If lighter mode applies the object pixel to the one underneath if it is lighter than the corresponding one. This time colour is involved. You can see that with the greyscale ramp, the roses pixels are the lighter ones more often than with the green ramp.

If darker makes the opposite decision, and uses the roses pixel only if it is darker than the one on the background beneath it. There is very little to choose between the two images this time, though the roses are reproduced very slightly more often with the green ramp than with the grey, which is, as you would expect, the opposite of the previous comparison. If the roses were more often lighter with the greyscale, it is logical that they would be more often darker with the green.

In lighting mode, the pixels are compared on the basis of their lighting. Where the lighting is much the same in both images, the pixel is coloured by calculating the effect of mixing their two colours in lighting. In the first one, pink is being blended with neutral greys, which have little effect on the hue of any light colour they are blended with other than to make it darker, so the pink is relatively unchanged. With the green one, the effect of mixing pink and green lighting would be a yellow colour, and so there is a distinct yellowish shade to this image.

Difference takes the colour value of a pixel in the object, the colour value of the one immediately below it, and colours the pixel with the colour whose value is exactly halfway in between. If you want surreal colour effects, difference is a great favourite of mine.

The addition mode adds the colour values of the corresponding pixels in the object and the layer below it, and uses the result as the colour value for displaying the pixel. It results in a much lighter colour, and since you are adding the lighting effects together, the hues are very much the same as those we saw in the lighting mode.