Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Neural mechanisms of dual-task interference and cognitive capacity limitation in the prefrontal cortex

Nature Neuroscience volume 17, pages 601–611 (2014)Cite this article

Subjects

Abstract

Simultaneous performance of two tasks often leads to performance deficits in the component tasks. This effect, known as dual-task interference, is thought to be a proof of capacity limitation in cognition, and the lateral prefrontal cortex (LPFC) has been highlighted as its putative neural substrate. Here we recorded single-neuron activities in LPFC while monkeys performed dual tasks that required the simultaneous performance of a varying-load spatial attention task and a spatial memory task. We found that the performance of the monkeys exhibited dual-task interference, and prefrontal neuron activities showed a decreased ability to represent task-relevant information to a degree proportional to the increased demand of the concurrent counterpart task. The locus of the interference was shown to originate in the simultaneous, overloaded recruitment of the same LPFC neural population by the two tasks. These results provide direct neurophysiological evidence for, and constraints to, psychological models of dual-task interference and capacity limitation.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Behavioral tasks.
Figure 2: Behavioral performance.
Figure 3: Cue-period activity of a representative neuron in the SMT and four DMT conditions (monkey S, right hemisphere).
Figure 4: Delay-period activities of representative neurons.
Figure 5: Population analyses.
Figure 6: Absence of influence of the spatial congruency between a neuron's maximum response location and an attention target ring position on memory task-related activity.
Figure 7: Neuronal responses against attention task events.
Figure 8: Temporal dynamics of neuronal signals representing attention and memory task information in the standard dual task.

Similar content being viewed by others

References

  1. Duncan, J. The demonstration of capacity limitation. Cognit. Psychol. 12, 75–96 (1980).

    Article  Google Scholar 

  2. Redelmeier, D.A. & Tibshirani, R.J. Association between cellular-telephone calls and motor vehicle collisions. N. Engl. J. Med. 336, 453–458 (1997).

    Article  CAS  Google Scholar 

  3. Strayer, D.L. & Johnston, W.A. Driven to distraction: dual-task studies of simulated driving and conversing on a cellular telephone. Psychol. Sci. 12, 462–466 (2001).

    Article  CAS  Google Scholar 

  4. Pashler, H. Dual-task interference in simple tasks: data and theory. Psychol. Bull. 116, 220–244 (1994).

    Article  CAS  Google Scholar 

  5. Marois, R. & Ivanoff, J. Capacity limits of information processing in the brain. Trends Cogn. Sci. 9, 296–305 (2005).

    Article  Google Scholar 

  6. Baddeley, A. & Hitch, G.J. Working memory. in The Psychology of Learning and Motivation Vol. 8 (ed. Bower, G.H.) 47–89 (Academic Press, New York, 1974).

  7. Telford, C.W. The refractory phase of voluntary and associative responses. J. Exp. Psychol. 14, 1–36 (1931).

    Article  Google Scholar 

  8. Welford, A. The 'psychological refractory period' and the timing of high speed performance–a review and theory. Br. J. Psychol. Gen. Sect. 43, 2–19 (1952).

    Article  Google Scholar 

  9. Kahneman, D. Attention and Effort (Prentice Hall, Englewood Cliffs, New Jersey, 1973).

  10. Wickens, C.D. The structure of attentional resources. in Attention and Performance VIII (ed. Nickerson, R.S.) 239–257 (Lawrence Erlbaum, Hillsdale, New Jersey, 1980).

  11. Wickens, C.D. Multiple resources and performance prediction. Theor. Issues Ergon. 3, 159–177 (2002).

    Article  Google Scholar 

  12. Klingberg, T. Concurrent performance of two working memory tasks: potential mechanisms of interference. Cereb. Cortex 8, 593–601 (1998).

    Article  CAS  Google Scholar 

  13. Adcock, R.A., Constable, R.T., Gore, J.C. & Goldman-Rakic, P.S. Functional neuroanatomy of executive processes involved in dual-task performance. Proc. Natl. Acad. Sci. USA 97, 3567–3572 (2000).

    Article  CAS  Google Scholar 

  14. Bunge, S.A., Klingberg, T., Jacobsen, R.B. & Gabrieli, J.D.E. A resource model of the neural basis of executive working memory. Proc. Natl. Acad. Sci. USA 97, 3573–3578 (2000).

