Resonance occurs when small amounts of energy accumulate because the rate at which they are being added matches the the frequency of the system itself and the phase. The classic example is pushing a child on a swing. If you add energy, ie push, at a frequency that matches the fundamental frequency of the pendulum motion of the swing the amplitude slowly increases. The phase also matters because for optimum energy transfer from you to the swing you need to add it at the correct phase ie at the end points of the cycle.
So if you want to (apparently) amplify a particular frequency of sound you need to put it in a chamber where that frequency is also the resonant frequency of the chamber (or a multiple thereof - think pushing the swing 1 in every every 2 or 3 swings etc).
A few observations.
First - if you record sound for a short time, the bandwidth of the sample will result in a smearing of the peaks. This only really matters if the sample is very short - with a 1 second sample you would have 1 Hz resolution, but if you sample for 0.01 second, the bandwidth is 100 Hz.
Second, you are using a scale that is quite compressed in the region of interest. That again makes me wonder whether you have not set up your sampling to be optimal for the frequencies of interest.
I have used a cheap iPhone app in the past to record sounds (Signal Spy - I am not connected to the product) and get an idea of their spectral content. I just played a simple scale on my guitar, and got the following (time along the horizontal axis, frequency vertical, intensity shows what frequencies are detected):
I have the feeling there's a problem with the labels on the logarithmic scale, but you can clearly see the fundamental and its harmonics; they are much better resolved than in your case.
This means that either your frequency is not constant, or your recording settings are very much not optimal for the task. Perhaps you can comment on the settings you used, and we can figure out how to get similar results for you. You used the words "a short region of the recording with minimal resonance was used..." - I wonder if your recording window was too short. I also wonder what windowing technique you used. When you do frequency analysis, you don't simply do the Fourier Transform of a snippet of sound - because if you do, you will generate a bunch of frequency content due to the way the signal "cuts off" at the start and end of the recording. Instead you need to apply an apodizing window (Hamming or Hanning window, usually) to get rid of extraneous peaks, and get cleaner frequency peaks.
If you have a pitch generator, the most accurate thing you could do would be to play the known pitch and slowly ramp it until the guitar string started to resonate. That works very well... as the resonance width is quite narrow. You would be able to determine the frequency within a fraction of a Hz (assuming your pitch generator produces enough output - perhaps you play it through a microphone and amplifier into a decent speaker).
Best Answer
It is not amplification! The purpose of the guitar body is to impedance and mode match between the string and the surrounding air.
Intuition
When a an object vibrates it pushes on the surrounding air creating pressure waves which we hear as sound. A string vibrating alone without the body of the instrument doesn't make a very loud sound because exchange of energy from the vibrating string to the air pressure waves is inefficient. Why is it inefficient?
The fundamental reason is that the string is a lot stiffer than the surrounding air and has a small cross sectional area. This means that as the string vibrates with a given amount of energy, it doesn't actually displace much air. With the same energy in the motion, a larger, more mechanically compliant object (e.g. the acoustic guitar body) would do a better job at transferring the energy into the air and thus into your ears.
Analysis
The equation of motion of a vibrating string with vertical displacement $u$ and horizontal position $x$ is
$$\frac{\partial^2 u}{\partial x^2} = \frac{1}{v^2} \frac{\partial^2 u}{\partial t^2}$$
where $v = \sqrt{T/\mu}$ is the velocity of sound in the string, $\mu$ is the linear mass density, and $T$ is the tension. If you consider, for example, the fundamental mode, then the solution is of the form
$$u(x,t) = \sin(k x) f(t)\,.$$
Plugging this into the equation of motion gives you
$$ \begin{align} -k^2 f &= \frac{1}{v^2} \ddot{f} \\ 0&= \ddot{f} + \omega_0^2 f \end{align} $$ where $\omega_0^2 = (vk)^2 = (2\pi)^2T/(\mu \lambda^2)$ and $\lambda$ is the wavelength of the mode ($k \equiv 2\pi / \lambda$). This is just the equation of a harmonic oscillator.
Now suppose we add air friction. We define a drag coefficient $\gamma$ by saying that the friction force on a piece of the moving string of length $\delta x$ is
$$F_{\text{friction}} = -\delta x \, \gamma \, \dot{u} \, .$$
Note that $\gamma$ has dimensions of force per velocity per length. Drag coefficients are usually force per velocity; the extra "per length" comes in because we defined $\gamma$ as the friction force per length of string.
Adding this drag term, re-deriving the equation of motion, and again specializing to a single mode, we wind up with
$$0 = \ddot{f} + \omega_0^2 f + \frac{\gamma}{\mu} \dot{f} \, .$$
Now we have a damped harmonic oscillator. The rate at which this damped oscillator decays tells you how fast (i.e. how efficiently) that oscillator transfers energy into the air. The energy loss rate $\kappa$ for a damped harmonic oscillator is just the coefficient of the $\dot{f}$ term, which for us is $\kappa = \gamma / \mu$.$^{[1]}$
The quality factor $Q$ of the resonator, which is the number of radians of oscillation that happen before the energy decays to $1/e$ of its initial value, is $$Q = \omega_0 / \kappa = \frac{2\pi}{\lambda} \frac{\sqrt{T \mu}}{\gamma} \, .$$
Lower $Q$ means less oscillations before the string's energy has dissipated away as sound. In other words, lower $Q$ means louder instrument. As we can see, $Q$ decreases if either $\mu$ or $T$ decreases. This is in perfect agreement with our intuitive discussion above: lower tension would allow the string to deflect more for a given amount of vibrational energy, thus pushing more air around and more quickly delivering its energy to the air.
Impedance and mode matching
Alright, so what's going on when we attach the string to a guitar body? We argued above that the lower tension string has more efficient sound production because it can move farther to push more air. However, you know this isn't the whole story with the guitar body because you plainly see with your eyes that the guitar body surface does not deflect even nearly as much as the string does. Note that the guitar surface has much more area than the string. This means that for a given velocity, the frictional force is much higher, i.e. $\gamma$ is larger than for the string.
So there you have it: the guitar body has lower $T$ and higher $\gamma$ than the string. These both contribute to making the $Q$ lower, which means that the vibrating guitar body more efficiently transfers energy to the surrounding air than does the bare string.
Lowering the $T$ to be more mechanically compliant like the air is "impedance matching". In general, two modes with similar response to external force (or voltage, or whatever), more efficiently transfer energy between themselves. This is precisely the same principle at work when you use index matching fluid in an immersion microscope to prevent diffraction, or an impedance matching network in a microwave circuit to prevent reflections.
Increasing the area to get larger $\gamma$ is "mode matching". It's called mode matching because you're taking a vibrational mode with a small cross section (the string) and transferring the energy to one with a larger cross section (the guitar body), which better matches the waves you're trying to get the energy into (the concert hall). This is the same reason horn instruments flare from a tiny, mouth sides aperture at one side, to a large, "concert hall" sized aperture at the other end.
[1] I may have messed up a factor of 2 here. It doesn't matter for the point of this calculation.