SummaryVocal-tract resonances (or formants) are acoustic signatures in the voice and are related to the shape and length of the vocal tract. Formants play an important role in human communication, helping us not only to distinguish several different speech sounds [1], but also to extract important information related to the physical characteristics of the speaker, so-called indexical cues. How did formants come to play such an important role in human vocal communication? One hypothesis suggests that the ancestral role of formant perception—a role that might be present in extant nonhuman primates—was to provide indexical cues [2–5]. Although formants are present in the acoustic structure of vowel-like calls of monkeys [3–8] and implicated in the discrimination of call types [8–10], it is not known whether they use this feature to extract indexical cues. Here, we investigate whether rhesus monkeys can use the formant structure in their “coo” calls to assess the age-related body size of conspecifics. Using a preferential-looking paradigm [11, 12] and synthetic coo calls in which formant structure simulated an adult/large- or juvenile/small-sounding individual, we demonstrate that untrained monkeys attend to formant cues and link large-sounding coos to large faces and small-sounding coos to small faces—in essence, they can, like humans [13], use formants as indicators of age-related body size.

Rodríguez-Sierra OE, Turesson HK, Pare D. Contrasting distribution of physiological cell types in different regions of the bed nucleus of the stria terminalis. J Neurophysiol 110: 2037-2049. First published August 7, 2013 doi:10.1152/jn.00408.2013.-We characterized the electroresponsive and morphological properties of neurons in the bed nucleus of the stria terminalis (BNST). Previously, Rainnie and colleagues distinguished three cell types in the anterolateral region of BNST (BNST-AL): low-threshold bursting cells (LTB; type II) and regular spiking neurons that display time-dependent (RS; type I) or fast (fIR; type III) inward rectification in the hyperpolarizing direction (Hammack SE, Mania I, Rainnie DG. J Neurophysiol 98: 638 -56, 2007). We report that the same neuronal types exist in the anteromedial (AM) and anteroventral (AV) regions of BNST. In addition, we observed two hitherto unreported cell types: late-firing (LF) cells, only seen in BNST-AL, that display a conspicuous delay to firing, and spontaneously active (SA) neurons, only present in BNST-AV, firing continuously at rest. However, the feature that most clearly distinguished the three BNST regions was the incidence of LTB cells (approximately 40 -70%) and the strength of their bursting behavior (both higher in BNST-AM and AV relative to AL). The incidence of RS cells was similar in the three regions (ϳ25%), whereas that of fIR cells was higher in BNST-AL (ϳ25%) than AV or AM (Յ8%). With the use of biocytin, two dominant morphological cell classes were identified but they were not consistently related to particular physiological phenotypes. One neuronal class had highly branched and spiny dendrites; the second had longer but poorly branched and sparsely spiny dendrites. Both often exhibited dendritic varicosities. Since LTB cells prevail in BNST, it will be important to determine what inputs set their firing mode (tonic vs. bursting) and in what behavioral states.bed nucleus of the stria terminalis; anxiety; fear; intrinsic properties; morphology DESPITE ANATOMICAL SIMILARITIES (Alheid and Heimer 1988; deOlmos and Heimer 1999;McDonald 2003), dense interconnections (Krettek and Price 1978a,b; Dong et al. 2001a), and functional kinship (Walker et al. 2003) between the amygdala and bed nucleus of the stria terminalis (BNST), there is a stark contrast between our understanding of these two structures. For instance, numerous in vitro studies have examined the physiological properties of amygdala neurons, mechanisms of synaptic transmission, neuromodulation, and activity-dependent plasticity (reviewed in Sah et al. 2003;Pape and Pare 2010). In contrast, relatively few reports on these themes are available for the BNST (reviewed in McElligott and Winder 2009; Hammack et al. 2009). As a result, the operations carried out by the BNST remain poorly understood.

The efficient cortical encoding of natural scenes is essential for guiding adaptive behavior. Because natural scenes and network activity in cortical circuits share similar temporal scales, it is necessary to understand how the temporal structure of natural scenes influences network dynamics in cortical circuits and spiking output. We examined the relationship between the structure of natural acoustic scenes and its impact on network activity [as indexed by local field potentials (LFPs)] and spiking responses in macaque primary auditory cortex. Natural auditory scenes led to a change in the power of the LFP in the 2-9 and 16 -30 Hz frequency ranges relative to the ongoing activity. In contrast, ongoing rhythmic activity in the 9 -16 Hz range was essentially unaffected by the natural scene. Phase coherence analysis showed that scene-related changes in LFP power were at least partially attributable to the locking of the LFP and spiking activity to the temporal structure in the scene, with locking extending up to 25 Hz for some scenes and cortical sites. Consistent with distributed place and temporal coding schemes, a key predictor of phase locking and power changes was the overlap between the spectral selectivity of a cortical site and the spectral structure of the scene. Finally, during the processing of natural acoustic scenes, spikes were locked to LFP phase at frequencies up to 30 Hz. These results are consistent with an idea that the cortical representation of natural scenes emerges from an interaction between network activity and stimulus dynamics.

Automatic classification of vocalization type could potentially become a useful tool for acoustic the monitoring of captive colonies of highly vocal primates. However, for classification to be useful in practice, a reliable algorithm that can be successfully trained on small datasets is necessary. In this work, we consider seven different classification algorithms with the goal of finding a robust classifier that can be successfully trained on small datasets. We found good classification performance (accuracy > 0.83 and F1-score > 0.84) using the Optimum Path Forest classifier. Dataset and algorithms are made publicly available.