    Article  CAS  Google Scholar 

  15. Passingham, R.E. Attention to action. Phil. Trans. R. Soc. Lond. B 351, 1473–1479 (1996).

    Article  CAS  Google Scholar 

  16. Klingberg, T. & Roland, P.E. Interference between two concurrent tasks is associated with activation of overlapping fields in the cortex. Brain Res. Cogn. Brain Res. 6, 1–8 (1997).

    Article  CAS  Google Scholar 

  17. Baddeley, A., Della Sala, S., Papagno, C. & Spinnler, H. Dual-task performance in dysexecutive and nondysexecutive patients with a frontal lesion. Neuropsychology 11, 187–194 (1997).

    Article  CAS  Google Scholar 

  18. D'Esposito, M. et al. The neural basis of the central executive system of working memory. Nature 378, 279–281 (1995).

    Article  CAS  Google Scholar 

  19. Jiang, Y., Saxe, R. & Kanwisher, N. Functional magnetic resonance imaging provides new constraints on the theories of the psychological refractory period. Psychol. Sci. 15, 390–396 (2004).

    Article  Google Scholar 

  20. Dux, P.E., Ivanoff, J., Asplund, C.L. & Marois, R. Isolation of a central bottleneck of information processing with time-resolved fMRI. Neuron 52, 1109–1120 (2006).

    Article  CAS  Google Scholar 

  21. Funahashi, S., Bruce, C.J. & Goldman-Rakic, P.S. Dorsolateral prefrontal lesions and oculomotor delayed response performance: evidence for mnemonic “scotomas.”. J. Neurosci. 13, 1479–1497 (1993).

    Article  CAS  Google Scholar 

  22. Rossi, A.F., Bichot, N.P., Desimone, R. & Ungerleider, L.G. Top down attentional deficits in macaques with lesions of lateral prefrontal cortex. J. Neurosci. 27, 11306–11314 (2007).

    Article  CAS  Google Scholar 

  23. Zanto, T.P., Rubens, M.T., Thangavel, A. & Gazzaley, A. Causal role of the prefrontal cortex in top-down modulation of visual processing and working memory. Nat. Neurosci. 14, 656–661 (2011).

    Article  CAS  Google Scholar 

  24. Funahashi, S., Bruce, C.J. & Goldman-Rakic, P.S. Mnemonic coding of visual space in the monkey's dorsolateral prefrontal cortex. J. Neurophysiol. 61, 331–349 (1989).

    Article  CAS  Google Scholar 

  25. Constantinidis, C., Franowicz, M.N. & Goldman-Rakic, P.S. The sensory nature of mnemonic representation in the primate prefrontal cortex. Nat. Neurosci. 4, 311–316 (2001).

    Article  CAS  Google Scholar 

  26. Everling, S., Tinsley, C.J., Gaffan, D. & Duncan, J. Filtering of neural signals by focused attention in the monkey prefrontal cortex. Nat. Neurosci. 5, 671–676 (2002).

    Article  CAS  Google Scholar 

  27. Lebedev, M.A., Messinger, A., Kralik, J.D. & Wise, S.P. Representation of attended versus remembered locations in prefrontal cortex. PLoS Biol. 2, e365 (2004).

    Article  Google Scholar 

  28. Desimone, R. & Duncan, J. Neural mechanisms of selective visual attention. Annu. Rev. Neurosci. 18, 193–222 (1995).

    Article  CAS  Google Scholar 

  29. Knudsen, E.I. Fundamental components of attention. Annu. Rev. Neurosci. 30, 57–78 (2007).

    Article  CAS  Google Scholar 

  30. Fougnie, D. & Marois, R. Distinct capacity limits for attention and working memory—evidence from attentive tracking and visual working memory. Psychol. Sci. 17, 526–534 (2006).

    Article  Google Scholar 

  31. Franconeri, S.L., Alvarez, G.A. & Cavanagh, P. Flexible cognitive resources: competitive content maps for attention and memory. Trends Cogn. Sci. 17, 134–141 (2013).

    Article  Google Scholar 

  32. Rao, S.G., Williams, G.V. & Goldman-Rakic, P.S. Destruction and creation of spatial tuning by disinhibition: GABAA blockade of prefrontal cortical neurons engaged by working memory. J. Neurosci. 20, 485–494 (2000).

    Article  CAS  Google Scholar 

  33. Wallis, J.D., Anderson, K.C. & Miller, E.K. Single neurons in prefrontal cortex encode abstract rules. Nature 411, 953–956 (2001).

