment x density x rep licate could be tested for significance against the within rep lic ates error term. When the interaction was not sig n ifican t (with p>.25), i t s mean square was pooled with the within rep lic ates mean square to form a combined error term for testin g the significance of the treatment x density in teraction (Winer, 1969).

Replacement series analysis

Differences between the high density replacement series treatments ( i.e . pure stand rad iate, pure stand non-radiate and the mixture) for survivorship and fecundity, were examined by a two factor ANOVA sim ilar to i th at described in Table I I . 1 (see Chapter I I . 3). Subsequent analysis followed the procedure outlined in Chapter I I . 3,

I I I , 3.2 - Autumn 1984 Experiment -

The structure of the ANOVA used for the analysis of survivorship, number of plants to f r u it , capitulum number per p lan t, and to ta l capitula produced per rep lic ate is described in Table I I I , 2, To determine differences between: (i) morphs raised in pure stand; and ( ii) pure stands and the mixture, two orthogonal contrasts were extracted from the treatment term as shown in the tab le. In addition, an examination of a difference in the effect of each of these contrasts with density was conducted by extracting two further orthogonal contrasts from the treatment x density in teractio n .

Further analysis of data was conducted in a manner sim ilar to that described in Chapter I I . 3, However, as the experimental design consisted of both high and low density pure stands and mixtures, differences in survivorship and fecundity for a given morph, raised in pure stand compared

■'1 Table 1X1,2 Three factor ANOVA model to examine differences in

survivorship and fecundity between &ure stands of radiate (R) and non-radiate (N) plants and a 1:1 mixture (Mix) of each morph, raised in high and low density p lo ts.

Orthogonal contrasts have been extracted from the treatment main e ffect and treatment x density in teractio n to

determine differences between: (i) the rad iate (R) and non-radiate (N) pure stands; and ( ii) pure

stands (P.S) and mixture (Mix),

Source of Variation d .f

Treatment (R v N v Mix) 2

Contrast (i) ( R v N ) ₍₁₎ Contrast ( ii) (PS v Mix) ₍₁₎

Density 1

Replication 7

Treatment x Density 2

Contrast (i) x Density ₍₁₎

Contrast ( ii) x Density (1)

Treatment x Replication 14

Density x Replication 7

Treatment x Density x Replication 14

Total 47

— 34 —

to mixture, u tilis e d a three factor ANOVA with density as the additional factor,

I I I . 4 - RESULTS -