dissenting:
I respectfully dissent. The evidence is sufficient to support Pena’s convictions of possessing, and conspiring to possess, with intent to distribute, less than 50 kilograms of marijuana, but it is clearly insufficient to support a jury’s finding beyond a reasonable doubt that Pena’s offenses involved in excess of 100 kilograms of marijuana.
The majority simply blinks the serious deficiency in the government’s proof of the quantity of marijuana involved. Because the indictment alleges a quantity greater than 100 kilograms of marijuana, triggering punishment at an elevated level within the statutory range of 21 U.S.C. § 841(b)(1)(B), Apprendi v. New Jersey, 530 U.S. 466, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000), applies and proof beyond a reasonable doubt of an amount greater than 100 kilograms became a fourth element of the offense alleged in this indictment. United States v. DeLeon, 247 F.3d 593, 596 (5th Cir.2001); United States v. Doggett, 230 F.3d 160, 164-65 (5th Cir. 2000). The government did not present sufficient evidence upon which a rational trier of fact could find beyond a reasonable doubt that Pena possessed more than 100 kilograms of marijuana, or indeed that he possessed more than 50 kilograms of marijuana; the evidence was sufficient only to show beyond a reasonable doubt that he possessed less than 50 kilograms of marijuana. Its officers testified they cut into only one of 150 thickly duct-tape wrapped, non-transparent bundles found in Pena’s tractor-trailer and took small samples of plant substance; they sent these samples from only one of the 150 bundles to the DEA laboratory. The DEA chemist testified that these small samples altogether weighed 926.2 grams and were found to contain marijuana. The government’s officers did not cut into or visually inspect the contents of the 149 unopened bundles. Although the 150 bundles were weighed in groups of four and found to have a total weight of 2,087 pounds, none of the bundles was weighed individually; and no effort was made to determine the weight of the duct tape wrappings. None of the bundles was measured for exact size or volume either; the most that any witness said about size was that the bundles were of approximately of the same size.
At trial, Officers Torres and Brizendine described in some detail the single bundle that they opened or cut into and extracted eleven small samples of green leafy material from. The leafy green substance was wrapped 5 to 10 times in cellophane, and *519then all of that was wrapped around by between one-half inch and one inch thickness of opaque duct tape. They testified that the other 149 bundles, which they did not open or sample, were approximately the same size and similarly wrapped in red or black opaque duct tape. They said that they could not see or detect what was contained in the other 149 bundles that remained wrapped. DEA forensic chemist Scott Wischnewsky testified that, as the result of his analysis, he found that each of the eleven small samples contained marijuana; and that the eleven samples of plant material collectively weighed 926.2 grams.1 The jury convicted Pena on both the conspiracy and possession with intent to distribute counts, as involving in excess of 100 kilograms of marijuana, and the district court sentenced him to 135 months on each conviction, consecutively, and eight years of supervised release. Pena made a timely motion for judgment of acquittal after the government completed its case, and again in a written motion after trial. The district court denied his motions, and Pena now appeals.
1.
Pena’s sufficiency-of-evidence challenge is reviewed de novo, United States v. Ragsdale, 426 F.3d 765, 770 (5th Cir.2005). The jury verdict of conviction should be affirmed if “after viewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt.” Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979) (second emphasis added); Ragsdale, 426 F.3d at 770-71. “This familiar standard gives full play to the responsibility of the trier of fact fairly to resolve conflicts in the testimony, to weigh the evidence, and to draw reasonable inferences from basic facts to ultimate facts. Once a defendant has been found guilty of the crime charged, the factfinder’s role as weigher of the evidence is preserved through a legal conclusion that upon judicial review all of the evidence is to be considered in the light most favorable to the prosecution. The criterion thus impinges upon ‘jury1 discretion only to the extent necessary to guarantee the fundamental protection of due process of law.” Jackson, 443 U.S. at 319, 99 S.Ct. 2781 (footnotes omitted).
2.
The record clearly demonstrates that only one bundle was opened or cut into, and that all of the eleven samples analyzed by the DEA forensic chemist were taken from that one bundle. On cross-examination, Torres testified that only one bundle was opened, that no other bundle was opened or cut into; that he saw that the single bundle they probed contained a green, leafy substance; and that he could not see the contents of any of the other 149 bundles. Brizendine’s testimony was consistent with Torres’ statements. Brizendine testified that he cut into only one of the 150 bundles, and removed the eleven samples from inside that bundle. Thus, all eleven samples tested by the DEA lab weighed less than one kilogram and came from a single bundle out of a total of 150 bundles. The officers testified that each bundle appeared to be about the same size, but because the 150 bundles were wrapped in opaque tape, no one could testify as to what was inside the 149 unopened bundles. Further, because none of the bundles was weighed individually, the officers did not *520attempt to testify as to the weight of the single bundle they opened, much less to say how much the total leafy green substance as opposed to the wrappings of that single bundle weighed. In fact, the officers did not venture in any way to quantify how much, if any, leafy green substance was left in the single bundle after they had removed the eleven samples.