    Article  CAS  Google Scholar 

  34. Fuster, J.M. & Alexander, G.E. Neuron activity related to short-term memory. Science 173, 652–654 (1971).

    Article  CAS  Google Scholar 

  35. Kubota, K. & Niki, H. Prefrontal cortical unit activity and delayed alternation performance in monkeys. J. Neurophysiol. 34, 337–347 (1971).

    Article  CAS  Google Scholar 

  36. Goldman-Rakic, P.S. Cellular basis of working memory. Neuron 14, 477–485 (1995).

    Article  CAS  Google Scholar 

  37. Curtis, C.E. & D'Esposito, M. Persistent activity in the prefrontal cortex during working memory. Trends Cogn. Sci. 7, 415–423 (2003).

    Article  Google Scholar 

  38. Wang, X.J. Synaptic reverberation underlying mnemonic persistent activity. Trends Neurosci. 24, 455–463 (2001).

    Article  CAS  Google Scholar 

  39. Meyers, E.M., Freedman, D.J., Kreiman, G., Miller, E.K. & Poggio, T. Dynamic population coding of category information in inferior temporal and prefrontal cortex. J. Neurophysiol. 100, 1407–1419 (2008).

    Article  Google Scholar 

  40. Stokes, M.G. et al. Dynamic coding for cognitive control in prefrontal cortex. Neuron 78, 364–375 (2013).

    Article  CAS  Google Scholar 

  41. Zucker, R.S. & Regehr, W.G. Short-term synaptic plasticity. Annu. Rev. Physiol. 64, 355–405 (2002).

    Article  CAS  Google Scholar 

  42. Mongillo, G., Barak, O. & Tsodyks, M. Synaptic theory of working memory. Science 319, 1543–1546 (2008).

    Article  CAS  Google Scholar 

  43. Katsuki, F. & Constantinidis, C. Unique and shared roles of the posterior parietal and dorsolateral prefrontal cortex in cognitive functions. Front. Integr. Neurosci. 6, 17 (2012).

    Article  Google Scholar 

  44. Suzuki, M. & Gottlieb, J. Distinct neural mechanisms of distractor suppression in the frontal and parietal lobe. Nat. Neurosci. 16, 98–104 (2013).

    Article  CAS  Google Scholar 

  45. Rissman, J., Gazzaley, A. & D'Esposito, M. The effect of non-visual working memory load on top-down modulation of visual processing. Neuropsychologia 47, 1637–1646 (2009).

    Article  Google Scholar 

  46. Bourke, P.A., Duncan, J. & Nimmo-Smith, I. A general factor involved in dual-task performance decrement. Q. J. Exp. Psychol. A 49A, 525–545 (1996).

    Google Scholar 

  47. Watanabe, K., Igaki, S. & Funahashi, S. Contribution of prefrontal cue-, delay-, and response-period activity to the decision process of saccade direction in a free-choice ODR task. Neural Netw. 19, 1203–1222 (2006).

    Article  Google Scholar 

  48. Robinson, D.A. A method of measuring eye movement using a scleral search coil in a magnetic field. IEEE Trans. Biomed. Eng. 10, 137–145 (1963).

    CAS  PubMed  Google Scholar 

  49. Bruce, C.J., Goldberg, M.E., Bushnell, M.C. & Stanton, G.B. Primate frontal eye fields. II. Physiological anatomical correlates of electrically evoked eye movements. J. Neurophysiol. 54, 714–734 (1985).

    Article  CAS  Google Scholar 

  50. Page, E.B. Ordered hypotheses for multiple treatments: a significance test for linear ranks. J. Am. Stat. Assoc. 58, 216–230 (1963).

    Article  Google Scholar 

  51. Kutner, M.H., Nachtsheim, C.J., Neter, J. & Li, W. Applied Linear Statistical Models (McGraw-Hill, New York, 2005).

  52. Buschman, T.J., Siegel, M., Roy, J.E. & Miller, E.K. Neural substrates of cognitive capacity limitations. Proc. Natl. Acad. Sci. USA 108, 11252–11255 (2011).

    Article  CAS  Google Scholar 

  53. Kline, R.B. Beyond Significance Testing: Reforming Data Analysis Methods in Behavioral Research (American Psychological Association, Washington DC, 2004).

