Link 1. Difference in processing lower and upper visual field
Alphabetically arranged citations of scientists who tell us about the difference of processing of the upper (lower retina) and lower (upper hemiretina) visual fields.
The anisotropic properties revealed by the present discrimination task, however, are radially symmetrical in the visual field only within a few degrees from the fovea. Farther away differences are found between the vertical and horizontal meridian This suggest that in addition to the local orientational anisotropies, which are radically symmetrical, there are functional asymmetries in the representation of the vertical and horizontal meridian in the human visual cortex. (Berardi and Fiorentini, 1991)
Color-selective cells are much less common in V3 than VP. This dorso-ventral difference in the distribution of neuronal response properties suggest an significant asymmetry in the way visual information is processed in upper and lower parts of the visual field (Burkhalter and Van Essen,1 986).
There are many reasons that there is something special about the tops of the objects. The top-bottom axis of our bodies and of most objects is correlated with the only asymmetric axis in the world, that of gravity. We typically view the objects from above. We cannot even see the bottom of objects because their tops get in the way and occlude the bottoms.
All written languages and musical notation read top to bottom. Some left to right some right to left, but all begin at the top of the page and proceed downward. Viewers tend to scan the stimuli the tops of figures, alphanumeric symbols, and faces tend to have higher information content. Viewers tend to scan such stimuli longer near the top and move from top down.
Objects in the real world really do have the tendency to remain upright and they really do have tops that are more informative. The occurrence of a generic base bias to experience tops as being more salient reflects how our perceptual mechanisms have internalized these regularities of nature (Chambers et al. 1999).
.. responses to global information were more accurate whereas responses to local information were faster and more accurate in the upper visual field… a left visual or field advantage for global processing terms of both reaction times and accuracy, no direct for local processing in the lower visual field (Christman, 1993) .
We find that attentional resolution is greater in the lower than in the upper visual field, where there is no corresponding asymmetry in the primary visual cortex. We suggest that the attentional filter act in one of higher visual cortical area to restrict the availability of visual information to conscious awareness. The dorsal parietal system and projections from the early visual area to the parietal regions are more numerous for the lower visual field than the upper field. We suggest that the dorsal parietal area may control attentional resolution and the information entering our conscious vision. (He et. al., 1996)
Patients with neglect from right hemispheric lesions omit more lines on the left lower than on the left upper portion of the page. Several lines of evidence from vision research in humans seem to imply an overall functional superiority of the upper hemiretinal system. Upper retinal areas have larger absolute and relative lightpeak amplitudes and sensitivity to change subserves visual functions below the horizontal meridian. The upper hemiretina has more rods and cone per mm2 than its lower of test light intensities. From a teleological point of view the powerful upper hemiretinal system is advantageous since it would ensure better performance in the lower visual field where both predators and prey are present.
The first stage where substantial asymmetry in the upper and lower fields has been reported is in the third visual area V3 (Koch, 1996).
The significance for vision, if any, remains to be determined. By contrast,
the change in rod/cone ratio between the dorsal and ventral halves of the retina
indicates a conspicuous asymmetry in the ground squirrel's visual system,
suggesting a specialization for maximizing visual sensitivity under dim levels
of illumination in the superior visual field. (Kruger et al. 1998)
The difference in P100 between upper and lower field of normal person is possibly related to the fact that the lower field is used more frequently than the upper in ordinary human or animal life, and to the differences in distribution of retinal ganglion cells and in retina-cortex pathway between the upper and lower retina. (Lan et. al. 1996).
In the upper visual field, 500 lux significantly suppressed melatonin, as
compared to 500 lux in the lower visual field or to 5 lux. In the lower visual
field, 500 lux produced intermediate suppression of borderline significance. The
results suggest that bright light treatment of depression or circadian phase
disorders might be most effective when applied in the upper visual field (Lasko,
1999).
In the human V1 and V2 there are little asymmetry between the upper and lower
visual field. Furthermore in the V5 (MT) of macaque lower visual field is
strongly represented than the upper field,, but there is no such difference in
V1, V2 and V4. Our present study support that activity of parieto-occipital
regions measured by N1 amplitude are stronger to LVF peripheral stimuli than UVF
peripheral stimuli(Naito and Osaka, 1999).
