...
But that still leaves the carbon just to the right of the amide group not respecting the octet rule by only having 2 carbon-carbon bonds and only one carbon-oxygen bond (normally there are 2 as well).
...
A couple pictures; first, the crop circle of topic:
Source:
Cropcircle Melatonin - C13H16N2O2 -- Sott.net
Second, a ball-and-stick (B&S) model for Melatonin:
Source: Melatonin - Wikipedia
https://__en.wikipedia.org/wiki/Melatonin
Let's call that "...not respecting the octect rule..." carbon the "carbonyl" carbon (a name for C=O double bond). As it is alongside that -NH- (amine), it is part of the amide ((O=)CN) group, a.k.a. carboxamide group, here.
From the B&S model, we can see the double line for the C=O (<<which is also represented in our "shorthand formula", a.k.a. "line-angle formula", a.k.a. "skeletal formula" as: C=O) bond clearly. This illustration, along with the double lines used in shorthand formula, function both as a book-keeping device (for counting electrons and bonds) and to describe what is occurring.
So for the crop circle, they show one line for that C=O, and it is not much distinguishable from the bonds in the "ether" group (C-O-C) on the other side of the molecule (the other O in the structure; they are small-ish white circles). They also do not demarcate the "lone pairs", or non-bonding pairs, of electrons, which each O should have two pairs and the N should have one pair. The more nuanced version of looking at it (one which I doubt I understand, and need to better study) is to develop the "molecular orbital", to describe how those atoms (and later, by extension, the whole molecule) are actually sharing electrons.
That involves the hybridization of the atomic orbitals , which goes into how the electron clouds deform to different shapes that allow for the lowest-energy situation in the bond's electron cloud. A molecule like this, especially given the double-bonds in the ring structure (extended "pi-bonding") would likely have potential to distribute it's electronic structure across the whole molecule. But, my understanding is limited: I think it would suffice to show that when the pi-bond orbital forms (one bond in C=O is a sigma bond, the second is a pi-bond), it is actually a distribution of possible locations in
two extended clouds; hence, demarcating it as equivalent to the sigma bond is not illustrative (a single dashed line - could show the sigma bond, the pi-bond would be better represented as two dashed lines above and below that, even though it is only "one bond"), but does serve for good book-keeping shorthand.
Source: http://__wps.prenhall.com/wps/media/objects/724/741576/Instructor_Resources/Chapter_17/Text_Images/FG17_01.JPG
From: https:__//i.ytimg.com/vi/RFxZ9YnltZ8/hqdefault.jpg
3D molecular structures of melatonin (MLT), of 6-hydroxymelatonin (6-OHM), and of 5,6-dihydroxytryptamine (5,6-DHT).
(A) 3D molecular structure to note the distances between 2 oxygen atoms of 6-OHM and 2 phenolic hydroxyls of 5,6-DHT, respectively 2.77 and 2.82 Ångström (Å);
(B) Frontier orbitals: HUMO (highest occupied molecular orbital);
(C) Frontier orbitals: LUMO (lowest unoccupied molecular orbital);
(D) Electrostatic Potential Surfaces (ESP). The geometry of the systems has been optimized considering the semi-empirical molecular orbital theory at the level of AM1 (Austin Model 1) and the electronic properties of the systems have been calculated by ab initio Restricted Hartree-Fock). ArgusLab 4.0.1 software (Mark Thompson and Planaria Software LLC, 2004) was used.
From: Pacini, Nicola & Borziani, Fabio. (2016).
Oncostatic-Cytoprotective Effect of Melatonin and Other Bioactive Molecules: A Common Target in Mitochondrial Respiration. International Journal of Molecular Sciences. 17. 341. 10.3390/ijms17030341.
available here: https://w__ww.researchgate.net/publication/297657923_Oncostatic-Cytoprotective_Effect_of_Melatonin_and_Other_Bioactive_Molecules_A_Common_Target_in_Mitochondrial_Respiration
To return back to the topic of this crop circle, I am most incredulous when imagining that this was made with lawnmowers. I find the overall roundness with which the molecule is depicted to be notable, too, and will likely take to bending the B&S models around a bit more when I use them from now on. Maybe if you're seeing the electrostatic potential surfaces (like in the last image for MLT(D)) without the B&S skeleton in the background it looks more like this.