Download references

Acknowledgements

We thank M. Buckley, J. Duncan, M. Kusunoki and M. Stokes for their comments on the manuscript and R. Akaishi, K. Mochizuki and A. Tanaka for their helpful discussions. This work was supported by Grant-in-Aids for Scientific Research (21240024 and 25240021) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan to S.F. and by Research Fellowships for Young Scientists from the Japan Society for the Promotion of Science to K.W. (20-8015). Two monkeys used for this experiment were supplied from the National Bioresource Project (Japanese Monkeys) supported by MEXT.

Author information

Author notes

  1. Kei Watanabe

    Present address: Present address: Department of Experimental Psychology, University of Oxford, Oxford, UK.,

Authors and Affiliations

  1. Japan Society for the Promotion of Science (JSPS), Tokyo, Japan

    Kei Watanabe

  2. Kokoro Research Center, Kyoto University, Kyoto, Japan

    Kei Watanabe & Shintaro Funahashi

Authors
  1. Kei Watanabe
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Shintaro Funahashi
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

K.W. designed the experiment, collected and analyzed the data and wrote the manuscript. S.F. designed the experiment, supervised all aspects of the project and wrote the manuscript.

Corresponding author

Correspondence to Shintaro Funahashi.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Event sequence of example trials.

(a) Example trials in the standard dual-task. The upper row depicts an example dual-task trial in which the attention task is performed as a short trial. The bottom row shows a trial in which the attention task is performed as a long trial. Note that, in the long trial, there were two possible patterns in the temporal order of catch change and memory cue presentation. (b) Same as in panel (a), but for the easy dual-task. Note that, in the long trial (bottom row), catch change was scheduled and executed as an ‘empty event’. (c) Example trials in the single memory task (SMT). The time course of the task was matched with that in the standard and easy dual-tasks. However, while all attention task events were scheduled, they were executed as ‘empty events’. Trials were automatically initiated by the appearance of FR after an intertrial interval (4.0–7.0 s).

Supplementary Figure 2 Additional evidence supporting the presence of dual-task interference effect in the DMT conditions.

(a) Mean percent correct rates in the SMT and four DMT conditions plotted separately for the trials with short (< 2.0 s), medium (2.0–4.0 s) and long (> 4.0 s) memory delay periods. In monkey S (left), a two-way mixed-design ANOVA showed significant main effects of both Task condition and Delay length (P < 10–4), and a nonsignificant interaction effect (P = 0.27). In monkey A (right), there were significant main effects of both Task condition and Delay length (P < 10–4), and an interaction effect (P = 0.02). Asterisks indicate the result of the simple effect ANOVA for the factor Delay length. (b) Time course of FB error rates relative to memory cue onset averaged across all sessions for monkeys S (left) and A (right). Inset bar graphs indicate the mean FB error rate during the 1-s period following memory cue onset. Error bars indicate s.e.m. During this period, monkey S showed a significant increase in the FB error rate in DMT-Up and DMT-Down compared with SMT (post-hoc Steel-Dwass test, P < 10–4; omnibus Kruskal-Wallis test, P < 10–4), indicating that the oculomotor aspect of memory task performance was substantially interfered with by the concurrent attention task. (c, d) Trajectories and end points of FB eye movements that occurred during the 1-s period following memory cue onset in monkeys S (c) and A (d). Three unfilled black circles indicate the ring stimuli for the attention task. Colors are assigned to each memory cue location (square) and each FB eye movement trajectory so that the color of a given FB eye movement trajectory indicates the location of the memory cue presentation that preceded this FB error. End points of FB eye movements are shown as black dots. Numbers shown on each colored square indicate the cumulative number of FB errors across sessions in each memory cue location. The total number of trials (N) in which the memory cue was presented is shown at the bottom of each panel. In the DMT-Up and DMT-Down conditions in monkey S, regardless of memory cue location, FB eye movements after memory cue onset were predominantly directed toward the attention target ring, rather than the memory cue location as observed in SMT. However, importantly, at the time of memory cue onset, the attention cue had been removed from the monitor for 1.6–5.1 s, and the monkeys were simply viewing the still images of three rings. This indicates that around the time of memory cue presentation, information regarding the position of the target ring for the attention task was occupying monkey S’s processing capacity, suggesting that the level of readiness for memory cue encoding was severely disrupted. Monkey A’s FB errors were characterized by short eye movements clustered on the vertical axis, regardless of the memory task conditions. The similarity in FB eye movement trajectories between the SMT and DMT conditions indicates that in all of the four DMT conditions, the preparatory state for memory cue encoding was not disrupted by the concurrent attention task. Thus, we concluded that, for monkey A, dual-task interference on the oculomotor aspect memory task performance was minimal.

