Available online at www.sciencedirect.com Behavioural Brain Research 187 (2007) 159–171 Research report The effects of hippocampal system lesions on a novel temporal discrimination task for rats R.J. Kyd a , J.M. Pearce b , M. Haselgrove c , E. Amin b , J.P. Aggleton b,∗ a Department of Pharmacology, Cambridge University, Tennis Court Road, Cambridge CB2 1PD, UK b School of Psychology, Cardiff University, Park Place, Cardiff CF10 3AT, UK c School of Psychology, University of Nottingham, University Park, Nottingham NG7 2RD, UK Received 14 June 2007; received in revised form 30 August 2007; accepted 7 September 2007 Available online 14 September 2007 Abstract A novel, appetitive, Pavlovian conditioning task was used to assess interval timing. Experiment 1 showed that normal rats could discriminate between tones of 1.5 s and 0.5 s duration, or between tones of 12.0 s and 3.0 s duration. Learning was demonstrated by a greater duration of magazine responding in the period before the delivery of a food reward and after cessation of the CS+ compared to the same time period after cessation of the CS-. Learning was, however, asymmetric as it was much quicker when the CS+ was the longer of the two durations (1.5 s and 12.0 s, respectively). Experiment 2 assessed the impact of fornix lesions on the acquisition of one version of this task (CS+ 1.5 s, CS- 0.5 s). No evidence was found of a change in discrimination learning following surgery. Experiment 3 examined whether rats with either fornix or hippocampal lesions affected discriminations between 12.0 s and 3.0 s stimuli. Again, there was no evidence of a lesion-induced deficit. T-maze alternation training confirmed the effectiveness of these lesions. The results not only reveal that neither the fornix nor the hippocampus is necessary for distinguishing temporal intervals within the ranges tested but also showed how under some circumstances these lesions can leave trace conditioning intact. © 2007 Elsevier B.V. All rights reserved. Keywords: Hippocampus; Fornix; Rat; Timing; Trace conditioning 1. Introduction A fundamental aspect of time perception is the ability to mea- sure intervals from seconds to minutes. This ability is regarded as a central element of cognition [5,39], and various models have been proposed to explain the nature of such interval tim- ing. Influential models based around scalar expectancy theory, e.g. the pacemaker–accumulator model [5,16,17] assume that interval timing involves at least two distinct cognitive com- ponents. The first is an internal-clock process (pacemaker), while the second is a device used to compare a reference memory of previous outputs of the pacemaker against the current output to determine the interval between a start and stop signal. Monitoring the current output is often assumed to use working memory [4,5,18,53]. Support for the notion of at least two distinct processes includes evidence that the pacemaker (clock) relies on dopaminergic systems, while the ∗ Corresponding author. Tel.: +44 29 2087 4563; fax: +44 29 2087 4858. E-mail address: aggleton@cardiff.ac.uk (J.P. Aggleton). reference memory component relies on cholinergic mechanisms [5,38–40]. A variety of brain structures have been implicated in tim- ing. Prominent among these structures are the striatum and the prefrontal cortex [5,21,32,36,38,39]. It is also plausible that the hippocampus should play a key role [27], particularly in view of its connections, via the fornix, with both the striatum and the prefrontal cortex, and its critical involvement in some forms of memory. In spite of these links, the role of the hippocampus in interval timing remains uncertain. Neuropsychological studies of interval timing by amnesics with bilateral medial temporal damage (and hence hippocampal pathology) show that inter- val timing can remain intact for intervals of up to 20 s [50,54], although patients with right temporal lobe damage (including the hippocampus) show changes in decision-making processes for temporal duration [43]. While the famous amnesic HM appears to underestimate time intervals longer than 20 s [50], this deficit has been interpreted as a failure to encode or retrieve the task demands [12]. More reliable deficits have been reported for longer durations (1–8 min) in patients with unilateral medial temporal lobe lesions [44] that included the hippocampus. 0166-4328/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bbr.2007.09.010