The influence of maximum and mean soil temperatures

The reaction of ants regarding soil temperature are very clear and Pearson’s correlation coefficient is

> 0.3926 (p << 0.001) in any of the four polynomial descriptions (Figs 8A–9B). For maximum calibrated soil temperature T_MAX, ranging in the study plots from 7.4°C to 33.5°C, the maximum species richness of 9.9 species/100 m² and the maximum biomass of 7.2 g/ m² is achieved at 27.5°C (Figs 8A, B). In mean calibrated soil temperature T_MEAN, ranging from 5.7°C to 18.4°C, the maximum species richness of 9.7 species/100 m² and the maximum biomass of 6.2 g/m² are achieved at 16.5°C (Figs 9A, B). These data confirm the view that ants are a very thermophilic insect group. The drop of richness and biomass in extremely hot habitats is probably an effect of too high temperatures for many species but also of increasing scarcity of food resources.

It has to be kept in mind that the habitat temperatures in Tabs 6–7 always refer to a standardized measuring depth of 35 mm below soil or substrate surface. It is reasonable to assume that these data give us a good indication of the within-genus ranking of species in thermal behavior.

Yet, comparative laboratory investigations to check these field data are nearly lacking. Elmes & Wardlaw (1983) investigated the temperature dependency of developmental time of the last larval stage until pupation in Myrmica ruginodis, M. rubra, M. scabrinodis and M.

sabuleti (all four species were correctly determined).

There was a full rank correlation of the integrated developmental temperatures of Elmes & Wardlaw with both the T_MAX and T_MEAN data in Tab. 7.

Conclusions on physiological adaptations of species are possible also from field data if T_MAX and T_MEAN data are considered against the position of the nest centers with the broods relative to surface and if there is knowledge on the temperature behavior of different soil strata in open and woodland habitats (Seifert &

Pannier 2007). This shall be explained in four examples.

Temnothorax and Leptothorax species prefer nest sites in 1–3 cm³ large microspaces placed few millimeters above

or below the soil surface and there is no option to avoid extreme temperatures by vertical movements within the nest throughout the year. The other ant genera in the study construct much deeper, vertically structured nests allowing a flexible up and down movement to levels with more favorable temperature and moisture conditions.

The example of two most different ant species frequently sharing the same macrohabitats shows these between-genus differences: The overall means of T_MAX and T_MEAN of 28.7 and 15.8°C in Temnothorax interruptus and of 29.0 and 16.7°C in Lasius psammophilus are nearly equal but the thermophysiological behavior differs strikingly. During a standard radiation day in a sandy xerothermous grassland with T_MAX = 29.0°C , broods of L. psammophilus will be concentrated in nest chambers in deeper soil layers where temperatures achieve a maximum of 27°C whereas broods and adults in a T. interruptus nest in the moss crust 10 mm below surface will experience a maximum temperature of 41°C – a temperature L. psammophilus larvae would not survive three hours. To give a further example, maximum calibrated soil temperatures in 35 mm depth were 30.9°C in study plot SP 122 – a former limestone quarry almost without vegetation and no humus horizon. Broods and adults of Temnothorax nigriceps nested here 12 mm below surface between loose-fitting limestone plates.

According to micro-thermoelement measurements of 17 July 2006, they survive noon temperatures of at least 50.4°C without any mortality and temperature is expected to drop in this nest spot to 9°C during the coldest days of the summer season. The adaptation of Temnothorax and other small ants of the tribe Formicoxenini to more or less strong circadian temperature amplitudes evolved in a way that these changes are virtually required for brood development (Buschinger 1973).

This situation is contrasted by the oligostenothermous species Stenamma debile which typically nests in topsoil of shady forests with well developed litter layer. The daily temperature in a typical nest site of this species under a basalt stone in study plot SP 164 varied from only 14.6 to 15.4° (0.8 K) while air temperature outside the forest varied from 18.0 to 36.5°C (18.5 K) as measured 30 May 2003. T_MAX and T_MEAN are a good indication for the true temperature conditions in Stenamma debile nests as their depth usually corresponds to the measuring depth for calibrated temperatures. S. debile populations were found within and outside this study in habitats with T_MEAN ranging between 12.5 and 17.0°C. Considering that these data are calibrated seasonal means and that mean soil temperatures increase by about 8°C from May to August in such woodland habitats (Seifert & Pannier 2007), one may expect an upper limit of soil temperature of about 21°C in years with normal macroclimatic conditions.

Figures 8. (A) Relation of maximum calibrated soil temperature and species richness of ant assemblages. (B) Relation of maximum calibrated soil temperature and biomass of ant assemblages

Figures 9. (A) Relation of mean calibrated soil temperature and species richness of ant assemblages. (B) Relation of mean calibrated soil temperature and biomass of ant assemblages.

Above-average nest temperatures during a critical phase of larval development may cause all diploid larvae to develop into workers and, accordingly, to a complete failure of gyne production – this was observed in the warm season of 1997 in most regions of South Germany (Buschinger 1999). Laboratory investigations (Lawitzky 1988) confirm the oligostenothermy of S. debile: The optimum for brood development was 20.5°C at 95 % relative air humidity and a temperature increase above 26°C at 95 % as well as lowering of air humidity below 80 % at 20.5°C caused a rapid emigration to other brood chambers.

A completely different example is Temnothorax tuberum. This ant occurs in the geographic area considered here in open, usually stony, xerothermous grasslands from the planar up to the subalpine zone (2300 m, here only nesting below heat-collecting stones at southern slopes). According to my own unpublished taxonomic investigations, the planar and subalpine populations in the reference area (but not outside!) are conspecific and we have to accept a very wide altitudinal range in a single species. Remarkably, there is a rather low niche width in maximum calibrated soil temperatures (w = 0.39) but a big one in mean calibrated soil temperatures (w = 0.79) over the eleven study plots where T. tuberum was found (Tab. 7). These data show that maximum temperatures are probably more important for niche formation in this species than mean temperatures which vary strongly due to the wide altitudinal range.

Increased niche width data found in the present regional study do also correlate with large-scale geographic distributions. Formica sanguinea is found in Europe from Sicily to North Cape in warm to very hot, more or less sun-exposed habitats and it shows a very wide factorial niche width for both T_MAX (w = 0.69) and T_MEAN (w = 0.61). There is also no morphological indication that this euryoecious species could be split into different species even over the whole range east to Mongolia and Tibet.