Proceedings of the XLV Italian Society of Agricultural Genetics - SIGA Annual Congress

Salsomaggiore Terme, Italy - 26/29 September, 2001

ISBN 88-900622-1-5

Poster Abstract

INFLUENCE OF MARKER AND PLANT SAMPLING ON GENETIC DIVERSITY MEASUREMENT IN CORN LANDRACES

PALLOTTINI L., PARRINI P.

Dipartimento di Agronomia Ambientale e Produzioni Vegetali Università degli Studi di Padova Via Romea 16, 35020 Legnaro, Padova

paolo.parrini@inipd.it

corn, local germoplasm, molecular markers

A research project aimed to characterize an old flint maize landrace, named “Nostrano di Storo” (Zea mays var. indurata) and to evaluate the influence of different conservation methods (“in situ”, on farm and “ex situ”) on the genetic structure of populations grown by farmers was set up. A genetic map of the landrace that includes 282 marker loci (222 AFLPs, 40 SAMPLEs, 19 ISSRs and 11 RAPDs) for a total length of 1917 cM was previously constructed.

The AFLP primer combinations and the RAPD and ISSR primers used for the map were applied to assess the genetic structure of two field populations of “Nostrano di Storo” (named NSt8 and NSt10) chosen because they were considered representative of the whole landrace in terms of morpho-phenological and agronomic traits. A total of 94 plants from 47 randomly chosen ears (2 plants per ear) were assayed at 584 marker loci (414 AFLPs, 40 SAMPLEs, 19 RAPDs and 11 ISSRs) of which 166 were mapped (145AFLPs, 10 SAMPLEs, 6 RAPDs and 5 ISSRs).

Here we intend: I) to evaluate the efficiency of different multi-locus PCR-based methods in detecting genetic diversity; II) to compare the PIC (polymorphism information content) recovered by mapped and random markers; III) to establish the lowest number of markers and plants per population required to describe the genetic structure of the landrace.

The contingency test allowed us to verify that the use of one or two plants per ear does not imply significant changes on allele frequency at the marker loci assayed. Moreover, the cluster analysis based on Dice’s genetic similarity estimates showed that the two field populations (NSt8 and NSt10) can be considered as a single population. Mapped markers showed higher mean genetic similarities based on pair-wise comparison than the total set of molecular markers.

Genetic diversity coefficients measured by the PIC using mapped markers were proven to be significantly higher than those based on total markers (respectively, 0,392±0,009 vs. 0,373±0,005). Within the mapped AFLP markers, primer combinations EcoRI/MseI scored PICs higher than PstI/MseI (0,402±0,012 vs. 0,389±0,014).

No significant changes of marker allele frequency and PIC value were observed when the overall sampled number of ears was reduced from 47 to 30. The adoption of a lower number of ears (e.g. 25, 20 or 15) resulted in a significant reduction of the PIC and in a higher standard error.

The influence of the number and type of molecular markers on genetic diversity measurement was also investigated. Different sample sizes in steps of 40 (from 40 to 160 for mapped markers and from 40 to 360 for total markers) were compared. Out of the total assessed markers, respectively mapped and random, 1000 samples for each size were taken at random. For each class of samples the mean PIC values, standard deviations and CVs of standard deviations were computed. Reduction of sample size, both for total and mapped markers did not cause any significant change in terms of mean PIC values. However, CVs of standard deviations were proven to increase from 2% with sample size of 360 to 10% with 40 random markers and from 1% to 10% with 160 and 40 mapped markers, respectively. A total of 120 random markers and 80 mapped markers is needed to get CVs lower than 5%.

Influence of markers and plant sampling on statistics used to measure genetic diversity should be useful to investigate the genetic consequences of different modes of conservation of corn landraces.

In conclusion, reliable and effective investigations of landrace population genetic structure with AFLP markers can be performed using at least 30 ears per population and one plant per ear and require at least 60-80 mapped marker loci.