Supplementary Figure 3 Behavioral performance in the modified standard dual-task.

(a) Schematic diagram of the event sequence for non-cued trials of the attention task that were randomly inserted among the Up, Down, FRstd conditions (cued trials). A catch change was scheduled but executed as an ‘empty event’ without actual changes in the display items. (b) Distribution of the session-by-session percent correct rates in the attention task in the modified standard dual-task (consecutive 25 sessions in each monkey) in monkeys S (left) and A (right). Only data from cued trials are shown. The results in FReasy (Fig. 2a) are also shown in the rightmost box plot. The statistical testing procedure and conventions were the same as in Fig. 2a. The results in FReasy (Fig. 2b) are also shown in the rightmost box plot. (c) Distribution of the session-by-session median RTs in the attention task for monkeys S (left) and A (right). (d) Distribution of the session-by-session percent correct rates in the three DMT conditions for monkeys S (left) and A (right). The results in SMT (Fig. 2c) are also shown in the rightmost box plot. The dotted line indicates the mean percent correct rates after “corrected-for-guessing” transformation. P-values were adjusted for three multiple comparisons between the SMT and three DMT conditions. (e) Comparison of the session-by-session percent correct rates between cued (C) and non-cued (NC) trials in the attention task. In monkey S, the percent correct rates in cued trials were significantly higher than those in non-cued trials (two-way repeated-measures ANOVA: main effect of Cueing, P = 2 × 10–4; Attention condition, P < 10–4; interaction, P < 10–4) (simple main effect of Cueing: Up, P < 10–4; FRstd, P = 0.006; both C > NC). (f) Comparison of the session-by-session median lever-release RTs between cued (C) and non-cued (NC) trials in the attention task. In both monkeys, the RTs in cued trials were significantly shorter than those in non-cued trials (monkey S: main effect of Cueing, P = 5 × 10–4; Attention condition, P < 10–4; interaction, P = 0.06; monkey A: main effect of Cueing, P < 10–4; Attention condition, P = 0.17; interaction, P = 0.09) (simple main effect of Cueing: P < 0.009 for the Up and Down conditions in monkey S; P < 2×10–4 for all three conditions in monkey A; all NC > C).

Supplementary Figure 4 Cue-period activity of example neurons.

(a) Activity of a single neuron (monkey S, right hemi.) recorded in the SMTpre, DMT-Up, DMT-Down, DMT-FRstd (standard dual-task), and SMTpost conditions. Conventions as in Fig. 4. (b) Same as in panel (a), but for a neuron recorded from monkey A (left hemi.) This neuron exhibited significant spatial selectivity in all three DMT conditions in the standard dual-task. However, the strength of cue-period activity in the maximum response location (270°) is significantly attenuated in DMT-Up and DMT-Down compared to that in SMTpre. (c) Activity of a single neuron (monkey A, left hemi.) recorded in the SMTpre, DMT-FReasy (easy dual-task), and SMTpost conditions. In DMT-FReasy, attenuation of both the magnitude and selectivity of cue-period activity was absent.

Supplementary Figure 5 Delay-period activity of example neurons.

(a) Activity of a single neuron (monkey S, right hemi.) recorded in the SMTpre, DMT-Up, DMT-Down, DMT-FRstd, DMT-FReasy, and SMTpost conditions. Conventions as in Fig. 4. From left to right, seven memory task conditions, including two SMTpost blocks (SMTpost-1 and SMTpost-2), are shown in the order of recording, except for DMT-Up, DMT-Down and DMT-FRstd. For this and three other neurons (monkey S) with spatially-selective delay-period activity in SMTpre, activities were obtained in both the standard and easy dual-tasks. (a) Activity of a single neuron (monkey A, left hemi.) recorded in SMTpre, DMT-Up, DMT-Down, DMT-FRstd, and SMTpost. Note that in DMT-Down, the activity level was elevated in all memory cue locations because the attention cue had been presented near the neuron’s maximum response location (315°). Nevertheless, spatial selectivity of delay-period activity was lost in this condition (P = 0.75). (b) Activity of a single neuron (monkey A, left hemi.) recorded in SMTpre, DMT-FReasy, and SMTpost. This neuron exhibited delay-period activity similar to that in panel (b). However, attenuation of both the magnitude and selectivity of delay-period activity was absent in DMT-FReasy.

Supplementary Figure 6 Population cue-period activities of individual monkeys.

