I looked at the EXIF information. In the 800 ISO shot, the Gain Control setting, it lists Low Gain Up. In the 100 ISO shot, we have Gain Control: None.
So...at 800 ISO, the JPEG processing was done with added Gain control added into the processing mixture.
Interestingly, additionally I see that in the ISO 100 shot, which was made at f/4.5, you have a slight bit of selective focus effect occurring: the foreground bush and its lovely red berries are rendered in crisp focus, but the background of the woods is clearly,and obviously, slightly de-focused. In the 800 ISO shot, made at f/22, the background is more subtly out of focus...it's not really blurry, but it's in-between sharp and de-focused.
Back in the 1980's, when Nikon _invented_ multi-segment light metering and scene analysis, the Nikon FA was fitted with an "8-bit microprocessor", as Nikon boasted, and the FA camera had a memory bank of over 100,000 actual scenes that had been photographed, measure,metered,and analyzed. Things like 3-D, color-aware light metering and dynamic range optimization have advanced markedly since those days if the FA and multi-segment light metering and exposure control. Today the camera not only meters the scene, but processes the final out of camera JPEG image!. One thing I notice is that, in the photo that has the greater depth of field, the f/22 shot done at ISO 800, there is more brightness in the slightly shadowed areas on the fir trees in the background. In other words, because the background is slightly more-recognizable, and more in-focus, I suspect that the JPEG processing has been biased to add a slight bit of brightness to the background due to 1) ISO 800 use and 2)slightly more in-focus, and thus deserving of being "seen". 3) Gain [ brightening-up] being added to the processing
In the f/4.5 shot, which is likely with the lens wide-open at the used 35mm focal length setting on the 18-55 lens, the camera's processing is favoring the foreground exposure, and the background brightness is what one gets with NO gain control added, a pretty straight-up scene analysis and image processing effort. The 100 ISO in use would likely be indicative of a well-lighted scene.
When one starts looking into how Nikon meters with new cameras, they have a color-aware, and distance-aware, "3-D RGB Color Matrix Metering" that uses at the low end cameras, 1,005 sensor measurements, up to many thousands of measurements on higher-priced camera models. The entire scene, the distances, the focused distance, the ISO,. the geographic location, and the time clock...all of those things play a part in the metering and scene analysis. The clock you say? Combined with the city/time zone location the user inputs, it tells the camera night from day; at 12 midnight in Seattle, a large, bright orb in a darker field is either the moon, or a street lamp; at 12 Noon in Seattle a large, bright orb seen against a darker field is...the Sun in the sky.
Reading the colors? White sand? Gray sand? Brown sand? The color-sensing ability of the metering can determine the colors of things AND determine their degree of reflectivity. The camera can then expose, and process, and thus determine how the scene might best be rendered as an out of camera JPEG.
Your shots are different in time by enough that some slight cloud cover over the sun _could_ possibly have altered actual scene brightness, but my experience with Nikon matrix metering is that is is quite affected by larger bright areas; center-weighted Nikon metering is less-affected by bright areas; this is where the slight framing variations Ysarex mentioned could come into play; still, we do not have enough information to really draw conclusive, valid conclusions about why the two shots are different, but my vote goes to ISO levels, 100 vs 800, and no gain versus Low gain Up on the 800 ISO shot, thus brightening it up.
Sorry for the long post on this, but hey...