Companion Architectures of Sub-Saturns: Distinct Migration Pathways Across the Neptunian Landscape

Abstract

Close-in sub-Saturns (4 - 8.5 R$_\oplus$) are depleted in the Neptunian desert, accumulate in a narrow overdensity near P = 3.2 - 5.7 d (the Neptunian ridge), and thin out into the more moderately populated savanna at longer periods. We test whether sub-Saturns have systematically different companion architectures, as predicted if desert and ridge planets arrived through high-eccentricity migration while savanna planets migrated quiescently. We compile 86 systems with both transit and RV data, construct completeness maps and combine them into detection probability surfaces for companions. The combined completeness maps are used to calculate companion occurrence rates across different companion types with a Poisson-Binomial framework. Companion architectures differ significantly across the landscape. $69.9_{-7.7}^{+6.9}\%$ of savanna sub-Saturns have nearby companions (P < 200 d) compared to only $10.5_{-5.3}^{+7.8}\%$ of desert and ridge sub-Saturns. Additionally, sub-Saturns in the savanna often reside in multi-planet systems with $64.4_{-9.8}^{+8.9}\%$ having more than one companion planet versus only $14.9_{-9.1}^{+13.7}\%$ in the desert and ridge. In both populations, the nearby companions that do exist are almost exclusively small (M < 20 M$_\oplus$), meaning the sub-Saturn is typically the dominant body of its inner system. These contrasts are robust to crosscuts in sub-Saturn radius, bulk density, eccentricity, and host-star properties. Desert and ridge sub-Saturns reside in dynamically emptied systems whose nearby companion rates match those of hot Jupiters, while savanna sub-Saturns inhabit compact multi-planet systems resembling those of warm Jupiters. This parallel supports two migration channels operating within a single population: HEM delivering planets to the desert and ridge, and quiescent disk migration or in-situ formation populating the savanna.