accidental cross-pollination, random mutation, and gene segregation. Since both types <strong>of</strong>selection pressure are highly variable, it is difficult to generalize about <strong>the</strong> rate <strong>of</strong> geneticchange; depending on <strong>the</strong> circumstances, <strong>the</strong> genotype <strong>of</strong> a landrace or improved OPV canchange significantly from one generation <strong>of</strong> plants to <strong>the</strong> next, or it can remain essentiallyunchanged across many generations <strong>of</strong> plants.In hybrids, by far <strong>the</strong> most important source <strong>of</strong> genetic changes is segregation — randomrecombination <strong>of</strong> alleles that occurs when seed is recycled. Key results <strong>of</strong> a simulationexercise designed to show <strong>the</strong> likely effects <strong>of</strong> inbreeding in maize hybrids appear to besupported by findings published in <strong>the</strong> empirical literature on seed recycling:♦ When maize hybrids are recycled, yield usually decreases significantly from <strong>the</strong> F1 to <strong>the</strong>F2 generation. Yield tends to stabilize in subsequent generations, however, and mayeventually begin to increase again if farmers are exerting selection pressure.♦ When maize hybrids are recycled, <strong>the</strong> size <strong>of</strong> <strong>the</strong> yield decrease observed between <strong>the</strong> F1and F2 generations depends in large part on <strong>the</strong> level <strong>of</strong> inbreeding <strong>of</strong> <strong>the</strong> originalparents. Generally speaking, <strong>the</strong> greater <strong>the</strong> degree <strong>of</strong> inbreeding in <strong>the</strong> parents, <strong>the</strong>greater <strong>the</strong> degree <strong>of</strong> heterosis in <strong>the</strong> F1 generation, and <strong>the</strong> greater <strong>the</strong> yield declineobserved between <strong>the</strong> F1 and F2 generations. This relationship may be confounded byenvironmental factors, however.♦ The degree <strong>of</strong> inbreeding <strong>of</strong> <strong>the</strong> parents affects not only <strong>the</strong> size <strong>of</strong> <strong>the</strong> expected yielddecrease but also its variability. The greater <strong>the</strong> level <strong>of</strong> heterozygosity in F1 plants, <strong>the</strong>greater <strong>the</strong> variability in inbreeding depression expected in F2 and F3 plants.♦ Whe<strong>the</strong>r or not advanced-generation hybrids outyield landraces and improved OPVsdepends on <strong>the</strong> original difference in yield and on <strong>the</strong> magnitude <strong>of</strong> <strong>the</strong> decline in yieldcaused by recycling. In some instances, recycled hybrids continue to outyield <strong>the</strong> o<strong>the</strong>rtypes <strong>of</strong> materials, which explains why hybrid recycling may make sense.♦ Recycling <strong>of</strong> hybrids may have little effect on qualitative traits such as kernel size andshape, grain texture, and pounding quality.The finding that seed recycling <strong>of</strong>ten leads to significant genetic changes in farmermaintainedcultivars suggests that <strong>the</strong>re may be a need to reassess <strong>the</strong> categoriestraditionally used to classify maize varieties (e.g., landraces, improved OPVs, hybrids). Inaddition, <strong>the</strong> rapid rate <strong>of</strong> genetic change observed to take place in farmers’ fields hasimportant implications for research impacts assessment studies. Practical guidelines for usein estimating <strong>the</strong> returns to maize breeding research are presented in <strong>the</strong> appendix.v
AcknowledgmentsAny report that is based on a comprehensive review <strong>of</strong> <strong>the</strong> literature owes a great deal topreviously published work, and this one is no exception. We would like to recognize <strong>the</strong>contributions <strong>of</strong> <strong>the</strong> many researchers whose findings we have liberally cited; without <strong>the</strong>irefforts, this syn<strong>the</strong>sis would not have been possible. We especially acknowledge <strong>the</strong>pioneering work on farmers’ varietal and seed management practices carried out by our<strong>CIMMYT</strong> colleagues Dominique Louette, Melinda Smale, and Mauricio Bellon.Drafts <strong>of</strong> this paper were reviewed by Gregory Edmeades, Melinda Smale, and MiguelLópez-Pereira, all <strong>of</strong> whom provided useful comments. In addition, many <strong>CIMMYT</strong>colleagues and national program collaborators took <strong>the</strong> time to go through <strong>the</strong>ir files insearch <strong>of</strong> studies <strong>of</strong> potential relevance. The manuscript was copyedited by Kelly Cassaday<strong>of</strong> <strong>the</strong> <strong>CIMMYT</strong> Publications Unit.vi
- Page 1 and 2: E C O N O M I C SWorking Paper 99-0
- Page 3 and 4: CIMMYT (www.cimmyt.mx or www.cimmyt
- Page 5: Executive SummaryThis paper summari
- Page 9 and 10: followed a very different path comp
- Page 11 and 12: Farmers’ Management of Maize Vari
- Page 13 and 14: In recent years, new evidence has e
- Page 15 and 16: state of Guanajuato and discusses t
- Page 17 and 18: Based on the results of a survey co
- Page 19 and 20: elied with greater frequency on the
- Page 21 and 22: hybrid seed. Initially, it was gene
- Page 23 and 24: In Veracruz State, Mexico, where mo
- Page 25 and 26: Recent work in the highlands of Mex
- Page 27 and 28: producers) tend to rely on family,
- Page 29 and 30: Unintentional seed mixingUnintentio
- Page 31 and 32: Table 11. Yield depression resultin
- Page 33 and 34: Genetic driftWhen farmers select ea
- Page 35 and 36: (a) Production of a single-cross hy
- Page 37 and 38: Breeding hybrid maize begins with t
- Page 39 and 40: According to this theory, genes tha
- Page 41 and 42: Wright’s finding, which is based
- Page 43 and 44: Similarly, the mean of F3 generatio
- Page 45 and 46: 2. Between the F1 and F2 generation
- Page 47 and 48: In a series of on-station trials co
- Page 49 and 50: As part of the same trial, Ramírez
- Page 51 and 52: Table 19. Inbreeding depression obs
- Page 53 and 54: to 41% for the single cross (Figure
- Page 55 and 56: DiscussionEvery time a farmer recyc
- Page 57 and 58:
The finding that genetic change in
- Page 59 and 60:
Brennan, J.P., and D. Byerlee. 1991
- Page 61 and 62:
Murillo Navarrete, P. 1978. Estimac
- Page 63 and 64:
AppendixGuidelines for Estimating t
- Page 65 and 66:
exactly how different the plants wo
- Page 67 and 68:
As a general rule, we propose that