Consequently, under these circumstances, the evidence that one bundle out of 150 contained less than one kilogram of marijuana was clearly insufficient evidence to support a rational finding or extrapolation beyond a reasonable doubt that the 150 bundles contained in excess of 100, or even in excess of 50, kilograms of marijuana.2 Under the principles of simple random sampling, logic, and common sense, because the eleven samples of marijuana were taken exclusively from only one of the 150 bundles, those samples were insufficient evidence to allow a jury to infer beyond a reasonable doubt that the 150 bundles, 149 of which were not opened or tested in any way, contained in excess of 100, or even 50, kilograms of marijuana. The samples were not drawn randomly from the whole of the 150 bundles, so that the contents of each of the 150 bundles would have an equal chance of inclusion or representation in the samples taken. Thus, the eleven samples were not a true “simple random sample,” and therefore not an unbiased representation of the contents of the whole 150 bundles. See, e.g., C. Philip Cox, A Handbook of Introductory Statistical Methods 1 (1987) (defining random sample as being a sample “obtained by taking observations independently of each other to ensure that each member of the population has an equal chance of being selected, that is, of appearing in the sample”); see also David H. Kaye & David A. Freeman, Reference Guide on Statistics, Reference Manual on Scientific Evidence 225 n. 34 (Fed. Judicial Ctr. 3d ed.2011) (explaining that “[i]n simple random sampling, units are drawn at random ... [and have] the same probability of being chosen for the sample”).
This court in In re Chevron U.S.A, Inc., 109 F.3d 1016 (5th Cir.1997), observed that the applicability of the science of inferential statistics has long been recognized by the courts;3 that the essence of inferential statistics is that one may confidently draw inferences about the whole from a representative sample of the whole; but that, in order to fairly and reliably draw such an inference, the sample must be randomly selected and of sufficient size so as to be representative of the whole to the desired level of confidence in the result obtained. *521Id. at 1019-20. Thus, in Chevron, which involved some 3,000 land contamination cases, this court barred the district court from using the trial of 30 of the cases as bellwether cases or sample trials, because the bellwether cases were not a randomly selected or a large enough sample to ensure that the results of the sample trials would be representative of those that would be obtained from the trials of the whole 3000 cases. Id. at 1020-21.
Under the same principles of logic and inferential statistics, it is self-evident that in the present case the eleven samples of marijuana taken from only one of the 150 bundles and weighing 926.2 grams cumulatively was not a randomly selected sample and was not reliably representative of the contents of the whole 150 bundles.4 What is more, the government does not even attempt to explain or present an argument for how a rational inference beyond a reasonable doubt may be drawn that the whole 150 bundles contained over 50 or 100 kilograms of marijuana (over 50,000 or 100,000 grams), based only upon eleven samples weighing 926.2 grams (less than one kilogram) taken exclusively from a single bundle. The government cites no authority and, in effect, waives its argument as to how the jury rationally could have found beyond a reasonable doubt that the quantity of marijuana involved weighed over 50 or 100 kilograms.
I did not find any other case in which the government through its officers took an obviously non-random, and thus non-representative, sample from the whole of the substances seized and then, without even a visual inspection of the whole of the suspected seized substances, asked a jury to extrapolate therefrom to find beyond a reasonable doubt that the defendant possessed the total quantity of illicit substance as charged. Usually, the government presents a witness or witnesses who can testify that a random sample was taken from the whole of the substances seized, that a visual inspection indicated the whole to be made up of a substance identical to the sample, and that scientific analysis of the random sample indicates that the whole quantity of the visually similar substances were illicit drugs. See, e.g., United States v. Fitzgerald, 89 F.3d 218, 223 n. 5 (5th Cir.1996); United States v. Jackson, 470 F.3d 299, 310-11 (6th Cir.2006); United States v. Roach, 28 F.3d 729, 735 (8th Cir.1994); United States v. Scalia, 993 F.2d 984, 988, 989-90 (1st Cir.1993); United States v. McCutchen, 992 F.2d 22, 26 (3d Cir.1993); United States v. Madkour, 930 F.2d 234, 239 (2d Cir.1991); United States v. Garza, 222 Fed.Appx. 433, 438 (5th Cir.2007).5 As observed by several Courts of Appeals, “[cjourts have endorsed statistically based drug-quantity extrapolations predicated on random test samples in *522circumstances where the government was able to demonstrate an ‘adequate basis in fact for the extrapolation and that the quantity was determined in a manner consistent with the accepted standards of [reasonable] reliability.’ ” Jackson, 470 F.3d at 310-11 (second alteration in original) (quoting Scalia, 993 F.2d at 989 (in turn quoting McCutchen, 992 F.2d at 25-26)). “[S]ufficient indicia of reliability may be found where a preponderance of the evidence demonstrates that (1) a proper ‘random’ selection procedure was employed; (2) the chemical testing method conformed with an accepted methodology; (3) the tested and untested samples were sufficiently similar in physical appearance; and (4) the tested and untested samples were contemporaneously seized at the search scene.” Id. at 311 (alteration in original) (quoting Scalia, 993 F.2d at 989). “In other words, ‘the overall margin of reliability in a drug-quantity approximation must be adequate to afford reasonable assurance that the defendant is not subjected to a guideline sentence or a mandated minimum sentence greater than warranted by the reliable evidence.’ ” Id. (quoting Scalia, 993 F.2d at 989); see also McCutchen, 992 F.2d at 25-26 & n. 8 (“[T]he government must show, and the court must find, that there is an adequate basis in fact for the extrapolation and that the quantity was determined in a manner consistent with accepted standards of reliability.”).