(a, b) Data are for monkeys S (a) and A (b). Upper row: population cue-period activity in the maximum (blue line) and minimum (red line) response locations across the six memory task conditions. In both monkeys, significant interaction effects [Task condition × Cue location] were observed among cue-period activity (monkey S: main effect of Task condition, F5,187 = 10.85, P < 10–4; Cue location,F1,187 = 114.88, P < 10–4; interaction, F5,187 = 10.52, P < 10–4; monkey A: main effect of Task condition, F5,185 = 0.54, P = 0.76; Cue location,F1,185 = 154.90, P < 10–4; interaction, F5,185 = 3.57, P = 0.004; two-way mixed design ANOVA). Bottom row: population spatial tuning during the cue-period. Conventions as in Fig. 5. Although, in both monkeys, cue-period activity exhibited a significant attenuation of spatial selectivity under DMT, the degree of attenuation was smaller in monkey A, whose cue-period activity exhibited a robust increase following memory cue onset. This suggests that, in monkey A, information processing of the memory task was rather unaffected in the initial phase (time period immediately following memory cue presentation), whereas in monkey S, processing of the memory task was considerably disrupted from this phase. In agreement with this notion, comparison of the behavioral performance between the two monkeys showed that monkey A exhibited a more moderate dual-task interference on memory task performance than monkey S (Fig. 2c). This was particularly evident in the trials that had short memory delay period (< 2.0 s) (Supplementary Fig. 2a). In addition, monkey A’s oculomotor behavior following memory cue presentation did not show signs of dual-task interference, whereas monkey S’s oculomotor behavior in the same epoch clearly exhibited interference by the attention task, as indicated by a significant increase in FB error rates in DMT-Up and DMT-Down relative to SMT (Supplementary Fig. 2b–d).

Supplementary Figure 7 Population delay-period activities of individual monkeys.

(a, b) Data are for monkeys S (a) and A (a). Upper row: population delay-period activity in the maximum (blue line) and minimum (red line) response locations across the six memory task conditions. In both monkeys, significant interaction effects [Task condition × Cue location] were observed among delay-period activity (monkey S: main effect of Task condition, F5,126 = 4.10, P = 0.002; Cue location,F1,126 = 66.05, P < 10–4; interaction, F5,126 = 5.13, P = 3 × 10–4; monkey A: main effect of Task condition, F5,165 = 0.64, P = 0.67; Cue location,F1,165 = 65.64, P < 10–4; interaction, F5,165 = 3.77, P = 0.003). Bottom row: population spatial tuning during the delay-period. Conventions as in Fig. 5. In contrast to the cue-period activity, the degree of selectivity attenuation among delay-period activity was comparable between the two monkeys in the DMT conditions, suggesting that, in both monkeys, memory task processing was substantially disrupted in the later stage by the presence of the concurrent attention task. In accordance with this notion, behavioral results showed that, in both monkeys, prominent dual-task interference was observed in the trials that had long memory delay-period (> 4.0 s) (Supplementary Fig. 2a). The close correspondence between the individual variability among behavioral performance and that among response patterns of cue- and delay-period activities further supports the notion that the attenuation of neuronal selectivity for the memory cue location under DMT is a direct neural correlate of the behavioral cost of dual-task performance.

Supplementary Figure 8 Comparison of cue- and delay-period activities in SMT between neurons assigned to the recording in the standard dual-task and the easy dual-task.

(a,b) Upper row: population cue-period activities in the SMTpre (a) and SMTpost (b) conditions for neurons assigned to the recording in the standard dual-task (left) and the easy dual-task (right). Conventions are the same as in Fig. 5. In both conditions, the activity patterns were highly similar between the groups of neurons assigned to the standard dual-task (left) and the easy dual-task (right) (SMTpre: main effect of Task assignment, F1,96 = 0.83, P = 0.36; Cue location,F1,96 = 170.30, P < 10–4; interaction, F1,96 = 0.06, P = 0.81; SMTpost: main effect of Task assignment, F1,77 = 0.40, P = 0.53; Cue location, F1,77 = 67.03, P < 10–4; interaction, F1,77 = 0.07, P = 0.79; two-way mixed-design ANOVA).Bottom row: population spatial tuning during the cue-period in SMTpre (a) and SMTpost (b). In both SMTpre and SMTpost, the tuning slopes and intercepts did not differ between the assigned task (SMTpre: slope, P = 0.92; intercept, P = 0.44; SMTpost: slope, P = 0.56; intercept, P = 0.60). (c,d) Same as in (a) and (b), but for delay-period activity (SMTpre: main effect of Task assignment, F1,73 = 0.16, P = 0.69; Cue location, F1,73 = 92.18, P < 10–4; interaction, F1,73 = 0.28, P = 0.60; SMTpost: main effect of Task assignment, F1,56 = 0.78, P = 0.38; Cue location, F1,56 = 43.21, P < 10–4; interaction, F1,56 = 1.54, P = 0.22). The tuning slopes and intercepts did not differ between the assigned task (SMTpre: slope, P = 0.65; intercept, P = 0.26; SMTpost: slope, P = 0.49; intercept, P = 0.10).

