About the question:
I failed to find the result of the PCR / electrophoresis. However, I will try to explain when to still consider a sample, and when to exclude it.
Answer:
Probably elephants 1, 2, 4, and 6 exposed a pattern of lines that differed in most alleles from the ivory´s pattern. So they were considered to be different and, hence, excluded. However, elephants 3 and 5 only were different in one allele, which might be due to the amplification failure. In these last situations, the experiment must be run again, and include many markers.
Explanation:
The genetic profile is the series of alleles, characteristic of each individual, concerning a series of DNA fragments.
Once genetic profiles are obtained by using PCR/electrophoresis, they must be subjected to interpretation. The researcher must compare profiles to find concordance between individuals or the lack of it.
In the exposed example, probably we have a profile of the ivory sample, and we need to find out to what individual or population it belongs to. So we need to compare the pattern of bands of the different lanes with the ivory´s pattern of bands.
In resumed terms, if the ivory´s patterns completely match with another lane pattern, then this is the individuals or population to whom it belongs the ivory. However, if they do not match, we need to exclude this individual.
But this is not as easy as it looks like. Sometimes it is difficult to interpret the results because bands are not so clear, and sometimes there are failures during amplification, as said before in the statement of the problem.
In these situations, what you need to do is look for the lanes that match the most with the ivory one and discard those that differ the most. These last patterns probably belong to a different animal or population, that why they can be excluded. Individuals suspected of being the same as the individual understudy or belonging to the same population, must be subducted to a repetition of the study. Also, the use of several markers in the study increases the matches chances.
So, in the exposed example, if the amplification fails in one allele, elephants 3 and 5 could be a match to the ivory sample, probably because all the other alleles do match. You can only exclude them if, after the experiment repetition, the involved allele is still different. On the other hand, you can eliminate the elephants 1, 2, 4, and 6 because their pattern bands are significantly different from the ivory´s patterns, so you do not even need to repeat the experiment.