Overall, the results of both experiments suggests that left-right distinction based on categorical and coordinate spatial relations may be extended to the specialization of the upper and lower visual fields. Finally there is some evidence for lower and left visual field advantages in the processing of lower spatial frequencies.. The consistency with which the lower and left visual fields processing versus upper and lower right field processing are links suggests something beyond just coincidence. Given that left versus right visual field differences primarily (although not exclusively) reflect hemispheric differences, whereas upper versus lower differences appear to be more functional in nature and are readily modulated by environmental factors. (Niebauer and Christman, 1998)
Responses to hemifield stimuli that extended to both lower and upper fields
resembled the responses to lower field stimuli in source current direction and
strength. These results agree with psychophysical data, which indicate lower
visual field advantage in complex visual processing. Parieto-occipital responses
in the putative V6 complex were similar to lower and upper field stimuli
(Portin, 1999).
In addition to the well-known anatomical break between left and right hemisphere, there is a wide separation between upper and lower hemifield representations between upper and lower hemifield in extrastriate cortex, which has received little attention so far.
The segmentation of an image into figures and background is shown to be performed much better in the lower visual field. This asymmetry indicates a neural specialization that may be related to the anatomical discontinuity along the representation of the horizontal meridian in extrastriate cortex. From an ecological point of view, specialization in scene segmentation in the lower field may be the result of a more frequent occurrence of occlusion situations there or of a greater survival importance for segmentation in the ground plane. This raises the possibility that the different specialization in scene segmentation reported here may in turn be related to functional specialization of higher visual areas in the ventral and dorsal systems streams. Because the lower and upper parts of early visual areas project more to the ventral and dorsal parts outside the occipital lobe (Rubin et al. 1996).
An important component of the organizational process, the segmentation of an
image into figures and background, is shown to be performed much better in the
lower visual field This asymmetry indicates a neural specialization that may be
related to the anatomical discontinuity along the representation of the
horizontal meridian in extrastriate visual cortex. (Rubin, 1996).
The recognition potential RP is an electrical response of the brain that occurs when a subject views recognizable image such as words pictures or faces. For the recognition potential, upper and lower field stimulation produced radically different profiles (Rudell et al. 1993).
Our findings show that upper retinal areas display a higher level of light-induced activity reflecting the interaction between the photoreceptors and the retinal pigment epithelium than lower retinal areas. The results are interpreted as a superiority of the upper over the lower retina and are related to other electrophysiological and functional differences between upper and lower retinal areas of man (Skradies and Baier ,1998).
An analysis of the potential convergence and divergence between rod bipolar cells and AII amacrine cells in the rabbit retina indicates that the neuronal architecture of the rod circuit is not organized in a uniform module that is simply scaled-up from central to peripheral retina. Moreover, peripheral fields in the superior and inferior retina that have equivalent densities of interneurons show markedly different rod bipolar----AII amacrine convergence ratios, with the result that many more rod photoreceptors converge on an AII amacrine cell in the superior retina than in the inferior retina. (Varey et al. 1991)
Region that appear in the lower visual field tend to be perceived as figure, suggesting that lower region may be a new gestalt principle of figure-ground segregation. The lower region bias may be the result of environmental constraints and the "near" depth perception processes associate with the lower visual field and the dorsal visual pathway. The visual system may be biased to see regions in the lower visual field as figure because objects in the lower visual field are usually closer to viewer. (Vecera et al. 1999).

Figure 1. Distribution of ganglion cells (somatostatin-immunoreactive small and large cells) in the retina of an adult cat preferentially in the inferior retina (upper visual field) (Chalupa and White, 1990)
The ganglion cell layer of the adult cat retina, subgroups of displaced
amacrine cells and alpha ganglion cells are immunoreactive for somatostatin or a
somatostatin like substance. Both types of immunoreactive cells are found
preferentially in inferior retina By contrast, the change in rod/cone ratio
between the dorsal and ventral halves of the retina indicates a conspicuous
asymmetry in the ground squirrel's visual system, suggesting a specialization
for maximizing visual sensitivity under dim levels of illumination in the
superior visual field"
(White and Chalupa 1992).