Supplementary Figure 9 Comparison of single-neuron PEV values between the SMT and DMT conditions.

(a) Upper row: scatter diagrams comparing PEV values of cue-period activity in SMTpre to those in the four DMT and SMTpost conditions. Integration time window for PEV calculation was 0.4 s (0.1–0.5 s from memory cue onset). Blue dashed lines indicate the mean PEV values in SMTpre. Red dashed lines indicate the same neurons’ mean PEV values in the corresponding conditions for comparison. Fractions show the number of neurons that showed a decrease in PEV relative to SMTpre, divided by the number of neurons that exhibited spatially-selective cue-period activity in SMTpre. Bottom row: histograms comparing the distribution of PEV values between SMTpre (blue bars) vs. each of the four DMT and SMTpost conditions (inverted red bars). (b) Same as in (a), but for delay-period activity. Integration time window was 1.0 s (0–1.0 s from memory cue offset). (c) Summary of the five paired-comparisons shown in (a). Note that n = 91 for SMTpre. The center of a notched bar indicates the median value, edges are CI68%, and the error bar is the CI95% of the median (bootstrap method). Open black circles indicate mean values. PEV values for cue-period activity were significantly different across memory task conditions (Kruskal-Wallis test, P = 4 × 10–4), and SMTpre showed a significantly greater PEV value than DMT-Up, DMT-Down, and DMT-FRstd (post-hoc Steel-Dwass test, P < 0.02). All six memory task conditions gave median PEV values significantly larger than zero (one-sample Wilcoxon signed-rank test). (d) Summary of the five paired-comparisons in (b). Note that n = 71 for SMTpre. PEV values in delay-period activity were significantly different across memory task conditions (Kruskal-Wallis test, P = 0.001). SMTpre showed a significantly greater PEV value than DMT-Up, DMT-Down (post-hoc Steel-Dwass test, P < 0.03) and a substantially greater PEV value than DMT-FRstd (P = 0.06). All six memory task conditions gave median PEV values significantly larger than zero.

Supplementary Figure 10 Comparison of memory task-related activity between the three-ring and one-ring layout types in the modified single memory task.

(a) Spatially-selective cue-period activity of a representative neuron that exhibited almost identical activities in the two layout types. Conventions as in Fig. 3. (b) Population activity in the 3-ring (top left) and 1-ring layout types (top right) for 13 spatially-selective cue neurons. A scatter diagram (bottom) shows a comparison of the strength of cue-period activity in the maximum (blue) and minimum (red) response locations that were selected from the five cue locations that were also used in DMT. Dotted lines indicate the mean cue-period activity across the population. The strength of cue-period activity was comparable between the two ring layout types at both the maximum (P = 0.31) and minimum (P = 0.19) response locations (Wilcoxon signed-rank test). (c) Same as in panel (b), but for 10 spatially-selective delay neurons. There was no significant difference in the strength of delay-period activity between the two layout types at both the maximum (P = 0.92) and minimum (P = 0.43) response locations.

Supplementary Figure 11 Temporal dynamics of neuronal signals representing attention and memory task information in the standard and easy dual-tasks.

