The aim of this paper is to quantitatively investigate the spatial and temporal biogeographical relationships of the recovery of ammonoid faunas after the Permian-Triassic mass extinction using three complementary numerical approaches among which is a new, non-hierarchical clustering strategy. The faunal data set consists of a taxonomically homogenised compilation of the spatial and temporal occurrences of ammonoid genera within 20 Early Triassic Tethyan and Panthalassic sites ranging from 40°S to 70°N in palaeolatitudes. In addition to hierarchical cluster analysis (hCA) and nonmetric multidimensional scaling (NMDS), we introduce a third, new non-hierarchical clustering technique allowing the visualisation of a nonmetric interassemblages similarity structure as a connected network constructed without inferring additional internal nodes. The resulting network, which we call a “ Bootstrapped Spanning Network” (BSN), allows the simultaneous identification of partially or totally nested as well as gradational linear or reticulated biogeographical structures. The identified interlocalities relationships indicate that the very beginning of the Early Triassic (Griesbachian) corresponds to a very simple biogeographical context, representing a time of great cosmopolitanism for ammonoids. This context shifts rapidly to a more complex configuration indicative of a more endemic and latitudinally-restricted distribution of the ammonoids during the middle and late Early Triassic (Smithian and Spathian). From an evolutionary dynamic point of view, our results illustrate a very rapid (less than ca. 1.4 myr) Early Triassic recovery of the ammonoid faunas, in contrast to many other marine organisms. This recovery is linked with a marked increase in the overall biogeographical heterogeneity, and parallels the formation of a latitudinal gradient of taxonomic richness, which may be essentially controlled by the progressive intensification of the gradient of sea surface temperature. From a methodological point of view, we show that a BSN is a simple, intuitively legible picture of the nested as well as gradational taxonomic similarity relationships, hence providing a good synthesis (and additional insights) between hierarchical clustering and ordination in reduced space results.