In the present case, the DEA agents did not take random samples from the 150 bundles in the truck. Instead, they took all of the eleven samples exclusively from one of the bundles without visually inspecting the contents of the other 149 bundles. Therefore, Wischnewsky’s chemical test on less than one kilogram of plant material from one bundle was not performed on a random sample of the contents of the whole load of 150 bundles. Consequently, the chemist’s analysis does not present a reliable representative basis for an extrapolation as to the quantity of marijuana in the whole 150 bundles, and did not afford the jury a rational basis from which it could find beyond a reasonable doubt that the 150 seized bundles contained either 100 or 50 or more kilograms of marijuana.
3.
However, the jury necessarily found all of the elements of the lesser offenses involving less than 50 kilograms of marijuana, and the evidence is sufficient to support Pena’s convictions and sentences at that level. Accordingly, this court should exercise its authority under 28 U.S.C. § 2106 to vacate the district court’s judgment of convictions and sentences and to remand the case to the district court with direction to enter judgment of conviction of conspiracy to possess, and possession of, less than 50 kilograms of marijuana, with intent to distribute; and to sentence Pena accordingly. See United States v. Skipper, 74 F.3d 608, 612 (5th Cir.1996). For these reasons, I respectfully dissent.
. Wischnewsky testified that after his testing the remaining plant material weighed 924.2 grams.
. The majority relies on the government's testimony that the bundles were similarly wrapped and the same size. But the testimony of Thomas Haskins at trial reflects only that they appeared to be the same size. The bundles were never individually measured or weighed. Moreover, the mere fact that the bundles appeared to be similar in size and wrapping cannot provide proof beyond reasonable doubt that the contents of the bundles were the same or that they contained marijuana.
. Id. at 1019-20 (citing Castaneda v. Partida, 430 U.S. 482, 498, 97 S.Ct. 1272, 51 L.Ed.2d 498 (1977) (using statistical data to prove discrimination in jury selection); Capad v. Katz & Besthoff, Inc., 711 F.2d 647, 653-57 (5th Cir.1983) (using census data in gender discrimination cases); Exxon Corp. v. Tex. Motor Exch. of Houston, Inc., 628 F.2d 500 (5th Cir.1980) (using statistical sampling in trademark infringement suit); Ageloff v. Delta Airlines, Inc., 860 F.2d 379 (11th Cir.1988) (using evidence of life-expectancy tables to determine damages); G.M. Brod & Co., Inc. v. U.S. Home Corp., 759 F.2d 1526, 1538-40 (11th Cir.1985) (using expert testimony as to profit projections based on industry norms); United States v. 449 Cases Containing Tomato Paste, 212 F.2d 567 (2d Cir.1954) (approving inspector's testing of samples, rather than requiring the opening of all cases)).
. Indeed, the samples as taken do not even reliably indicate how much marijuana material was in the single bundle that Torres and Brizendine sampled. It is undisputed that the single tested bundle was not weighed individually, and neither Torres or Brizendine testified about how much green leafy substance was left after they extracted the samples. Therefore, while the evidence was sufficient to demonstrate that the single bundle from which the samples were removed contained marijuana, there was not sufficient evidence to determine or estimate the quantity of marijuana the bundle contained.
. In Jackson, Fitzgerald, and Roach, the appellate courts were reviewing cases in which quantity was an element of the crime that needed to be proved beyond reasonable doubt. The remaining cases, including Scalia, McCutchen, and Madkour, involved cases in which quantity was not an element of the crime and was relevant only to the sentencing determination. In those cases, the government needed only to establish the quantity by a preponderance of the evidence, and review was for clear error. See, e.g., McCutchen, 992 F.2d at 26.