(a) Time course of neuronal signals of an example neuron (the neuron shown in Supplementary Fig. 5b) representing the location of the attention cue (PEVattention, magenta), the memory cue (PEVmemory, blue), and their interaction (PEVinteraction, green) in the standard dual-task. Dashed cyan line indicates the same neuron’s PEVmemory in SMTpre. Conventions are the same as in Fig. 8a. (b) Time course of neuronal signals of an example neuron (the neuron shown in Supplementary Fig. 5c) representing the location of the memory cue (PEVmemory, blue) in the easy dual-task. Dashed cyan line indicates the same neuron’s PEVmemory obtained in SMTpre. (c) Population-averaged time course of PEVmemory in the easy dual-task (solid blue line, n = 24). Shaded areas indicate s.e.m. The same neurons’ population-averaged PEVmemory time series in SMTpre are plotted as a solid cyan line. Dashed blue line and dashed cyan line indicate population-averaged PEVmemory time series in the standard dual-task and SMTpre, respectively for 51 neurons analyzed in Fig. 8 (the curves are the same as those shown in Fig. 8a). (d) Time course of the proportion of neurons that exhibited significant information (P < 0.05) about the memory cue location (solid blue line). The same neurons’ results in SMTpre are plotted as a solid cyan line. Dashed blue line and dashed cyan line indicate the results of the same analysis in the standard dual-task and SMTpre, respectively (both n = 51, the curves are the same as in Fig. 8f). Horizontal dashed lines indicate the proportion expected by chance (5%).

Supplementary Figure 12 Comparison of spatial selectivity between the SMT and DMT conditions.

(a) Comparison of behavioral performance between the SMTpre sessions with low percent correct rates and the DMT (standard dual-task) sessions with high percent correct rates. To perform this analysis, session-by-session percent correct rates in each memory task condition were rank-ordered and split at the median. The bottom half of SMTpre sessions and the top half of DMT sessions were selected. This analysis included 51 sessions where spatially-selective delay-period activity was recorded in SMTpre. Data from the three DMT conditions in the standard dual-task (DMT-Up, DMT-Down and DMT-FRstd) were collapsed. The sessions from the individual monkey were separately rank-ordered to avoid a biased subsampling from one monkey. The subsampled sessions gave highly similar percent correct rates between SMTpre and DMT (SMTpre: 95.4%, n = 26; DMT: 96.2%, n = 26; Wilcoxon rank-sum test, P = 0.99). Conventions as in Fig. 2c. (b) Time course of PEVattention (magenta), PEVmemory (blue), and PEVinteraction (green) in the standard dual-task for the 26 subsampled sessions. The magnitude of PEVmemory during the delay-period (D) was significantly attenuated relative to that in SMTpre (n = 26, dashed cyan line) (Wilcoxon rank-sum test, P = 0.03). Conventions as in Fig. 8. (c) Comparison of PEVmemory between the pre-Tcol change period and the follow-up fixation period in the standard dual-task. Following the conclusion of the attention task events, PEVmemory in the standard dual-task exhibited significant reawakening. (d) Comparison of PEVmemory and PEVattention during the follow-up fixation period in the standard dual-task. The reawakening of PEVmemory during the follow-up fixation period coincided with the reprioritization of task processing between the attention and memory tasks. (e) Normalized population-averaged delay-period activity (grey shaded area) in the maximum (blue line) and minimum (red line) response locations in SMTpre and the three DMT conditions in the standard dual-task for the subsampled sessions. For comparing delay-period activity across the four conditions, behavioral performance in DMT-Up, DMT-Down and DMT-FRstd were rank-ordered separately. The subsampled sessions gave similar percent correct rates across the four conditions (P = 0.14, n = 26 for each of the four conditions,). For each neuron, firing rate in each 50-ms bin was divided by the peak delay-period firing rate at the maximum response location in the SMTpre condition. Compared with SMTpre, the difference in activity between the maximum and minimum response locations was remarkably attenuated in DMT-Up, DMT-Down and DMT-FRstd (main effect of Task condition, F3,100 = 0.15, P = 0.93; Cue location, F1,100 = 52.95, P < 10–4; interaction, F3,100 = 6.25, P = 6 × 10–4; two-way mixed design ANOVA). Conventions as in Fig. 5c. (f) Comparison of PEV values of the delay-period activity between the SMT and three DMT conditions. The three DMT conditions in the standard dual-task exhibited attenuation in spatial selectivity relative to SMTpre (Kruskal-Wallis test, P = 0.03). Conventions as in Supplementary Fig. 9d. Similar result was obtained when the rank-order of sessions was done over monkey-collapsed data (dotted line, P = 0.03).

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–12 (PDF 11028 kb)

Source data

Rights and permissions

Reprints and permissions

About this article

Cite this article

Watanabe, K., Funahashi, S. Neural mechanisms of dual-task interference and cognitive capacity limitation in the prefrontal cortex. Nat Neurosci 17, 601–611 (2014). https://doi.org/10.1038/nn.3667

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nn.3667

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing